The Trump administration has effectively unwound the Biden era chip controls by selling the H200 to China; I agree with the decision, which is a return to longstanding U.S. policy.
The Dutch Raad van State in 2019 has argued that a nitrogen deposition of 5.09 mol per acre per year is damaging De Heide (Heath) too much. This is the same as putting down about one one grain of fertilizer the size of a sugar grain per two square meters per week. How ridiculous this may sound, this verdict has blocked thousands of farmers and builders from expanding their business or homes, and even caused many farms to close down. Furthermore, many farmers in the Netherlands, which have often been farmers for many generations, are not sure whether they will be allowed to continue farming.
The gap between Europe's self-image and reality has grown into a chasm of delulu. One that's threatening to swallow the continent's future whole, as dangerous dependencies on others for energy, security, software, and manufacturing stack up to strangle Europe's sovereignty. But its current political class continues to double down on everything that hasn't worked for the past forty years. Let's start with free speech, and the €120 million fine just levied against X. The fig leaf for this was painted as "deceptive design" and "transparency for researchers", but the EU already bared its real intentions when they announced this authoritarian quest back in 2023 with charges of "dissemination of illegal content" and "information manipulation" (aka censorship). Besides, even the fig leaf itself is rotten. Meta offers the very same paid verification scheme as X but, according to Musk, has chosen to play ball with the EU censorship apparatus, so no investigation for them. And the citizens of Europe clearly don't seem bothered much by any "deceptive design", as X continues to be a top-ranked download across every country on the continent. But you can see why many politicians in Europe are eager to punish X for giving Europeans a social media that doesn't cooperate with its crackdown on wrongthink. The German chancellor, Friedrich Merz, is personally responsible for 5,000(!!) cases pursuing his subjects for insults online, which has led to house raids for utterances as banal as calling him a "filthy drunk". Germany is not an outlier either. The UK has been arresting over 10,000 people per year since 2020 for illicit tweets, Facebook posts, and silent prayers. France has thousands of yearly cases for speech-related offenses too. No wonder people on X aren't eager to volunteer their name and address when their elected officials crash out over their tweets. It's against this backdrop — thousands of yearly arrests for banal insults or crass opposition to government policies — that some Europeans still try to convince themselves they're the true champions of free speech and freedom of the press. Delulu indeed. But this isn't just about the lack of free speech in Europe. The X fine also highlights just how weak and puny the European tech sector has become. Get this: The EU's tech-fine operation produced more income for European coffers than all the income taxes paid by its public internet tech companies in 2024!! That's primarily because Europe basically stopped creating new, large companies more than half a century ago. So as the likes of Nokia died off, there was nobody new to replace them. In the last fifty years, the number and size of new European companies worth $10 billion or more is alarmingly small: But even the old industrial titans of Europe are now struggling. Germany hasn't grown its real GDP in five years. The net-zero nonsense has seriously hurt its competitiveness, and its energy costs are now 2-3x that of America and China. This is after Germany spent a staggering ~€700 billion on green energy projects — despite Europe as a whole being just 6% of world emissions. All the while, the EU as a whole sent over twenty billion euros to Russia to pay for energy in 2024. So cue the talk about security. European leaders are incensed by getting excluded from the discussion about ending the war in Ukraine, which is currently just happening between America and Russia directly. But they only have themselves to thank for a seat on the sidelines. Here's a breakdown of the NATO spending by country: This used to be a joke to Europeans. That America would spend so much on its military might. Since the invasion of Ukraine, there's been a lot less laughing, and now the new official NATO target for member states is to spend 5% of GDP on defense. But even this target fails to acknowledge the fact that even if European countries should meet their new obligations (and currently only Poland among the larger EU countries is even close), they'd still lag far behind America, simply because the EU is comparatively a much smaller and shrinking economic zone. In 2025, the combined GDP for the European Union was $20 trillion. America was fifty percent larger with a GDP of $30 trillion. And the gap continues to widen, as EU growth is pegged at around 1% in 2024 compared to almost 3% for the US. Now this is usually when the euro cope begins to screech the loudest. Trying every which way to explain that actually Europe is a better place to live than America, despite having a GDP per capita that's almost half. And on a subjective level, that might well be true! There are plenty of reasons to prefer living in Europe, but that doesn't offset the fact that America is simply a vastly richer country, and that matters when it comes to everything from commercial dominance to military power. But it's the trajectory that's most damning. In 2008, Europe was on near-parity in GDP with America! But if the 1% vs 3% growth-rate disparity continues for another decade, America will grow its economy by another third to $40 trillion, while Europe will grow just 10% to $22 trillion. Making the American economy nearly twice as large as the European one. Yikes. These should all be sobering numbers to any European. Whether it's the 10,000 yearly arrests in the UK for social media posts or the risk of an economy that's half the size of the American one in a decade. But Europe isn't doomed to fulfill this tragic destiny. It's full of some of the most creative, capable, and ambitious people in the world (like the fifth of US startup unicorns with European founders!). But they need much better reasons to stay than what the EU (and now a separate UK) is currently giving them. Like drastically lower energy costs to for a competitive industrial base and to power the AI revolution, so best we quickly revive European nuclear ambitions. Like an immigration policy designed to rival America's cherry-picking of the world's best, rather than mass immigration from low-average-IQ regions of net-negative contributors to the economy (and society). Like dropping the censorship ambitions and bureaucratic boondoggles like the DSA. Like actually offering a European internal market for remote labor and a unified stock exchange for listings. There are plenty of paths to take that do not end in a low-growth, censorious regime that continues to export many of its best brains to America and elsewhere. So: make haste, the shadows lengthen.
An interview with MoffettNathanson's Michael Nathanson about Netflix's acquisition of Warner Bros. and the Hollywood end game.
Elon Musk is not happy with the EU fining his X platform and is currently on a tweet rampage complaining about it. Among other things, he wants the whole EU to be abolished. He sadly is hardly the first wealthy American to share their opinions on European politics lately. I’m not a fan of this outside attention but I believe it’s noteworthy and something to pay attention to. In particular because the idea of destroying and ripping apart the EU is not just popular in the US; it’s popular over here too. Something that greatly concerns me. There is definitely a bunch of stuff we might want to fix over here. I have complained about our culture before. Unfortunately, I happen to think that our challenges are not coming from politicians or civil servants, but from us, the people. Europeans don’t like to take risks and are quite pessimistic about the future compared to their US counterparts. Additionally, we Europeans have been trained to feel a lot of guilt over the years, which makes us hesitant to stand up for ourselves. This has led to all kinds of interesting counter-cultural movements in Europe, like years of significant support for unregulated immigration and an unhealthy obsession with the idea of degrowth. Today, though, neither seems quite as popular as it once was. Morally these things may be defensible, but in practice they have led to Europe losing its competitive edge and eroding social cohesion. The combination of a strong social state and high taxes in particular does not mix well with the kind of immigration we have seen in the last decade: mostly people escaping wars ending up in low-skilled jobs. That means it’s not unlikely that certain classes of immigrants are going to be net-negative for a very long time, if not forever, and increasingly society is starting to think about what the implications of that might be. Yet even all of that is not where our problems lie, and it’s certainly not our presumed lack of free speech. Any conversation on that topic is foolish because it’s too nuanced. Society clearly wants to place some limits to free speech here, but the same is true in the US. In the US we can currently see a significant push-back against “woke ideologies,” and a lot of that push-back involves restricting freedom of expression through different avenues. The US might try to lecture Europe right now on free speech, but what it should be lecturing us on is our economic model. Europe has too much fragmentation, incredibly strict regulation that harms innovation, ineffective capital markets, and a massive dependency on both the United States and China. If the US were to cut us off from their cloud providers, we would not be able to operate anything over here. If China were to stop shipping us chips, we would be in deep trouble too ( we have seen this ). This is painful because the US is historically a great example when it comes to freedom of information, direct democracy at the state level, and rather low corruption. These are all areas where we’re not faring well, at least not consistently, and we should be lectured. Fundamentally, the US approach to capitalism is about as good as it’s going to get. If there was any doubt that alternative approaches might have worked out better, at this point there’s very little evidence in favor of that. Yet because of increased loss of civil liberties in the US, many Europeans now see everything that the US is doing as bad. A grave mistake. Both China and the US are quite happy with the dependency we have on them and with us falling short of our potential. Europe’s attempt at dealing with the dependency so far has been to regulate and tax US corporations more heavily. That’s not a good strategy. The solution must be to become competitive again so that we can redirect that tax revenue to local companies instead. The Digital Services Act is a good example: we’re punishing Apple and forcing them to open up their platform, but we have no company that can take advantage of that opening. If you read my blog here, you might remember my musings about the lack of clarity of what a foreigner is in Europe. The reality is that Europe has been deeply integrated for a long time now as a result of how the EU works — but still not at the same level as the US. I think this is still the biggest problem. People point to languages as the challenge, but underneath the hood, the countries are still fighting each other. Austria wants to protect its local stores from larger competition in Germany and its carpenters from the cheaper ones coming from Slovenia. You can replace Austria with any other EU country and you will find the same thing. The EU might not be perfect, but it’s hard to imagine that abolishing it would solve any problem given how national states have shown to behave. The moment the EU fell away, we would be warming up all border struggles again. We have already seen similar issues pop up in Northern Ireland after the UK left. And we just have so much bureaucracy, so many non-functioning social systems, and such a tremendous amount of incoming governmental debt to support our flailing pension schemes. We need growth more than any other bloc, and we have such a low probability of actually accomplishing that. Given how the EU is structured, it’s also acting as the punching bag for the failure of the nation states to come to agreements. It’s not that EU bureaucrats are telling Europeans to take in immigrants, to enact chat control or to enact cookie banners or attached plastic caps. Those are all initiatives that come from one or more member states. But the EU in the end will always take the blame because even local politicians that voted in support of some of these things can easily point towards “Brussels” as having created a problem. A Europe in pieces does not sound appealing to me at all, and that’s because I can look at what China and the US have. What China and the US have that Europe lacks is a strong national identity. Both countries have recognized that strength comes from unity. China in particular is fighting any kind of regionalism tooth and nail. The US has accomplished this through the pledge of allegiance, a civil war, the Department of Education pushing a common narrative in schools, and historically putting post offices and infrastructure everywhere. Europe has none of that. More importantly, Europeans don’t even want it. There is a mistaken belief that we can just become these tiny states again and be fine. If Europe wants to be competitive, it seems unlikely that this can be accomplished without becoming a unified superpower. Yet there is no belief in Europe that this can or should happen, and the other superpowers have little interest in seeing it happen either. If I had to propose something constructive, it would be this: Europe needs to stop pretending it can be 27 different countries with 27 different economic policies while also being a single market. The half-measures are killing us. We have a common currency in the Eurozone but no common fiscal policy. We have freedom of movement but wildly different social systems. We have common regulations but fragmented enforcement. 27 labor laws, 27 different legal systems, tax codes, complex VAT rules and so on. The Draghi report from last year laid out many of these issues quite clearly: Europe needs massive investment in technology and infrastructure. It needs a genuine single market for services, not just goods. It needs capital markets that can actually fund startups at scale. None of this is news to anyone paying attention. But here’s the uncomfortable truth: none of this will happen without Europeans accepting that more integration is the answer, not less. And right now, the political momentum is in the opposite direction. Every country wants the benefits of the EU without the obligations. Every country wants to protect its own industries while accessing everyone else’s markets. One of the arguments against deeper integration is that Europe hinges on some quite unrelated issues. For instance, the EU is seen as non-democratic, but some of the criticism just does not sit right with me. Sure, I too would welcome more democracy in the EU, but at the same time, the system really is not undemocratic today. Take things like chat control: the reason this thing does not die, is because some member states and their elected representatives are pushing for it. What stands in the way is that the member countries and their people don’t actually want to strengthen the EU further. The “lack of democracy” is very much intentional and the exact outcome you get if you want to keep the power with the national states. So back to where we started: should the EU be abolished as Musk suggests? I think this is a profoundly unserious proposal from someone who has little understanding of European history and even less interest in learning. The EU exists because two world wars taught Europeans that nationalism without checks leads to catastrophe. It exists because small countries recognized they have more leverage negotiating as a bloc than individually. I also take a lot of issue with the idea that European politics should be driven by foreign interests. Neither Russians nor Americans have any good reason for why they should be having so much interest in European politics. They are not living here; we are. Would Europe be more “free” without the EU? Perhaps in some narrow regulatory sense. But it would also be weaker, more divided, and more susceptible to manipulation by larger powers — including the United States. I also find it somewhat rich that American tech billionaires are calling for the dissolution of the EU while they are greatly benefiting from the open market it provides. Their companies extract enormous value from the European market, more than even local companies are able to. The real question isn’t whether Europe should have less regulation or more freedom. It’s whether we Europeans can find the political will to actually complete the project we started. A genuine federation with real fiscal transfers, a common defense policy, and a unified foreign policy would be a superpower. What we have now is a compromise that satisfies nobody and leaves us vulnerable to exactly the kind of pressure Musk and other oligarchs represent. Europe doesn’t need fixing in the way the loud present-day critics suggest. It doesn’t need to become more like America or abandon its social model entirely. What it needs is to decide what it actually wants to be. The current state of perpetual ambiguity is unsustainable. It also should not lose its values. Europeans might no longer be quite as hot on the human rights that the EU provides, and they might no longer want to have the same level of immigration. Yet simultaneously, Europeans are presented with a reality that needs all of these things. We’re all highly dependent on movement of labour, and that includes people from abroad. Unfortunately, the wars of the last decade have dominated any migration discourse, and that has created ground for populists to thrive. Any skilled tech migrant is running into the same walls as everyone else, which has made it less and less appealing to come. Or perhaps we’ll continue muddling through, which historically has been Europe’s preferred approach. It’s not inspiring, but it’s also not going to be the catastrophe the internet would have you believe either. Is there reason to be optimistic? On a long enough timeline the graph goes up and to the right. We might be going through some rough patches, but structurally the whole thing here is still pretty solid. And it’s not as if the rest of the world is cruising along smoothly: the US, China, and Russia are each dealing with their own crises. That shouldn’t serve as an excuse, but it does offer context. As bleak as things can feel, we’re not alone in having challenges, but ours are uniquely ours and we will face them. One way or another.
At the end of November, NVIDIA put out an internal memo ( that was leaked to Barron's reporter Tae Kim, who is a huge NVIDIA fan and knows the company very well , so take from that what you will) that sought to get ahead of a few things that had been bubbling up in the news, a lot of which I covered in my Hater’s Guide To NVIDIA (which includes a generous free intro). Long story short, people have a few concerns about NVIDIA, and guess what, you shouldn’t have any concerns, because NVIDIA’s very secret, not-to-be-leaked-immediately document spent thousands of words very specifically explaining how NVIDIA was fine and, most importantly, nothing like Enron . Anyway, all of this is fine and normal . Companies do this all the time, especially successful ones, and there is nothing to be worried about here , because after reading all seven pages of the document, we can all agree that NVIDIA is nothing like Enron. No, really! NVIDIA is nothing like Enron, and it’s kind of weird that you’re saying that it is! Why would you say anything about Enron? NVIDIA didn’t say anything about Enron. Okay, well now NVIDIA said something about Enron, but that’s because fools and vagabonds kept suggesting that NVIDIA was like Enron, and very normally, NVIDIA has decided it was time to set the record straight. And I agree! I truly agree. NVIDIA is nothing like Enron. Putting aside how I might feel about the ethics or underlying economics of generative AI, NVIDIA is an incredibly successful business that has incredible profits, holds an effective monopoly on CUDA ( explained here ), which powers the underlying software layer to running software on GPUs, specifically generative AI, and not really much else that has any kind of revenue potential. And yes, while I believe that one day this will all be seen as one of the most egregious wastes of capital of all time, for the time being, Jensen Huang may be one of the most successful salespeople in business history. Nevertheless, people have somewhat run away with the idea that NVIDIA is Enron , in part because of the weird, circular deals it’s built with Neoclouds — dedicated AI-focused cloud companies — like CoreWeave, Lambda and Nebius , who run data centers full of GPUs sold by NVIDIA, which they then use as collateral for loans to buy more GPUs from NVIDIA . Yet as dodgy and weird and unsustainable as this is, it isn’t illegal , and it certainly isn’t Enron, because, as NVIDIA has been trying to tell you, it is nothing like Enron! Now, you may be a little confused — I get it! — that NVIDIA is bringing up Enron at all. Nobody seriously thought that NVIDIA was like Enron before (though JustDario, who has been questioning its accounting practices for years , is a little suspicious), because Enron was one of the largest criminal enterprises in history, and NVIDIA is at worst, I believe, a big, dodgy entity that is doing whatever it can to survive. Wait, what’s that? You still think NVIDIA is Enron ? What’s it going to take to convince you? I just told you NVIDIA isn’t Enron! NVIDIA itself has shown it’s not Enron, and I’m not sure why you keep bringing up Enron all the time! Stop being an asshole. NVIDIA is not Enron! Look, NVIDIA’s own memo said that “NVIDIA does not resemble historical accounting frauds because NVIDIA's underlying business is economically sound, [its] reporting is complete and transparent, and [it] cares about [its] reputation for integrity.” Now, I know what you’re thinking. Why is the largest company on the stock market having to reassure us about its underlying business economics and reporting? One might immediately begin to think — Streisand Effect style — that there might be something up with NVIDIA’s underlying business. But nevertheless, NVIDIA really is nothing like Enron. But you know what? I’m good. I’m fine. NVIDIA, grab your coat, we’re going out, let’s forget any of this ever happened. Wait, what was that? First, unlike Enron, NVIDIA does not use Special Purpose Entities to hide debt and inflate revenue. NVIDIA has one guarantee for which the maximum exposure is disclosed in Note 9 ($860M) and mitigated by $470M escrow. The fair value of the guarantee is accrued and disclosed as having an insignificant value. NVIDIA neither controls nor provides most of the financing for the companies in which NVIDIA invests. Oh, okay! I wasn’t even thinking about that at all, I was literally just saying how you were nothing like Enron , we’re good. Let’s go home- Second, the article claims that NVIDIA resembles WorldCom but provides no support for the analogy. WorldCom overstated earnings by capitalizing operating expenses as capital expenditures. We are not aware of any claims that NVIDIA has improperly capitalized operating expenses. Several commentators allege that customers have overstated earnings by extending GPU depreciation schedules beyond economic useful life. Rebutting this claim, some companies have increased useful life estimates to reflect the fact that GPUs remain useful and profitable for longer than originally anticipated; in many cases, for six years or more. We provide additional context on the depreciation topic below. I…okay, NVIDIA is also not like WorldCom either. I wasn’t even thinking about WorldCom. I haven’t thought of them in a while. Per Adam Berger of Ebsco : …NVIDIA, are you doing something WorldCommy? Why are you bringing up WorldCom? To be clear, WorldCom was doing capital F fraud , and its CEO Bernie Ebbers went to prison after an internal team of auditors led by WorldCom VP of internal auditing Cynthia Cooper reported $3.8 billion in “misallocated expenses and phony accounting entries.” So, yeah, NVIDIA, you were really specific about saying you didn’t capitalize operating expenses as capital expenditures. You’re…not doing that, I guess? That’s great. Great stuff. I had literally never thought you had done that before. I genuinely agree that NVIDIA is nothing like WorldCom. Anyway, also glad to hear about the depreciation stuff, looking forward to reading- Third, unlike Lucent, NVIDIA does not rely on vendor financing arrangements to grow revenue. In typical vendor financing arrangements, customers pay for products over years. NVIDIA's DSO was 53 in Q3. NVIDIA discloses our standard payment terms, with payment generally due shortly after delivery of products. We do not disclose any vendor financing arrangements. Our customers are subject to strict credit evaluation to ensure collectability. NVIDIA would disclose any receivable longer than one year in long-term other assets. The $632M "Other" balance as of Q3 does not include extended receivables; even if it did, the amount would be immaterial to revenue. Erm… Alright man, if anyone asks about whether you’re like famed dot-com crashout Lucent Technologies, I’ll be sure to correct them. After all, Lucent’s situation was really different — well…sort of. Lucent was a giant telecommunications equipment company, one that was, for a time, extremely successful, really really successful, in fact, turned around by the now-infamous Carly Fiorina. From a 2010 profile in CNN : NVIDIA, this sounds great — why wouldn’t you want to be compared to Lucen- Oh. So, to put it simply, Lucent was classifying debt as an asset (we're getting into technicalities here, but it sort of was but was really counting money from loans as revenue, which is dodgy and bad and accountants hate it ), and did something called “vendor financing,” which means you lend somebody money to buy something from you. It turns out Lucent did a lot of this. Okay, NVIDIA, I hate to say this, but I kind of get why somebody might say you’re doing Lucent stuff. After all, rumour has it that your deal with OpenAI — a company that burns billions of dollars a year — will involve it leasing your GPUs , which sure sounds like you’re doing vendor financing... -we do not disclose any vendor financing arrangements- Fine! Fine. Anyway, Lucent really fucked up big time, indulging in the dark art of circular vendor financing. In 1998 it signed its largest deal — a $2 billion “equipment and finance agreement” — with telecommunications company Winstar , which promised to bring in “$100 million in new business over the next five years” and build a giant wireless broadband network, along with expanding Winstar’s optical networking. To quote The Wall Street Journal : In December 1999, WIRED would say that Winstar’s “small white dish antennas…[heralded] a new era and new mind-set in telecommunications,” and included this awesome quote about Lucent from CEO and founder Will Rouhana: Fuck yeah! But that’s not the only great part of this piece: Annualized revenues, very nice. We love annualized revenues don't we folks? A company making about $25 million a month a year after taking on $2 billion in financing from Lucent. Weirdly, Winstar’s Wikipedia page says that revenues were $445.6 million for the year ending 1999 — or around $37.1 million a month. Winstar loved raising money — two years later in November 2000, it would raise $1.02 billion, for example — and it raised a remarkable $5.6 billion between February 1999 and July 2001 according to the Wall Street Journal. $900 million of that came in December 1999 from an investment from Microsoft and “several investment firms,” with analyst Greg Miller of Jefferies & Co saying: Another fun thing happened in November 2000 too. Lucent would admit it had overstated its fourth-quarter profits by improperly recording $125 million in sales , reducing that quarter’s revenue from “profitable” to “break-even.” Things would eventually collapse when Winstar couldn’t pay its debts, filing for Chapter 11 bankruptcy protection on April 18 2001 after failing to pay $75 million in interest payments to Lucent, which had cut access to the remaining $400 million of its $1 billion loan to Winstar as a result. Winstar would file a $10 billion lawsuit in bankruptcy court in Delaware the very same day, claiming that Lucent breached its contract and forced Winstar into bankruptcy by, well, not offering to give it more money that it couldn’t pay off. Elsewhere, things had begun to unravel for Lucent. A January 2001 story from the New York Times told a strange story of Lucent, a company that had made over $33 billion in revenue in its previous fiscal year, asking to defer the final tranche of payment — $20 million — for an acquisition due to “accounting and financial reporting considerations.” Why? Because Lucent needed to keep that money on the books to boost its earnings, as its stock was in the toilet, and was about to announce it was laying off 10,000 people and a quarterly loss of $1.02 billion . Over the course of the next few years, Lucent would sell off various entities , and by the end of September 2005 it would have 30,500 staff and have a stock price of $2.99 — down from a high of $75 a share at the end of 1999 and 157,000 employees. According to VC Tomasz Tunguz, Lucent had $8.1 billion of vendor financing deals at its height . Lucent was still a real company selling real things, but had massively overextended itself in an attempt to meet demand that didn’t really exist, and when Lucent realized that, it decided to create demand itself to please the markets. To quote MIT Tech Review (and author Lisa Endlich), it believed that “setting and meeting [the expectations of Wall Street] “subsumed all other goals,” and that “Lucent had little choice but to ride the wave.” To be clear, NVIDIA is quite different from Lucent. It has plenty of money, and the circular deals it does with CoreWeave and Lambda don’t involve the same levels of risk. NVIDIA is not (to my knowledge) backstopping CoreWeave’s business or providing it with loans , though NVIDIA has agreed to buy $6.3 billion of compute as the “buyer of last resort” of any unsold capacity . NVIDIA can actually afford this, and it isn’t illegal , though it is obviously propping up a company with flagging demand. NVIDIA also doesn’t appear to be taking on masses of debt to fund its empire, with over $56 billion in cash on hand and a mere $8.4 billion in long term debt . Okay, phew. We got through this man. NVIDIA is nothing like Lucent either . Okay, maybe it’s got some similarities — but it’s different! No worries at all. I know I’m relaxed. You still seem nervous, NVIDIA. I promise you, if anyone asks me if you’re like Lucent I’ll tell them you’re not. I’ll be sure to tell them you’re nothing like that. Are you okay, dude? When did you last sleep? Inventory growth indicates waning demand Claim: Growing inventory in Q3 (+32% QoQ) suggests that demand is weak and chips are accumulating unsold, or customers are accepting delivery without payment capability, causing inventory to convert to receivables rather than cash. Woah, woah, woah, slow down. Who has been saying this? Oh, everybody ? Did Michael Burry scare you? Did you watch The Big Short and say “ah, fuck, Christian Bale is going to get me! I can’t believe he played drums to Pantera ! Ahh!” Anyway, now you’ve woken up everybody else in the house and they’re all wondering why you’re talking about receivables. Shouldn’t that be fine? NVIDIA is a big business, and it’s totally reasonable to believe that a company planning to sell $63 billion of GPUs in the next quarter would have ballooning receivables ( $33 billion, up from $27 billion last quarter ) and growing inventory ( $19.78 billion, up from $14.96 billion the last quarter ). It’s a big, asset-heavy business, which means NVIDIA’s clients likely get decent payment terms to raise debt or move cash around to get them paid. Everybody calm down! Like my buddy NVIDIA, who is nothing like Enron by the way, just said: Response: First, growing inventory does not necessarily indicate weak demand. In addition to finished goods, inventory includes significant raw materials and work-in-progress. Companies with sophisticated supply chains typically build inventory in advance of new product launches to avoid stockouts. NVIDIA's current supply levels are consistent with historical trends and anticipate strong future growth. Second, growing inventory does not indicate customers are accepting delivery without payment capability. NVIDIA recognizes revenue upon shipping a product and deeming collectability probable. The shipment reduces inventory, which is not related to customer payments. Our customers are subject to strict credit evaluation to ensure collectability. Payment is due shortly after product delivery; some customers prepay. NVIDIA's DSO actually decreased sequentially from 54 days to 53 days. Haha, nice dude, you’re totally right, it’s pretty common for companies, especially large ones, to deliver something before they receive the cash, it happens , I’m being sincere. Sounds like companies are paying! Great! But, you know, just, can you be a little more specific? Like about the whole “shipping things before they’re paid” thing. NVIDIA recognizes revenue upon shipping a product and deeming collectability probable- Alright, yeah, thought I heard you right the first time. What does “deeming collectability probable” mean? You could’ve just said “we get paid 95% of the time within 2 months” or whatever. Unless it’s not 95%? Or 90%? How often is it? Most companies don’t break this down by the way, but then again, most companies are not NVIDIA, the largest company on the stock market, and if I’m honest, nobody else has recently had to put out anything that said “I’m not like Enron,” and I want to be clear that NVIDIA is not like Enron. For real, Enron was a criminal enterprise. It broke the law, it committed real deal, actual fraud, and NVIDIA is nothing like Enron. In fact, before NVIDIA put out a letter saying how it was nothing like Enron I would have staunchly defended the company against the Enron allegations, because I truly do not think NVIDIA is committing fraud. That being said, it is very strange that NVIDIA wants somebody to think about how it’s nothing like Enron. This was, technically, an internal memo, and thus there is a chance its existence was built for only internal NVIDIANs worried about the value of their stock, and we know it was definitely written to try and deflect Michael Burry’s criticism, as well as that of a random Substacker who clearly had AI help him write a right-adjacent piece that made all sorts of insane and made up statements (including several about Arrow Electronics that did not happen) — and no, I won’t link it, it’s straight up misinformation. Nevertheless, I think it’s fair to ask: why does NVIDIA need you to know that it’s nothing like Enron? Did it do something like Enron? Is there a chance that I, or you, may mistakenly say “hey, is NVIDIA doing Enron?” Heeeeeeyyyy NVIDIA. How’re you feeling? Yeah, haha, you had a rough night. You were saying all this crazy stuff about Enron last night, are you doing okay? No, no, I get it, you’re nothing like Enron, you said that a lot last night. So, while you were asleep — yeah it’s been sixteen hours dude, you were pretty messed up, you brought up Lucent then puked in my sink — I did some digging and like, I get it, you are definitely not like Enron, Enron was breaking the law . NVIDIA is definitely not doing that. But…you did kind of use Special Purpose Vehicles recently? I’m sorry, I know, you’re not like Enron! You’re investing $2 billion in Elon Musk’s special purpose vehicle that will then use that money to raise debt to buy GPUs from NVIDIA that will then be rented to Elon Musk . This is very different to what Enron did! I am with you dude , don’t let the haters keep you down! No, I don’t think a t-shirt that says “NVIDIA is not like Enron for these specific reasons” helps. Wait, wait, okay, look. One thing. You had this theoretical deal lined up with Sam Altman and OpenAI to invest $100 billion — and yes, you said in your latest earnings that "it was actually a Letter of Intent with the opportunity to invest," which doesn’t mean anything, got it — and the plan was that you would “ lease the GPUs to OpenAI .” Now how would you go about doing that NVIDIA? You’d probably need to do exactly the same deal as you just did with xAI. Right? Because you can’t very well rent these GPUs directly to Elon Musk , you need to sell them to somebody so that you can book the revenue, you were telling me that’s how you make money. I dunno, it’s either that or vendor financing. Oh, you mentioned that already- -unlike Lucent, NVIDIA does not rely on vendor financing arrangements to grow revenue. In typical vendor financing arrangements, customers pay for products over years. NVIDIA's DSO was 53 in Q3. NVIDIA discloses our standard payment terms, with payment generally due shortly after delivery of products. We do not disclose any vendor financing arrangements- Let me stop you right there a second, you were on about this last night before you scared my cats when you were crying about something to do with “two nanometer.” First of all, why are you bringing up typical vendor financing agreements? Do you have atypical ones? Also I’m jazzed to hear you “disclose your standard payment terms,” but uh, standard payment terms for what exactly? Where can I find those? For every contract? Also, you are straight up saying you don’t disclose any vendor financing arrangements , that’s not the same as “not having any vendor financing arrangements.” I “do not disclose” when I go to the bathroom but I absolutely do use the toilet. Let’s not pretend like you don’t have a history in helping get your buddies funding. You have deals with both Lambda and CoreWeave to guarantee that they will have compute revenue, which they in turn use to raise debt, which is used to buy more of your GPUs. You have learned how to feed debt into yourself quite well, I’m genuinely impressed . This is great stuff, I’m having the time of my life with how not like Enron you are, and I’m serious that I 100% do not believe you are like Enron. But…what exactly are you doing man? What’re you going to do about what Wall Street wants? Enron was a criminal enterprise! NVIDIA is not. More than likely NVIDIA is doing relatively boring vendor financing stuff and getting people to pay them on 50-60 day time scales — probably net 60, and, like it said, it gets paid upfront sometimes. NVIDIA truly isn’t like Enron — after all, Meta is the one getting into ENERGY TRADING — to the point that I think it’s time to explain to you what exactly happened with Enron. Or, at least as much as is possible within the confines of a newsletter that isn’t exclusively about Enron… The collapse of Enron wasn’t just — in retrospect — a large business that ultimately failed. If that was all it was, Enron wouldn’t command the same space in our heads as other failures from that era, like WorldCom (which I mentioned earlier) and Nortel (which I’ll get to later), both of whom were similarly considered giants in their fields. It’s also not just about the fact that Enron failed because of proven business and accounting malfeasance. WorldCom entered bankruptcy due to similar circumstances (though, rather than being liquidated, it was acquired as part of Verizon’s acquisition of MCI , the name of a company that had previously merged with WorldCom that WorldCom renamed itself to after bankruptcy ), and unlike Enron, isn’t the subject of flashy Academy-nominated films , or even a Broadway production . It’s not the size of Enron that makes its downfall so intriguing. Nor, for that matter, is it the fact that Enron did a lot of legally and ethically dubious stuff to bring about its downfall. No, what makes Enron special is the sheer gravity of its malfeasance, the rotten culture at the heart of the company that encouraged said malfeasance, and the creative ways Enron’s leaders crafted an image of success around what was, at its heart, a dog of a company. Enron was born in 1985 on the foundations of two older, much less interesting businesses. The first, Houston Natural Gas (HNG), started life as a utility provider, pumping natural gas from the oilfields of Texas to customers throughout the region, before later exiting the industry to focus on other opportunities. The other, InterNorth, was based in Omaha, Nebraska and was in the same business — pipelines. In the mid-1980s, HNG was the subject of a hostile take-over from Coastal Corporation (which, until 2001, operated a chain of refineries and gas stations throughout much of the US mainland). Unable to fend it off by itself, HNG merged with InterNorth, with the combined corporation renamed Enron . The CEO of this new entity was Ken Lay, an economist by trade who spent most of his career in the energy sector who also enjoyed deep political connections with the Bush family . He co-chaired George H. W. Bush’s failed 1992 re-election campaign , and allowed Enron’s corporate jet to ferry Bush Sr. and Barbara Bush back and forth to Washington. Center for Public Integrity Director Charles Lewis said that “ there was no company in America closer to George W. Bush than Enron. ” George W. Bush (the second one) even had a nickname for Lay. Kenny Boy . Anyway, in 1987, Enron hired McKinsey — the world’s most evil management consultancy firm — to help the company create a futures market for natural gas. What that means isn’t particularly important to the story, but essentially, a futures contract is where a company agrees to buy or sell an asset in the future at a fixed price. It’s a way of hedging against risk, whether that be from something like price or currency fluctuations, or from default. If you’re buying oil in dollars, for example, buying a futures contract for oil to be delivered in six months time at a predetermined price means that if your currency weakens against the dollar, your costs won’t spiral. That bit isn’t terribly important. What does matter is while working with McKinsey, Lay met someone called Jeff Skilling — a young engineer-turned-consultant who impressed the company’s CEO deeply, so much so that Lay decided to poach him from McKinsey in 1990 and give him the role of chairman and CEO of Enron Finance Group. Anyway, Skilling continued to impress Lay, who gave him greater and greater responsibility, eventually crowning him Chief Operating Officer (COO) of Enron. With Skilling in a key leadership position, he was able to shape the organization’s culture. He appreciated those who took risks — even if those risks, when viewed with impartial eyes, were deemed reckless, or even criminal. He introduced the practice of stack-ranking (also known as “rank and yank”) to Enron, which had previously been pioneered by Jack Welch at GE (see The Shareholder Supremacy from last year ). Here, employees were graded on a scale, and those at the bottom of the scale were terminated. Managers had to place at least 10% (other reports say closer to 15%) of employees in the lowest bracket, which created an almost Darwinian drive to survive. Staffers worked brutal hours. They cut corners. They did some really, really dodgy shit. None of this bothered Skilling in the slightest. How dodgy, you ask? Well, in 2000 and 2001, California suffered a series of electricity blackouts. This shouldn’t have happened, because California’s total energy demand (at the time) was 28GW and its production capacity was 45GW. California also shares a transmission grid with other states (and, for what it’s worth, the Canadian provinces of Alberta and British Colombia, as well as part of Baja California in Mexico), meaning that in the event of a shortage, it could simply draw capacity from elsewhere. So, how did it happen? Well, remember, Enron traded electricity like a commodity, and as a result, it was incentivized to get the highest possible price for that commodity . So, it took power plants off line during peak hours, and exported power to other states when there was real domestic demand. How does a company like Enron shut down a power station? Well, it just asked . In one taped phone conversation released after the company’s collapse , an Enron employee called Bill called an official at a Las Vegas power plant (California shares the same grid with Nevada) and asked him to “ get a little creative, and come up with a reason to go down. Anything you want to do over there? Any cleaning, anything like that? " This power crisis had dramatic consequences — for the people of California, who faced outages and price hikes; for Governor Gray Davis, who was recalled by voters and later replaced by Arnold Schwarzenegger; for PG&E, which entered Chapter 11 bankruptcy that year ; and for Southern California Edison, which was pushed to the brink of bankruptcy as a result. This kind of stuff could only happen in an organization whose culture actively rewarded bad behavior . In fact, Skilling was seemingly determined to elevate the dodgiest of characters to the highest positions within the company, and few were more-ethically-dubious than Andy Fastow, who Skilling mentored like a protegé, and who would later become Enron’s Chief Financial Officer. Even before vaulting to the top of Enron’s nasty little empire, Fastow was able to shape its accounting practices, with the company adopting mark-to-market accounting practices in 1991 . Mark-to-market sounds complicated, but it’s really simple. When listing assets on a balance sheet, you don’t use the acquisition cost, but rather the fair-market value of that asset. So, if I buy a baseball card for a dollar, and I see that it’s currently selling for $10 on eBay, I’d say that said asset is worth $10, not the dollar I paid for it, even though I haven’t actually sold it yet. This sounds simple — reasonable, even — but the problem is that the way you determine the value of that asset matters, and mark-to-market accounting allows companies and individuals to exercise some…creativity. Sure, for publicly-traded companies (where the price of a share is verifiable, open knowledge), it’s not too bad, but for assets with limited liquidity, limited buyers, or where the price has to be engineered somehow, you have a lot of latitude for fraud. Let’s go back to the baseball card example. How do you know it’s actually worth $10, and not $1? What if the “fair value” isn’t something you can check on eBay, but what somebody told me in-person it’s worth? What’s to stop me from lying and saying that the card is actually worth $100, or $1000? Well, other than the fact I’d be committing fraud. What if I have ten $1 baseball cards, and I give my friend $10 and tell him to buy one of the cards using the $10 bill I just handed him, allowing me to say that I’ve realized a $9 profit on one of my $1 cards, and my other cards are worth $90 and not $9? And then, what if I use the phony valuation of my remaining cards to get a $50 loan, using the cards as collateral, even though the collateral isn’t even one-fifth of the value of the loan? You get the idea. While a lot of the things people can do to alter the mark-to-market value of an asset are illegal (and would be covered under generic fraud laws), it doesn’t change the fact that mark-to-market accounting allows for some shenanigans to take place. Another trait of mark-to-market accounting, as employed by Enron, is that it would count all the long-term potential revenue from a deal as quarterly revenue — even if that revenue would be delivered over the course of a decades-long contract, or if the contract would be terminated before its intended expiration date. It would also realize potential revenue as actual revenue, even before money changed hands, and when the conclusion of the deal wasn’t a certainty. For example, in 1999, Enron sold a stake in four electricity-generating barges in Nigeria (essentially floating power stations) to Merrill Lynch , which allowed the company to register $12m in profit. That sale ultimately didn’t happen, though that didn’t stop Enron from selling pieces to Merrill Lynch, which — I’m not kidding — Merrill Lynch quickly sold back to a Special Purpose Vehicle called “LJM2” controlled by Andrew Fastow. You’re gonna hear that name again. Although the Merrill Lynch bankers who participated in the deal were eventually convicted of conspiracy and fraud charges (long after the collapse of Enron), their convictions were later quashed on appeal. But still, for a moment, it gave a jolt to Enron’s quarterly earnings. Anyway, Enron was incredibly creative when it came to how it valued its assets. Take, for example, fiber optic cables. As the Dot Com bubble swelled, Enron saw an opportunity, and wanted to be able to trade and control the supply of bandwidth, just like it does with other more conventional commodities (like oil and gas) . It built, bought, and leased fiber-optic cables throughout the country, and then, using exaggerated estimates of their value and potential long-term revenue, released glowing financial reports that made the company look a lot healthier and more successful than it actually was. Enron also loved to create special-purpose entities that existed either to generate revenue that didn’t exist, or to hold toxic assets that would otherwise need to be disclosed (with Enron then using its holdings in said entities to boost its balance sheet), or to disguise its debt. One, Whitewing, was created and capitalized by Enron (and an outside investor), and pretty much exclusively bought assets from Enron — which allowed the company to recognize sales and profits on its balance sheets, even if they were fundamentally contrived. Another set of entities — known as LJM, named after the first initial of Andy Fastow’s wife and two children , and which I mentioned earlier — did the same thing, allowing the company to hide risky or failing investments, to limit its perceived debt, and to generate artificial profits and revenues. LJM2 was, creatively, the second version of the idea. Even though the assets that LJM held were, ultimately, dogshit, the distance that LJM provided, combined with Enron’s use of mark-to-market accounting, allowed the company to turn a multi-billion collective failure into a resounding and (on paper) profitable triumph. So, how did this happen, and how did it go on for so long? Well, first, Enron was, at its peak, worth $70bn. Its failure would be a failure for its investors and shareholders, and nobody — besides the press, that is — wanted to ask tough questions. It had auditors, but they were paid handsomely, turning a blind eye to the criminal malfeasance at the heart of the company. Auditor Arthur Andersen surrendered its license in 2002, bringing an end to the company — and resulting in 85,000 employees losing their jobs. Well, it’s not so much as it only turned a blind eye, so much as it turned on a big paper shredder , shredding tons — and I’m using that as a measure of weight, and not figuratively — of documents as Enron started to implode , a crime for which it was later convicted of obstruction of justice. I’ve talked about Enron’s culture, but I’d be remiss if I didn’t mention that Enron’s highest-performers and its leadership received hefty bonuses in company equity, motivating them to keep the charade going. Enron’s pension scheme, I add, was basically entirely Enron stock, and employees were regularly encouraged to buy more, with Kenneth Lay telling employees weeks before the company’s collapse that “the company is fundamentally sound” and to “hang on to their stock.” Additionally, per the terms of the Enron pension plan, employees were prevented from shifting their holdings into other pension funds, or other investments, until they turned 50 . When the company collapsed, those people lost everything, even those who didn’t know anything about Enron’s criminality. George Maddox, a retired former Enron employee, had his entire retirement tied up in 14,000 Enron shares (worth at the time more than $1.3 million), was “forced to spend his golden years making ends meet by mowing pastures and living in a run-down East Texas farmhouse.” The US Government brought criminal charges against Enron’s top leadership. Ken Lay was convicted of four counts of fraud and making false statements , but died on a skiing vacation to Aspen before sentencing . May he burn in Hell. Skilling was convicted on 24 counts of fraud and conspiracy and sentenced to 24 years in jail. This was reduced in 2013 on appeal to 14 years, and he was released to a halfway house in 2018 , and then freed in 2019. He’s since tried to re-enter the energy sector — with one venture combining energy trading and, I kid you not, blockchain technology — although nothing really came out of it. Andy Fastow pled guilty to two counts — one of manipulation of financial statements, and one of self-dealing . and received ten years in prison. This was later reduced to six years, including two years of probation , in part because he cooperated with the investigations against other Enron executives. He is now a public speaker and a tech investor in an AI company, KeenCorp . His wife, Lea, who also worked at Enron, received twelve months for conspiracy to commit wire fraud and money laundering and for submitting false tax returns. She was released from custody in July, 2005 . Enron’s implosion was entirely self-inflicted and horrifyingly, painfully criminal, yet, it had plenty of collateral damage — to the US economy, to those companies that had lent it money, to its employees who lost their jobs and their life savings and their retirements, and to those employees at companies most entangled with Enron, like those at auditing firm Arthur Andersen. This isn’t unique among corporate failures. WorldCom had some dodgy accounting practices. Nortel too. Both companies failed, both companies wrecked the lives of their employees, and the failure of these companies had systemic economic consequences (especially in Canada, where Nortel, at its peak, accounted for one-third of the market cap of all companies on the Toronto Stock Exchange). The reason why Enron remains captured in our imagination — and why NVIDIA is so vociferously opposed to being compared with Enron — is the extent to which Enron manipulated reality to appear stronger and more successful than it was, and how long it was able to get away with it. While we may have forgotten the memory of Enron — it happened over two decades ago, after all — we haven’t forgotten the instincts that it gave us. It’s why our noses twitch when we see special-purpose vehicles being used to buy GPUs, and why we gag when we see mark-to-market accounting. It’s entirely possible that everything NVIDIA is doing is above board. Great! But that doesn’t do anything for the deep pit of dread in my stomach. A few weeks ago, I published the Hater’s Guide to NVIDIA, and included within it a guide to what this company does . If you’re looking at this through the cold, unthinking lenses of late-stage capitalism. This all sounds really good! I’ve basically described a company that has an essential monopoly in the one thing required for a high-growth (if we’re talking exclusively about capex spending) industry to exist. Moreover, that monopoly is all-but assured, thanks to NVIDIA’s CUDA moat, its first-mover advantage, and the actual capabilities of the products themselves — thereby allowing the company to charge a pretty penny to customers. And those customers? If we temporarily forget about the likes of Nebius and CoreWeave (oh, how I wish I could forget about CoreWeave permanently), we’re talking about the biggest companies on the planet. Ones that, surely, will have no problems paying their bills. Back in February 2023, I wrote about The Rot Economy , and how everything in tech had become oriented around growth — even if it meant making products harder to use as a means of increasing user engagement or funnelling them toward more-profitable parts of an app. Back in June 2024, I wrote about the Rot-Com Bubble , and my greater theory that the tech industry has run out of hypergrowth ideas: In simple terms, big tech — Amazon, Google, Microsoft and Meta, but also a number of other companies — no longer has the “next big thing,” and jumped on AI out of an abundance of desperation. Hell, look at Oracle. This company started off by selling databases and ERP systems to big companies, and then trapping said companies by making it really, really difficult to migrate to cheaper (and better) solutions, and then bleeding said companies with onerous licensing terms (including some where you pay by the number of CPU cores that use the application). It doesn’t do anything new, or exciting, or impressive, and even when presented with the opportunity to do things that are useful or innovative (like when it bought Sun Microsystems), it turns away. I imagine that, deep down, it recognizes that its current model just isn’t viable in the long-term, and so, it needs something else. When you haven’t thought about innovation… well… ever, it’s hard to start. Generative AI, on the face of it, probably seemed like a godsend to Larry Ellison. We also live in an era where nobody knows what big tech CEOs do other than make nearly $100 million a year , meaning that somebody like Satya Nadella can get called a “ thoughtful leader with striking humility ” for pushing Copilot AI in every single part of your Microsoft experience, even Notepad, a place that no human being would want it , and accelerating capital expenditures from $28 billion across the entirey of FY 2023 to $34.9 billion in its latest quarter . In simpler terms, spending money makes a CEO look busy. And at a time when there were no other potential growth avenues, AI was a convenient way to make everybody look busy. Every department can “have an AI strategy,” and every useless manager and executive can yell, as ServiceNow CEO did back in 2022 , “ let me make it clear to everybody here, everything you do: AI, AI, AI, AI, AI. ” I should also add that ChatGPT was the first real, meaningful hit that the American tech industry had produced in a long, long time — the last being, if I’m honest, Uber, and that’s if we allow “successful yet not particularly good businesses” into the pile. If we insist on things like “profitability” and “sustainability,” US tech hasn’t done so great. Snowflake runs at a loss , Snap runs at a loss , and while Uber has turned things around somewhat , it’s hardly created the next cloud computing or smartphone. Putting aside finances, the last major “hit” was probably Venmo or Zelle, and maybe, if I’m feeling generous, smart speakers like Amazon Echo and Apple Homepod. Much like Uber, none of these were “the next big thing,” which would be fine except big tech needs more growth forever right now, pig! This is why Google, Amazon and Meta all do 20 different things — although rarely for any length of time, with these “things” often having a shelf life shorter than a can of peaches — because The Rot Economy’s growth-at-all-costs mindset exists only to please the markets, and the markets demanded growth. ChatGPT was different. Not only did it do something new, it did so in a way that was relatively easy to get people to try and “see the potential” of. It was also really easy to convince people it would become something bigger and better , because that’s what tech does. To quote Bender and Hanna, AI is a “marketing term ” — a squishy way of evoking futuristic visions of autonomous computers that can do anything and everything from us, and because both consumers and analysts have been primed to believe and trust the tech industry, everybody believed that whatever ChatGPT was would be the Next Big Thing. And said “Next Big Thing” is powered by Large Language Models, which require GPUs sold by one company — NVIDIA. AI became a very useful thing to do. If a company wanted to seem futuristic and attract investors, it could now “integrate AI.” If a hyperscaler wanted to seem enterprising and like it was “building for the future,” it could buy a bunch of GPUs, or invest in its own silicon, or, as Google, Microsoft, Amazon and Meta have done, shove AI in every imaginable crevice of the app. Investors could invest in AI companies, retail investors (IE: regular people) could invest in AI stocks, tech reporters could write about something new in AI, LinkedIn perverts could write long screeds about AI, the markets could become obsessed with AI… …and yeah, you can kind of see how things got out of control. Everybody now had something to do . An excuse to do AI, regardless of whether it made sense, because everybody else was doing it. ChatGPT quickly became one of the most popular websites on the internet — all while OpenAI burned billions of dollars — and because the media effectively published every single thought that Sam Altman had (such as that GPT-4 would “automate away some jobs and create others ” and that he was a “ little bit scared of it ”), AI, as an idea, technology, symbolic stock trope, marketing tool and myth became so powerful that it could do anything, replace anyone, and be worth anything, even the future of your company. Amongst the hype, there was an assumption related to scaling laws ( summarized well by Charlie Meyer ): In simple terms, the paper suggested that shoving more training data and using more compute power would exponentially increase the ability of a model to do stuff. And to make a model that did more stuff, you needed more GPUs and more data centers. Did it matter that there was compelling evidence in 2022 ( Gary Marcus was right! ) that there were limits to scaling laws, and that we would hit the point of diminishing returns? Amidst all this, NVIDIA has sold over $200 billion of GPUs since the beginning of 2023 , becoming the largest company on the stock market and trading at over $170 as of writing this sentence only a few years after being worth $19.52 a share . You see, Meta, Google, Microsoft and Amazon all wanted to be “part of the future,” so they sunk a lot of their money into NVIDIA, making up 42% of its revenue in its fiscal year 2025. Though there are some arguments about exactly how much of big tech’s billowing capital expenditures are spent on GPUs, some estimate somewhere between 41% to more than 50% of a data center’s capex is spent on them. If you’re wondering what the payoff is, well, you’re in good company. I estimate that there’s only around $61 billion in total generative AI revenue , and that includes every hyperscaler and neocloud. Large Language Models are limited, AI agents are a pipedream and simply do not work , AI-powered products are unreliable and coding LLMs make developers slower , and the cost of inference — the way in which a model produces its output — keeps going up . So, due to the fact that so much money has now been piled into building AI infrastructure, and big tech has promised to spend hundreds of billions of dollars more in the next year , big tech has found itself in a bit of a hole. How big a hole? Well, By the end of the year, Microsoft, Amazon, Google and Meta will have spent over $400bn in capital expenditures, much of it focused on building AI infrastructure, on top of $228.4 billion in capital expenditures in 2024 and around $148bn in capital expenditures in 2023, for a total of around $776bn in the space of three years, and intends to spend $400 billion or more in 2026. As a result, based on my analysis, big tech needs to make $2 trillion in brand new revenue, specifically from AI by 2030, or all of this was for nothing. I go into detail here in my premium piece , but I’m going to give you a short explanation here. Sadly you’re going to have to learn stuff. I know! I’m sorry. Introducing a term: depreciation. From my October, 31 newsletter : Nobody seems to be able to come to a consensus about how long this should be. In Microsoft’s case, depreciation for its servers is spread over six years — a convenient change it made in August 2022, a few months before the launch of ChatGPT. This means that Microsoft can spread the cost of the tens of thousands of A100 GPUs bought in 2020, or the 450,000 H100 GPUs it bought in 2024 , across six years, regardless of whether those are the years they will be either A) generating revenue or B) still functional. CoreWeave, for what it’s worth, says the same thing — but largely because it’s betting that it’ll still be able to find users for older silicon after its initial contracts with companies like OpenAI expire. The problem is, as the aforementioned linked CNBC article points out, is that this is pretty much untested ground. Whereas we know how much, say, a truck or a piece of heavy machinery can last, and how long it can deliver value to an organization, we don’t know the same thing about the kind of data center GPUs that hyperscapers are spending tens of billions of dollars on each year. Any kind of depreciation schedule is based on, at best, assumptions, and at worst, hope. The assumption that the cards won’t degrade with heavy usage. The assumption that future generations of GPUs won’t be so powerful and impressive, they’ll render the previous ones more obsolete than expected, kind of like how the first jet-powered planes of the 1950s did to those manufactured just one decade prior. The assumption that there will, in fact, be a market for older cards, and that there’ll be a way to lease them profitably. What if those assumptions are wrong? What if that hope is, ultimately, irrational? Mihir Kshirsagar of the Center for Information Technology Policy framed the problem well : This is why Michael Burry brought it up recently — because spreading out these costs allows big tech to make their net income (IE: profits) look better. In simple terms, by spreading out costs over six years rather than three, hyperscalers are able to reduce a line item that eats into their earnings, which makes their companies look better to the markets. So, why does this create an artificial time limit? In really, really simple terms: So, now that you know this, there’s a fairly obvious question to ask: why are they still buying GPUs? Also…where the fuck are they going? As I covered in the Hater’s Guide To NVIDIA : While I’m not going to copy-paste my whole (premium) piece, I was only able to find, at most, a few hundred thousand Blackwell GPUs — many of which aren’t even online! — including OpenAI’s Stargate Abilene (allegedly 400,000, though only two buildings are handed over); a theoretical 131,000 GPU cluster owned by Oracle announced in March 2025 ; 5000 Blackwell GPUs at the University of Texas, Austin ; “more than 1500” in a Lambda data center in Columbus, Ohio ; The Department of Energy’s still-in-development 100,000 GPU supercluster, as well as “10,000 NVIDIA Blackwell GPUs” that are “expected to be available in 2026 in its “Equinox” cluster ; 50,000 going into the still-unbuilt Musk-run Colossus 2 supercluster ; CoreWeave’s “largest GB200 Blackwell cluster” of 2496 Blackwell GPUs ; “tens of thousands” of them deployed globally by Microsoft ( including 4600 Blackwell Ultra GPUs ); 260,000 GPUs for five AI data centers for the South Korean government …and I am still having trouble finding one million of these things that are actually allocated anywhere , let alone in a data center, let alone one with sufficient power. I do not know where these six million Blackwell GPUs have gone, but they certainly haven’t gone into data centers that are powered and turned on. In fact, power has become one of the biggest issues with building these things, in that it’s really difficult (and maybe impossible!) to get the amount of power these things need. In really simple terms: there isn’t enough power or built data centers for those six million Blackwell GPUs, in part because the data centers aren’t built, and in part because there isn’t enough power for the ones that are. Microsoft CEO Satya Nadella recently said on a podcast that his company “[didn’t] have the warm shells to plug into,” meaning buildings with sufficient power, and heavily suggested Microsoft “may actually have a bunch of chips sitting in inventory that [he] couldn’t plug in.” The news that HPE’s (Hewlett Packard Enterprise) AI server business underperformed, and by a significant margin, only raises more questions about where these chips are going . So why, pray tell, is Jensen Huang of NVIDIA saying that he has 20 million Blackwell and Vera Rubin GPUs ordered through the end of 2026 ? Where are they going to go? I truly don’t know! AI bulls will tell you about the “insatiable demand for AI” and that these massive amounts of orders are proof of something or rather, and you know what, I’ll give them that — people sure are buying a lot of NVIDIA GPUs! I just don’t know why . Nobody has made a profit from AI, and those making revenue aren’t really making much. For example, my reporting on OpenAI from a few weeks ago suggests that the company only made $4.329 billion in revenue through the end of September, extrapolated from the 20% revenue share that Microsoft receives from the company. As some people have argued with the figures, claiming they are either A) delayed or B) not inclusive of the revenue that OpenAI is paid from Microsoft as part of Bing’s AI integration and sales of OpenAI’s models via Microsoft Azure, I wanted to be clear of two things: In the same period, it spent $8.67 billion on inference (the process in which an LLM creates an output). This is the biggest company in the generative AI space, with 800 million weekly active users and the mandate of heaven in the eyes of the media. Anthropic, its largest competitor, alleges it will make $833 million in revenue in December 2025 , and based on my estimates will end up having $5 billion in revenue by end of year. Based on my reporting from October, Anthropic spent $2.66 billion on Amazon Web Services through the end of September, meaning that it (based on my own analysis of reported revenues) spent 104% of its $2.55 billion in revenue up until that point just on AWS , and likely spent just as much on Google Cloud. While everybody wants to tell the story of Anthropic’s “efficiency” and “ only burning $2.8 billion this year ,” one has to ask why a company that is allegedly “reducing costs” had to raise $13 billion in September 2025 after raising $3.5 billion in March 2025 , and after raising $4 billion in November 2024 ? Am I really meant to read stories about Anthropic hitting break even in 2028 with a straight face? Especially as other stories say Anthropic will be cash flow positive “ as soon as 2027 .” These are the two largest companies in the generative AI space, and by extension the two largest consumers of GPU compute. Both companies burn billions of dollars, and require an infinite amount of venture capital to keep alive at a time when the Saudi Public Investment Fund is struggling and the US venture capital system is set to run out of cash in the next year and a half . The two largest sources of actual revenue for selling AI compute are subsidized by venture capital and debt. What happens if these sources dry up? And, in all seriousness, who else is buying AI compute? What are they doing with it? Hyperscalers (other than Microsoft, which chose to stop reporting its AI revenue back in January, when it claimed a $13 billion, or about $1 billion a month, in revenue ) don’t disclose anything about their AI revenue, which in turn means we have no real idea about how much real, actual money is coming in to justify these GPUs. CoreWeave made $1.36 billion in revenue (and lost $110 million doing so) in its last quarter — and if that’s indicative of the kind of actual, real demand for AI compute, I think it’s time to start panicking about whether all of this was for nothing. CoreWeave has a backlog of over $50 billion in compute , but $22 billion of that is OpenAI (a company that burns billions of dollars a year and lives on venture subsidies), $14 billion of that is Meta (which has yet to work out how to make any kind of real money from generative AI, and no, its “ generative AI ads ” are not the future, sorry), and the rest is likely a mixture of Microsoft and NVIDIA, which agreed to buy $6.3 billion of any unused compute from CoreWeave through 2032 . Sorry, I also forgot Google, which is renting capacity from CoreWeave to rent to OpenAI . Also, I also forgot to mention that CoreWeave’s backlog problem stems from data center construction delays . That and CoreWeave has $14 billion in debt mostly from buying GPUs, which it was able to raise by using GPUs as collateral and that it had contracts from customers willing to pay it, such as NVIDIA, which is also selling it the GPUs. So, just to be abundantly clear: CoreWeave has bought all those GPUs to rent to OpenAI, Microsoft (for OpenAI), Meta, Google (OpenAI), and NVIDIA, which is the company that benefits from CoreWeave’s continued ability to buy GPUs. Otherwise, where’s the fucking business, exactly? Who are the customers? Who are the people renting these GPUs, and for what purpose are they being rented? How much money is renting those GPUs? You can sit and waffle on about the supposedly glorious “AI revolution” all you want, but where’s the money, exactly? And why, exactly, are we buying more GPUs? What are they doing? To whom are they being rented? For what purpose? And why isn’t it creating the kind of revenue that is actually worth sharing? Is it because the revenue sucks? Is it because it’s unprofitable to provide it? And why, at this point in history, do we not know? Hundreds of billions of dollars that have made NVIDIA the biggest company on the stock market and we still do not know why people are buying these fucking things. NVIDIA is currently making hundreds of billions in revenue selling GPUs to companies that either plug them in and start losing money or, I assume, put them in a warehouse for safe keeping. This brings me to my core anxiety: why, exactly, are companies pre-ordering GPUs? What benefit is there in doing so? Blackwell does not appear to be “more efficient” in a way that actually makes anybody a profit, and we’re potentially years from seeing these GPUs in operation in data centers at the scale they’re being shipped — so why would anybody be buying more? I doubt these are new customers — they’re likely hyperscalers, neoclouds like CoreWeave and resellers like Dell and SuperMicro — because the only companies that can actually afford to buy them are those with massive amounts of cash or debt, to the point that even Google , Amazon , Meta and Oracle are taking on massive amounts of new debt, all without a plan to make a profit. NVIDIA’s largest customers are increasingly unable to afford its GPUs, which appear to be increasing in price with every subsequent generation. NVIDIA’s GPUs are so expensive that the only way you can buy them is by already having billions of dollars or being able to raise billions of dollars, which means, in a very real sense, that NVIDIA is dependent not on its customers , but on its customers’ credit ratings and financial backers. To make matters worse, the key reason that one would buy a GPU is to either run services using it or rent it to somebody else, and the two largest parties spending money on these services are OpenAI and Anthropic, both of whom lose billions of dollars, and are thus dependent on venture capital and debt (remember, OpenAI has a $4 billion line of credit , and Anthropic a $2.5 billion one too ). In simple terms, NVIDIA’s customers rely on debt to buy its GPUs, and NVIDIA’s customers’ customers rely on debt to pay to rent them. Yet it gets worse from there. Who, after all, are the biggest customers renting AI compute? That’s right, AI startups, all of which are deeply unprofitable. Cursor — Anthropic’s largest customer and now its biggest competitor in the AI coding sphere — raised $2.3 billion in November after raising $900 million in June . Perplexity, one of the most “popular” AI companies, raised $200 million in September after raising $100 million in July after seeming to fail to raise $500 million in May (I’ve not seen any proof this round closed) after raising $500 million in December 2024 . Cognition raised $400 million in September after raising $300 million in March . Cohere raised $100 million in September a month after it raised $500 million . Venture capital is feeding money to either OpenAI or Anthropic to use their models, or in some cases hyperscalers or neoclouds like CoreWeave or Lambda to rent NVIDIA GPUs. OpenAI and Anthropic then raise venture capital or debt to pay hyperscalers or neoclouds to rent NVIDIA GPUs. Hyperscalers and neoclouds then use either debt or existent cashflow (in the case of hyperscalers, though not for long!) to buy more NVIDIA GPUs. Only one company actually makes a profit here: NVIDIA. At some point, a link in this debt-backed chain breaks, because very little cashflow exists to prop it up. At some point, venture capitalists will be forced to stop funnelling money into unprofitable, unsustainable AI companies, which will make those companies unable to funnel money into the pockets of those buying GPUs, which will make it harder for those companies buying GPUs to justify (or raise debt for) buying more GPUs. And if I’m honest, none of NVIDIA’s success really makes any sense. Who is buying so many GPUs? Where are they going? Why are inventories increasing ? Is it really just pre-buying parts for future orders? Why are accounts receivable climbing , and how much product is NVIDIA shipping before it gets paid? While these are both explainable as “this is a big company and that’s how big companies do business” (which is true!), why do receivables not seem to be coming down? And how long, realistically, can the largest company on the stock market continue to grow revenues selling assets that only seem to lose its customers money? I worry about NVIDIA, not because I believe there’s a massive scandal, but because so much rides on its success, and its success rides on the back of dwindling amounts of venture capital and debt, because nobody is actually making money to pay for these GPUs. In fact, I’m not even saying it goes tits up. Hell, it might even have another good quarter or two. It really comes down to how long people are willing to be stupid and how long Jensen Huang is able to call hyperscalers at three in the morning and say “buy one billion dollars of GPUs, pig.” No, really! I think much of the US stock market’s growth is held up by how long everybody is willing to be gaslit by Jensen Huang into believing that they need more GPUs. At this point it’s barely about AI anymore, as AI revenue — real, actual cash made from selling services run on GPUs — doesn’t even cover its own costs, let alone create the cash flow necessary to buy $70,000 GPUs thousands at a time. It’s not like any actual innovation or progress is driving this bullshit! In any case, the markets crave a healthy NVIDIA, as so many hundreds of billions of dollars of NVIDIA stock sit in the hands of retail investors and people’s 401ks, and its endless growth has helped paper over the pallid growth of the US stock market and, by extension, the decay of the tech industry’s ability to innovate. Once this pops — and it will pop, because there is simply not enough money to do this forever — there must be a referendum on those that chose to ignore the naked instability of this era, and the endless lies that inflated the AI bubble. Until then, everybody is betting billions on the idea that Wile E. Coyote won’t look down. Let’s start with a horrible fact: it takes about 2.5 years of construction time and $50 billion per gigawatt of data center capacity . One way or another, these GPUs are depreciating in value, either through death (or reduced efficacy through wear and tear) or becoming obsolete, which is very likely as NVIDIA has committed to releasing a new GPU every year . At some point, Wall Street is going to need to see some sort of return on this investment, and right now that return is “negative dollars.” I break it down in my premium piece, but I estimate that big tech needs to make $2 for every $1 of capex . This revenue must also be brand spanking new, as this capex is only for AI. Meta, Amazon, Google and Microsoft are already years and hundreds of billions of dollars in , and are yet to see a dollar of profit , creating a $1.21 trillion hole just to justify the expenses (so around $605 billion in capex all told, at the time I calculated it). You might argue that there’s a scenario where, say, an A100 GPU is “useful” past the 3 or 6 year shelf life. Even if that were the case, the average rental price of an A100 GPU is 99 cents an hour . This is a four or five-year-old GPU, and customers are paying for it like they would a five-year-old piece of hardware. The same fate awaits H100 GPUs too. Every year, NVIDIA releases a new GPU, lowering the value of all the other GPUs in the process, making it harder to fill in the holes created by all the other GPUs. This whole time, nobody appears to have found a way to make a profit, meaning that the hole created by these GPUs remains unfilled, all while big tech firms buy more GPUs, creating more holes to fill. Big tech keeps buying more GPUs despite the old GPUs failing to pay for themselves. To fix this problem, big tech is buying more GPUs. Newer generation GPUs — like NVIDIA’s Blackwell and Vera Rubin — require entirely new data center architecture, meaning that one has to either build a brand new data center or retrofit an old one. Big tech is spending billions of dollars to make sure it’s able to turn on these new GPUs, at which point you may think that they’ll make a profit. Even when they’re turned on, these things don’t make money. The Information reports that Oracle’s Blackwell GPUs have a negative 100% gross margin . How exactly are these bloody things meant to make more money than they cost in the next six years, let alone three? They don’t make a profit now and have no path to doing so in the future! I feel like I’m going INSANE! This is accrual accounting, meaning that these numbers are revenue booked in the quarter I reported them. Any comments about quarter-long delays in payments are incorrect. Microsoft’s revenue share payments to OpenAI are pathetic — totalling, based on documents reviewed by this publication, $69.1 million in CY (calendar year) Q3 2025.
Listen to this post : Warner Bros. started with distribution. Just after the turn of the century, Harry, Albert, Sam, and Jack Warner bought a second-hand projector and started showing short films in Ohio and Pennsylvania mining towns; in 1907 they bought their first permanent theater in New Castle, Pennsylvania. Around the same time the brothers also began distributing films to other theaters, and in 1908 began producing their own movies in California. In 1923 the brothers formally incorporated as Warner Bros. Pictures, Inc., becoming one of the five major Hollywood Studios. What the brothers realized early on was that distribution just wasn’t a very good business: you had to maintain the theater and find films to show, and your profit was capped by your capacity, which you had to work diligently to fill out; after all, every empty seat in a showing was potential revenue that disappeared forever. What was far more lucrative was making the films shown in those theaters: you could film a movie once and make money on it again and again. In this Hollywood was the tech industry before there was a tech industry, which is to say the studios were the industry that focused its investment on large-up-front costs that could be leveraged repeatedly to make money. Granted, Warner Bros., along with the rest of Hollywood, did come to own large theater chains as well as part of fully integrated companies, but when the Supreme Court, with 1948’s Paramount decrees, forced them to split, it was the theaters that got spun out: making content was simply a much better business than distributing it. That business only got better over time. First, television provided an expansive new licensing opportunity for films and eventually TV shows; not only were there more televisions than theaters, but they were accessible at all hours in the home. Then, home video added a new window: movies could not only make money in theaters and on TV, but there were entirely new opportunities to rent and sell recordings. The real bonanza, however, was the cable bundle: now, instead of needing to earn discrete revenue, the majority of Hollywood revenue became a de facto annuity, as 90% of households paid an ever increasing amount of money every month to have access to a universe of content they mostly didn’t watch. Netflix, which was founded in 1997, also started with distribution, specifically of DVDs-by-mail; the streaming service that the company is known for today launched in 2007, 100 years after the Warner brothers bought their theater. The differences were profound: because Netflix was on the Internet, it was available literally everywhere; there were no seats to clean or projectors to maintain, and every incremental customer was profit. More importantly, the number of potential customers was, at least in theory, the entire population of the world. That, in a nutshell, is why the Internet is different : you can, from day one, reach anyone, with zero marginal cost. Netflix did, over time, like Warner Bros. before them, backwards integrate into producing their own content. Unlike Warner Bros., however, that content production was and has always only ever been in service of Netflix’s distribution. What Netflix has understood — and what Hollywood, Warner Bros. included, was far too slow to realize — is that because of the Internet distribution is even more scalable than content. The specifics of this are not obvious; after all, content is scarce and exclusive, while everyone can access the Internet. However, it’s precisely because everyone can access the Internet that there is an abundance of content, far too much for anyone to consume; this gives power to Aggregators who sort that content on consumers’ behalf, delivering a satisfying user experience. Consumers flock to the Aggregator, which makes the Aggregator attractive to suppliers, giving them more content, which attracts more consumers, all in a virtuous cycle. Over time the largest Aggregators gain overwhelming advantages in customer acquisition costs and simply don’t churn users; that is the ultimate source of their economic power. This is the lesson Hollywood studios have painfully learned over the last decade. As Netflix grew — and importantly, had a far more desirable stock multiple despite making inferior content — Hollywood studios wanted in on the game, and the multiple, and they were confident they would win because they had the content. Content is king, right? Well, it was, in a world of distribution limited by physical constraints; on the Internet, customer acquisition and churn mitigation in a world of infinite alternatives matters more, and that’s the advantage Netflix had, and that advantage has only grown. On Friday, Netflix announced it was buying Warner Bros.; from the Wall Street Journal : Netflix has agreed to buy Warner Bros. for $72 billion after the entertainment company splits its studios and HBO Max streaming business from its cable networks, a deal that would reshape the entertainment and media industry. The cash-and-stock transaction was announced Friday after the two sides entered into exclusive negotiations for the media company known for Superman and the Harry Potter movies, as well as hit TV shows such as “Friends.” The offer is valued at $27.75 per Warner Discovery share and has an enterprise value of roughly $82.7 billion. Rival Paramount, which sought to buy the entire company, including Warner’s cable networks, bid $30 per share all-cash for Warner Discovery, according to people familiar with the matter. Paramount is weighing its next move, which could involve pivoting to other potential acquisitions, people familiar with its plans said. Paramount’s bid, it should be noted, was for the entire Warner Bros. Discovery business, including the TV and cable networks that will be split off next year; Netflix is only buying the Warner Bros. part. The Puck reported that the stub Netflix is leaving behind is being valued at $5/share, which would mean that Netflix outbid Paramount. And, it should be noted, that Paramount money wouldn’t be from the actual business, which is valued at a mere $14 billion; new owner David Ellison is the son of Oracle founder Larry Ellison, who is worth $275 billion. Netflix, meanwhile, is worth $425 billion and generated $9 billion in cash flow over the last year. Absent family money this wouldn’t be anywhere close to a fair fight. That’s exactly what you would expect given Netflix’s position — and the most optimistic scenario I painted back in 2016 : Much of this analysis about the impact of subscriber numbers, growth rates, and churn apply to any SaaS company, but for Netflix the stakes are higher: the company has the potential to be an Aggregator , with the dominance and profits that follow from such a position. To review: Netflix has acquired users through, among other things, a superior TV viewing experience. That customer base has given the company the ability to secure suppliers, which improve the attractiveness of the company’s offerings to users, which gives Netflix even more power over suppliers. The most bullish outcome in this scenario is Netflix as not simply another cable channel with a unique delivery method, but as the only TV you need with all of the market dominance over suppliers that entails. The most obvious way that this scenario might have developed is that Netflix ends up being the only buyer for Hollywood suppliers, thanks to their ability to pay more by virtue of having the most customers; that is the nature of the company’s relationship with Sony , which had the foresight (and lack of lost TV network revenue to compensate for) to avoid the streaming wars and simply sell its content to the highest bidder. There are three specific properties I think of, however, that might be examples of what convinced Netflix it was worth simply buying one of the biggest suppliers entirely: With regards to KPop Demon Hunters , I wrote in an Update : How much of the struggle for original animation comes from the fact that no one goes to see movies on a lark anymore? Simply making it to the silver screen used to be the biggest hurdle; now that the theater is a destination — something you have to explicitly choose to do, instead of do on a Friday night by default — you need to actually sell, and that favors IP the audience is already familiar with. In fact, this is the most ironic capstone to Netflix’s rise and the misguided chase by studios seeking to replicate their success: the latter thought that content mattered most, but in truth great content — and again, KPop Demon Hunters is legitimately good — needs distribution and “free” access in the most convenient way possible to prove its worth. To put it another way, KPop Demon Hunters is succeeding on its own merits, but those merits only ever had a chance to matter because they were accessible on the largest streaming service. In short, I think that Netflix executives have become convinced that simply licensing shows is leaving money on the table: if Netflix is uniquely able to make IP more valuable, then the obvious answer is to own the IP. If the process of acquiring said IP helps force the long overdue consolidation of Hollywood studios, and takes a rival streamer off the board (and denies content to another rival), all the better. There are certainly obvious risks, and the price is high, but the argument is plausible. That phrase — “takes a rival streamer off the board” — also raises regulatory questions, and no industry gets more scrutiny than the media in this regard. That is sure to be the case for Netflix; from Bloomberg : US President Donald Trump raised potential antitrust concerns around Netflix Inc.’s planned $72 billion acquisition of Warner Bros. Discovery Inc., noting that the market share of the combined entity may pose problems. Trump’s comments, made as he arrived at the Kennedy Center for an event on Sunday, may spur concerns regulators will oppose the coupling of the world’s dominant streaming service with a Hollywood icon. The company faces a lengthy Justice Department review of a deal that would reshape the entertainment industry. “Well, that’s got to go through a process, and we’ll see what happens,” Trump said when asked about the deal, confirming he met Netflix co-Chief Executive Officer Ted Sarandos recently. “But it is a big market share. It could be a problem.” It’s important to note that the President does not have final say in the matter: President Trump directed the DOJ to oppose AT&T’s acquisition of Time Warner, but the DOJ lost in federal court , much to AT&T’s detriment. Indeed, the irony of mergers and regulatory review is that is that the success of the latter is often inversely correlated to the wisdom of the former: the AT&T deal for Time Warner never made much sense, which is directly related to why it (correctly) was approved. It would have been economically destructive for AT&T to, say, limit Time Warner content to its networks, so suing over that theoretical possibility was ultimately unsuccessful. This deal is more interesting. The complaint, if there ends up being one, will, as is so often the case, come down to market definition. If the market is defined extremely narrowly as subscription streaming services, then Netflix will have a harder time; if the market is defined as TV viewing broadly, then Netflix has a good defense: that definition includes linear TV, YouTube, etc., where Netflix’s share is both much smaller and also (correctly) includes their biggest threat (YouTube). That YouTube is Netflix’s biggest threat speaks to a broader point: because of the Internet there is no scarcity in terms of access to customers; it’s not as if there are a limited number of Internet packets, as there once were a limited number of TV channels. Everything is available to everyone, which means the only scarce resource is people’s time and attention. If this were the market definition — which is the market all of these companies actually care about — then the list of competitors expands beyond TV and YouTube to include social media and user-generated content broadly: TikTok, to take an extreme example, really is a Netflix competitor for the only scarce resource that is left. Ultimately, however, I think that everything Netflix does has to be framed in the context of the aforementioned YouTube threat. YouTube has not only long surpassed Netflix in consumer time spent generally, but also TV time specifically, and has done so with content it has acquired for free. That is very difficult to compete with in the long run: YouTube will always have more new content than anyone else. The one big advantage professionally-produced content has, however, is that it tends to be more evergreen and have higher re-watchability. That’s where we come back to the library: implicit in Netflix making library content more valuable is that library content has longevity in a way that YouTube content does not. That, by extension, may speak to why Netflix has decided to initiate the Hollywood end game now: the real threat to Hollywood isn’t (just) that the Internet made distribution free, favoring the Aggregators; it’s that technology has made it possible for anyone to create content, and the threat isn’t theoretical: it’s winning in the market. Netflix may be feared by the town, but everyone in Hollywood should fear the fact that anyone can be a creator much more. In 2019, Netflix launched Formula 1: Drive to Survive , which has been a massive success. The biggest upside recipient of that series, however, has not been Netflix, but Formula 1 owner Liberty Media. In 2018 Liberty Media offered the U.S. TV rights to ESPN for free; seven years later Apple signed a deal to broadcast Formula 1 for $150 million a year. That upside was largely generated by Netflix, who captured none of it. In 2023, NBCUniversal licensed Suits to Netflix, and the show, long since stuck in the Peacock backwater, suddenly became the hottest thing in streaming. Netflix didn’t pay much, because the deal wasn’t exclusive, but it was suddenly apparent to everyone that Netflix had a unique ability to increase the value of library content. In 2025, KPop Demon Hunters became a global phenomenon, and it’s difficult to see that happening absent the Netflix algorithm. First, it is in part a vertical merger, wherein a distributor is acquiring a supplier, which is generally approved. However, it seems likely that Netflix will, over time, make Warner Bros. content, particularly its vast libraries, exclusive to Netflix, instead of selling it to other distributors. This will be economically destructive in the short term, but it very well may be outweighed by the aforementioned increase in value that Netflix can drive to established IP, giving Netflix more pricing power over time (which will increase regulatory scrutiny). Second, it is also in part a horizontal merger, because Netflix is acquiring a rival streaming service, and presumably taking it off the market. Horizontal mergers get much more scrutiny, because the explicit outcome is to reduce competition. The frustrating point for Netflix is that the company probably doesn’t weigh this point that heavily: it’s difficult to see HBO Max providing incremental customers to Netflix, as most HBO Max customers are also Netflix customers. Indeed, Netflix may argue that they will, at least in the short to medium term, be providing consumers benefit by giving them the same content for a price that is actually lower, since you’re only paying for one service (although again, the long-term goal would be to increase pricing power).
Delegation doesn't scale your business. What does: Creating a team that is better than the current team. Including you.
I founded Moonshine back in 2022, together with Manjunath, another engineer and researcher. My entire career up until that point had been working on consumer products, so I felt very comfortable with how those are sold, and I thought to myself “How hard can B2B sales be?”. The answer, of course, is very hard! My investors knew that before I did, and pushed me to hire a senior sales person to make up for my lack of experience. It’s taken me three years and multiple failed attempts to build a working sales team, mostly because I didn’t even know enough to ask the right questions. The biggest mistake I kept making was hiring people with ten or twenty years of enterprise sales experience. This wasn’t because they were bad at their jobs, everyone who made it through our interview process had done amazing things at larger companies, but I set them up to fail at my startup. Here’s why: Startup Sales aren’t Enterprise Sales Experienced sales people are used to being given a list of qualified leads, a clear set of sales materials, and in general a “repeatable sales motion” that they can follow to close deals. There’s a whole world of Sales Development Representatives (SDRs) who handle finding and qualifying leads through cold-calling, linkedin, searching the web, etc. These are junior roles that hires new to sales are given when they start, and people who want to focus on sales usually graduate from them within six months to a year. Any sales person with experience won’t have had to generate their own leads for a long time, they’re used to having a team behind them. Even if they’re willing to roll up their sleeves and commit to what’s consider a low-status job, they won’t have a good idea of how to do SDR for a novel product. Startup Incentives are Long Term One of the best sales people I met described himself as “coin operated”, and the usual incentive structure is set up to reinforce that attitude, since sales people make most of their earnings through commissions on a quarterly basis. This isn’t a good fit with an early-stage startup because you’re probably going to be making proof-of-concept deals initially where the time to close is uncertain and the revenue is small. A 10% slice of that isn’t interesting compared to the steady, large income stream they get at an established company. The alternative is setting up performance-based bonuses (for example $x for each paid pilot signed) but even that is unlikely to be a very compelling amount for them. The hope of course is that you can convince candidates to focus on the stock they can earn, but coming from a world where incentives are liquid cash they get within a couple of months, it’s a hard perspective switch to make. They’ve chosen comparatively low-risk compensation for years, why are they going to change now? Market Discovery If there’s one thing I’m certain of, it’s that you won’t end up selling to the companies you thought you would at the start. As you learn more about your product and people’s needs, you’ll inevitably adjust who you’re targeting. This is a problem because most senior sales people have a lot of experience in a particular industry, but those skills aren’t portable. They may know the customer needs and have warm relationships with key players in one market, but when your startup changes focus they’ve lost all of those advantages that they’ve spent years building. Even changing the sales model within a single industry will have a big impact on their effectiveness. Someone who has spent years doing high-touch, long sales cycle engagements is going to be starting from scratch if you move to self-serve subscriptions. So, What Has Worked? If hiring established sales leaders didn’t work for us, what has? The first thing I had to learn was that a lot of the work I was thinking of as sales was actually business development. Closing deals is a job for sales, but there will be a lot of other steps before that, like figuring out which role in an organization to reach out to, developing materials, finding conferences where decision makers attend) that are much more about BD. Think about hiring someone with those skills first, before you get a sales person. What worked for us was finding somebody super-keen who has a business background, but was early in their career, and willing to take on the time-consuming BD work with a song in their heart. The feedback has been that it’s great experience for them, and a lot more interesting than most MBA jobs at that level. You should also prepare to spend a lot of time on sales yourself. The first few sales are going to be founder-led, and there’s a lot to learn to be successful, so take it as a serious time commitment. Customers prefer talking to founders over salespeople. Founders know the product better than anyone, can answer technical questions, and bring the passion. If you can get to the point where there’s a license to be closed, you have a much better chance of making it happen than anyone else in the company. Happily you don’t have to go it alone. Good advisors can be incredibly helpful in figuring out domain-specific and process-related questions, as well as being able to introduce you to the people you should be talking to. Find someone who’s got a lot of experience and contacts in the industry and get them excited about what you’re doing, they can be a massive help. A lot of good later-career people are bored because their job is no longer as challenging, so they can be surprisingly open to taking an advisory role for equity. Think about people like lawyers in your field too, they are often very well connected and will know a lot about the actual sales process. There’s so much inertia at most companies, cultivating champions within your target companies is the only effective way I’ve found to make things happen. You need someone who’s willing to be a pest on your behalf to avoid getting stuck in an endless sales purgatory. To get that level of engagement you have to make sure they feel included in your decision making and invested in the success of your startup. One way is to set up an advisory board that includes any promising champions, that way they get bragging rights if you succeed, they can network with other key industry people, and you can give them an advisory stake too, as long as that works ethically. I’d imagine that having another founder with good sales experience would have save me learning a lot of these lessons the hard way, but if you’re starting with a technical team, resist the urge to bring in somebody to “handle sales”. It’s so critical to the existence of your startup, it’s not something you can hire your way out of. As CEO, getting those early sales across the line has taken up the majority of my time, even more than product direction and hiring, and I wish I’d embraced that earlier. There are a lot of ways to get help from other people, but at the end of the day only a founder can close those crucial deals.
Welcome back to This Week in Stratechery! As a reminder, each week, every Friday, we’re sending out this overview of content in the Stratechery bundle; highlighted links are free for everyone . Additionally, you have complete control over what we send to you. If you don’t want to receive This Week in Stratechery emails (there is no podcast), please uncheck the box in your delivery settings . On that note, here were a few of our favorites this week. This week’s Stratechery video is on Robotaxis and Suburbia . What the Times Missed in Its David Sacks Story. On Sharp Text this week, I wrote about the commotion that ensued in tech and media after the New York Times profiled Trump Crypto and AI Czar, David Sacks, including an OpenAI-style outpouring of Sacks support, why the piece failed on its own terms, and an entirely different story that went unexplored. While the Times focused on the private interests that may benefit under Sacks’ watch, there are better questions about the public’s interest in leaning on someone like Sacks , and why the government might need Silicon Valley expertise as it confronts a variety of tech questions that have enormous implications for the future of the Western world. — Andrew Sharp Atlassian’s History and the Near Future. My favorite part of every Stratechery Interview is Ben’s “how did you get here?” question to first-time interview guests, and this week’s interview with Atlassian CEO Mike Cannon-Brookes is a terrific entry in the series. Come for the story of how a Qantas Frequent Flyer program eventually led to a $40 billion software business in Sydney, and stay for Cannon-Brookes on how his company is adapting to the AI era, as well as his take on “correct, but chronologically challenged” snake oil salesmen. Finally, as a rabid F1 fan, I’d be remiss if I didn’t recommend the end, where Cannon-Brookes expounds on Atlassian’s role sponsoring and helping to transform the once moribund Williams team (a story that can also be marketed to enterprises the world over). — AS Code Red at OpenAI. I have, for three years now — i.e. ever since ChatGPT took the world by storm in November 2022 — been convinced that we were witnessing the birth of the next great consumer tech company. Today, however, there are very legitimate reasons to be concerned that OpenAI is going to eventually succumb to the Google behemoth, just as Yahoo, Microsoft, Blackberry, and countless others have; I still want to believe that OpenAI can be an Aggregator, but they don’t have the business model to match, and that may be fatal. I summarized all of these feelings in this week’s episode of Sharp Tech , which covered both this week’s Article about OpenAI and Nvidia angst , and Tuesday’s Update about the bear case for OpenAI . — Ben Thompson Google, Nvidia, and OpenAI — OpenAI and Nvidia are both under threat from Google; I like OpenAI’s chances best, but they need an advertising model to beat Google as an Aggregator. OpenAI Code Red, AWS and Google Cloud Networking — OpenAI is declaring code red and doubling down on ChatGPT, highlighting the company’s bear case. Then, AWS makes it easier to run AI workloads on other clouds. AWS re:Invent, Agents for AWS, Nova Forge — AWS re:Invent sought to present AI solutions in the spirit of AWS’ original impact on startups; the real targets may be the startups from that era, not the current one. An Interview with Atlassian CEO Mike Cannon-Brookes About Atlassian and AI — An interview with Atlassian founder and CEO Mike Cannon-Brookes about building Atlassian and why he is optimistic about AI. The Forest the New York Times Missed Among the David Sacks Trees — The New York Times failed to support its David Sacks headline, and ignored better questions about the how U.S. devises modern tech policy. Google Looms Alan Dye Leaves Apple Let’s Break Down the 45nm Process Node A Quiet Chinese Mobile Giant in Africa Trump, Takaichi and a Game of Telephone; Japan Jawboning Continues; An Internet Governance Study Session; China Making Trade ‘Impossible’ Wolves and Cavs Concerns, The NBA Cup in Year 3, Questions on the Magic, Suns, Thunder and Raptors The Game of the Week, A Giannis Inc. Emergency Board Meeting, Chris Paul Gets Cut at 2 a.m. in Atlanta OpenAI Declares a ‘Code Red,’ Alan Dye Leaves Apple for Meta, Questions on Tranium 3, Substack, and F1
[Editor's Note: this piece previously said "Blackstone" instead of "Blackrock," which has now been fixed.] I've been struggling to think about what to write this week, if only because I've written so much recently and because, if I'm honest, things aren't really making a lot of sense. NVIDIA claims to have shipped six million Blackwell GPUs in the last four quarters — as I went into in my last premium piece — working out to somewhere between 10GW and 12GW of power (based on the power draw of B100 and B200 GPUs and GB200 and GB300 racks), which...does not make sense based on the amount of actual data center capacity brought online. Similarly, Anthropic claims to be approaching $10 billion in annualized revenue — so around $833 million in a month — which would make it competitive with OpenAI's projected $13 billion in revenue, though I should add that based on my reporting extrapolating OpenAI's revenues from Microsoft's revenue share , I estimate the company will miss that projection by several billion dollars, especially now that Google's Gemini 3 launch has put OpenAI on a " Code Red, " shortly after an internal memo revealed that Gemini 3 could “create some temporary economic headwinds for [OpenAI]." Which leads me to another question: why? Gemini 3 is "better," in the same way that every single new AI model is some indeterminate level of "better." Nano Banana Pro is, to Simon Willison, " the best available image generation model. " But I can't find a clear, definitive answer as to why A) this is "so much better," B) why everybody is freaking out about Gemini 3, and C) why this would have created "headwinds" for OpenAI, headwinds so severe that it has had to rush out a model called Garlic "as soon as possible" according to The Information : Right, sure, cool, another model. Again, why is Gemini 3 so much better and making OpenAI worried about "economic headwinds"? Could this simply be a convenient excuse to cover over, as Alex Heath reported a few weeks ago , ChatGPT's slowing download and usage growth ? Experts I've talked to arrived at two conclusions: I don't know about garlic or shallotpeat or whatever , but one has to wonder at some point what it is that OpenAI is doing all day : So, OpenAI's big plan is to improve ChatGPT , make the image generation better , make people like the models better , improve rankings , make it faster, and make it answer more stuff. I think it's fair to ask: what the fuck has OpenAI been doing this whole time if it isn't "make the model better" and "make people like ChatGPT more"? I guess the company shoved Sora 2 out the door — which is already off the top 30 free Android apps in the US and at 17 on the US free iPhone apps rankings as of writing this sentence after everybody freaked out about it hitting number one . All that attention, and for what? Indeed, signs seem to be pointing towards reduced demand for these services. As The Information reported a few days ago ... Microsoft, of course, disputed this, and said... Well, I don't think Microsoft has any problems selling compute to OpenAI — which paid it $8.67 billion just for inference between January and September — as I doubt there is any "sales team" having to sell compute to OpenAI. But I also want to be clear that Microsoft added a word: "aggregate." The Information never used that word, and indeed nobody seems to have bothered to ask what "aggregate" means. I do, however, know that Microsoft has had trouble selling stuff. As I reported a few months ago, in August 2025 Redmond only had 8 million active paying licenses for Microsoft 365 Copilot out of the more-than-440 million people paying for Microsoft 365 . In fact, here's a rundown of how well AI is going for Microsoft: Yet things are getting weird. Remember that OpenAI-NVIDIA deal? The supposedly "sealed" one where NVIDIA would invest $100 billion in OpenAI , with each tranche of $10 billion gated behind a gigawatt of compute? The one that never really seemed to have any fundament to it, but people reported as closed anyway? Well, per NVIDIA's most-recent 10-Q (emphasis mine): A letter of intent "with an opportunity" means jack diddly squat. My evidence? NVIDIA's follow-up mention of its investment in Anthropic: This deal, as ever, was reported as effectively done , with NVIDIA investing $10 billion and Microsoft $5 billion, saying the word "will" as if the money had been wired, despite the "closing conditions" and the words "up to" suggesting NVIDIA hasn't really agreed how much it will really invest. A few weeks later, the Financial Times would report that Anthropic is trying to go public as early as 2026 and that Microsoft and NVIDIA's money would "form part of a funding round expected to value the group between $300bn and $350bn." For some reason, Anthropic is hailed as some sort of "efficient" competitor to OpenAI, at least based on what both The Information and Wall Street Journal have said, yet it appears to be raising and burning just as much as OpenAI . Why did a company that's allegedly “reducing costs” have to raise $13 billion in September 2025 after raising $3.5 billion in March 2025 , and after raising $4 billion in November 2024 ? Am I really meant to read stories about Anthropic hitting break even in 2028 with a straight face? Especially as other stories say Anthropic will be cash flow positive “ as soon as 2027 .” And if this company is so efficient and so good with money , why does it need another $15 billion, likely only a few months after it raised $13 billion? Though I doubt the $15 billion round closes this year, if it does, it would mean that Anthropic would have raised $31.5 billion in 2025 — which is, assuming the remaining $22.5 billion comes from SoftBank, not far from the $40.8 billion OpenAI would have raised this year. In the event that SoftBank doesn't fund that money in 2025, Anthropic will have raised a little under $2 billion less ($16.5 billion) than OpenAI ($18.3 billion, consisting of $10 billion in June split between $7.5 billion from SoftBank and $2.5 billion from other investors, and an $8.3 billion round in August ) this year. I think it's likely that Anthropic is just as disastrous a business as OpenAI, and I'm genuinely surprised that nobody has done the simple maths here, though at this point I think we're in the era of "not thinking too hard because when you do so everything feels crazy.” Which is why I'm about to think harder than ever! I feel like I'm asked multiple times a day both how and when the bubble will burst, and the truth is that it could be weeks or months or another year , because so little of this is based on actual, real stuff. While our markets are supported by NVIDIA's eternal growth engine, said growth engine isn't supported by revenues or real growth or really much of anything beyond vibes. As a result, it's hard to say exactly what the catalyst might be, or indeed what the bubble bursting might look like. Today, I'm going to sit down and give you the scenarios — the systemic shocks — that would potentially start the unravelling of this era, as well as explain what a bubble bursting might actually look like, both for private and public companies. This is the spiritual successor to August's AI Bubble 2027 , except I'm going to have a little more fun and write out a few scenarios that range from likely to possible , and try and give you an enjoyable romp through the potential apocalypses waiting for us in 2026. Gemini 3 is good/better at the stuff tested on benchmarks compared to what OpenAI has. OpenAI's growth and usage was decelerating before this happened, and this just allows OpenAI to point to something. Its chips effort is falling behind , with its "Maya" AI chip delayed to 2026, and according to The Information, "when it finally goes into mass production next year, it’s expected to fall well short of the performance of Nvidia’s flagship Blackwell chip." According to The Information in late October 2025 , "more customers have been using Microsoft’s suite of AI copilots, but many of them aren’t paying for it." In October , Australian's Competition and Consumer Commission sued Microsoft for "allegedly misleading 2.7 million Australians over Microsoft 365 subscriptions," by making it seem like they had to pay extra and integrate Copilot into their subscription rather than buy the, and I quote, "undisclosed third option, the Microsoft 365 Personal or Family Classic plans, which allowed subscribers to retain the features of their existing plan, without Copilot, at the previous lower price." This is what a company does when it can't sell shit. Google did the same thing with its workspace accounts earlier in the year . This should be illegal! According to The Information in September 2025 , Microsoft had to "partly" replace OpenAI's models with Anthropic's for some of its Copilot software. Microsoft has, at this point, sunk over ten billion dollars into OpenAI, and part of its return for doing so was exclusively being able to use its models. Cool! According to The Information in September 2025 , Microsoft has had to push discounts for Office 365 Copilot as customers had "found Copilot adoption slow due to high cost and unproven ROI." In late 2024 , customers had paused purchasing further Copilot assistants due to performance and cost issues.
Good morning, This week’s Stratechery Interview is with Atlassian founder and CEO Mike Cannon-Brookes . Cannon-Brookes and Scott Farquhar — whom I interviewed in 2017 — founded Atlassian in 2002; their first product was Jira, a project and issue-tracking tool, followed by Confluence, a team collaboration platform. Atlassian, thanks in part to their location in Australia, pioneered several critical innovations, including downloadable software and a self-serve business model; over the ensuing two decades Atlassian has moved to the cloud and greatly expanded their offering, and is now leaning into AI. In this interview we discuss that entire journey, including Cannon-Brookes’ desire to not have a job, how the absence of venture capital shaped the company, and how the company’s go-to-market approach has evolved. We then dive into AI, including why Cannon-Brookes believes that there will be more developers doing more, and why Atlassian’s position in the enterprise lets them create compelling offerings. Finally we discuss Atlassian’s sponsorship of Williams, the F1 race team, and why Cannon-Brookes thinks they can both help Williams win and also accrue big benefits for Atlassian. To repeat a disclosure I have long made in my Ethics Statement , I did, in the earliest years of Stratechery, take on consulting work for a limited number of companies, including Atlassian. And, for what it’s worth, I’m also a huge F1 fan! Go Max. As a reminder, all Stratechery content, including interviews, is available as a podcast; click the link at the top of this email to add Stratechery to your podcast player. On to the Interview: This interview is lightly edited for content and clarity. Mike Cannon-Brooks, welcome to Stratechery. MCB: Thank you for having me, Ben. So this is admittedly a new experience for me, I’ve already interviewed the founder of Atlassian , but it wasn’t you. I’m of course referring to Scott [Farquhar] . That was eight years ago, actually, before I even had podcasts. It was very brief, but hey, like I said, new experiences. MCB: That’s true. That’s true. And you wrote a consulting paper for us in 2014! I was going to disclose, yes, in the very brief period where I did consulting work, you flew me down to Sydney for a week, I had a chance to learn a lot about Atlassian. And on a personal note, that consulting contract helped me a lot, that was when I was just starting. It’s funny how small the numbers seem in retrospect, but maybe that’s why I’ve shied away from writing about you too much over the years, because it meant a lot to me. So I appreciate it and there’s my disclosure for the interview. MCB: Thank you. It’s a good piece of work. Don’t forget, ironically, we started as a consulting and services business and then decided that software was a better business model, so I think you did the same thing. You went the scalability route instead of the consulting work via Sydney. Absolutely. I’m not doing anything that doesn’t scale anymore, but I did love visiting Sydney, so it was great. MCB: Still, we pulled out the old consulting paper you wrote for us in 2014. Why are we going to win, why are we going to lose, everything else, it was classic Ben work. Was it good? MCB: It’s pretty good! It’s interesting, I’d probably be embarrassed if I read it today. Anyhow, the good news is that since it’s the first time I’m interviewing you, I do get to do my favorite segment, which is learning more about you. Where did you grow up, but also, where were you born? I know they were different places. Then, how’d you get interested in technology and what’s your version of the Atlassian origin story? MCB: Sure, I feel like I’ve heard this question 1,000 times! Where to start? My dad was in banking, he joined the glorious institution that is Citibank today, from England. Parents are both from Cambridge and bounced around the world a lot as part of that job. Took the, “Hey, we need someone to go to this country”, and he was like, “I’ll take that”. So I was born in America, in a period I lived in New York. To be honest, lived there for three months before I moved to Taiwan. Really? Whoa. I didn’t know that. MCB: Yeah, in 1980 when it was very different than what it is today. Yeah. Were you saving that to drop that off me? I had no idea. I thought you went straight from America to Australia. MCB: I only just thought about it about 30 seconds ago, actually. No, I went to Taiwan for a few years, lived in Hong Kong for a few years, went to Australia for a few years. So how I got into technology is actually related because my parents were moving around so much, the logic was being English, that they would send us to English boarding schools and that would be a stable thing while they were moving once we got old enough. So at the mighty age of seven, I was put on Qantas and sent to England and back four times a year to go to boarding school in England for about five, six years. Because of that boarding school, I have one of the lowest frequent flyer numbers in Australia, they introduced the frequent flyer program and that was at the end of year one or end of year two. I get given this catalog by my parents and how you’ve earned all these points, “What do you want to buy?”, and it’s like, “I don’t know, trips, winery things, booze”, I’m flicking through this catalog and I’m like, “There’s literally nothing in this catalog”, of gear that you used to be able to get that I wanted and at the back is this computer, so I was like, “I guess I’ll get that”. The only thing that was potentially age appropriate. MCB: That was the only thing in the catalog, I didn’t want a toaster, I didn’t want wine, so that became my first computer, the mighty Amstrad PC20 . Four colors, no hard drive. Eventually, I bought an external floppy drive, so you could put in two and did buy magazines and type in programs and write games and stuff from magazines and play with it, played a lot of video games basically back in that era. I was into computers peripherally all through high school, came back to Australia at 12, my parents had settled here by then and weren’t moving, and so I came back here, did all high school and university here. In high school, I was always going to be an architect, that was my dream the entire way through, but come to the end of grade 12, applied for a bunch of scholarships, because university, applied for the scholarships, ended up getting one and so I thought, “Oh, well, maybe I’ll take that”, and it was in a course called BIT. Basically, half computer science, half finance and economics, but it was 15 grand a year, tax-free, so I was like, “Well, I’ll do that for a while and go back to the architecture thing”. Of course, famously in that scholarship, I met my first business partner of my first startup, met my second business partner of the second startup, they went in radically different directions in terms of outcome, but it was just 30 kids right at the right time, did the dot-com era thing. Now, ironically, as a part of that scholarship, you had to spend six months in three industrial placements, so the origin story of Atlassian comes from then a little bit, because those industrial placements were so boring. Scott spent six months installing Windows at a large corporate and he was crazy freaking smart and it was like, “Hey, go from computer to computer and upgrade to Windows 98”, or whatever it was. It was like, “Guys, this is our life, this is going to be horrible”. I worked for Nortel Bay Networks, which was a good, at the time, massive competitor, Cisco then completely disappeared and so a good tech lesson in and of itself, I basically cataloged the room full of networking gear and routers, it was mind-numbingly boring. So towards the end of the university course, I famously sent an email to a few people saying, “Look, I don’t really want to get a real job, why don’t we start a company and we’ll try some stuff?”. And this was after the dot-com era? This was the early 2000s? MCB: This was after the dot-com era, yeah. So I lived through the dot-com era actually as a journalist and writer, analyst and technology. I worked for a company called Internet.com, which became Jupiter Media and Jupiter Research and that was great, that was an amazing era for me. We ran events, newsletters, what would’ve been podcasts, didn’t have them back then. And we ran events on Mobile Monday, I think one of them was called and it was all about WAP and— Well, the real secret is you’re not the only one. There are some founders that are very successful, that they’re like, “Look, I just want to pontificate about technology”. MCB: A little bit like you, I remember getting in a lot of trouble from some of the startups, because some company would launch and I wrote basically 500 words on, “This thing’s never going to work, this is a disaster of an idea”, and they would ring up and yell at my boss and he was awesome, he’d be like, “Dude, just keep writing what you think”, and it didn’t make you very popular as a journalist type. Anyway, emailed some people, tried to start a business, we didn’t actually know what we were going to do. Atlassian has, I always tell people, a terrible origin story. You should not copy us. You just didn’t want to be installing Windows or upgrading software. MCB: We literally did not want to get a real job. And Scott replied and said, “Yeah, sure, I’m in for trying that”. He was one of the smartest kids in our class and his nickname is Skip, because he was the president of our student association and always a leader type and Eagle Scout and everything else, so we’re like, “Yeah, okay, let’s do that, we’re good mates” — and that started Atlassian. We picked the name in about five minutes, which if you consulted any branding company, would not have been chosen. Ironically, originally, we were going to do customer service and consulting, that was what the gig was. Hence the name, because Atlas was a Greek titan whose job was to stand on top of the Atlas Mountains and hold up the sky, that’s what he was supposed to be doing. He was a bad guy, so his punishment was to hold the sky up and we thought that was an act of legendary service, and so we were going to provide legendary service by holding up the sky for customers and as I said, did the service thing for about six months, decided that this is a terrible business. People paying us $350 US to answer their questions and didn’t scale and was at crazy hours of the morning and night and everything else. So in the meantime, we wrote the first version of what became Jira . We actually wrote three pieces of software, one was a knowledge basey type tool, one was a mail archiving tool for groups, so you could see each other’s email as a shared archiving. And were you seeing this and you were building tools for yourself, for your consulting business? MCB: Literally, yes, exactly. So all three were tools that we needed for ourselves. People would email us and I couldn’t see Scott’s email and he couldn’t see mine at the time and it was like this is silly, and we built Jira to handle questions and issues and problems that we were having ourselves that became a teeny bit popular. There was this glimmer that someone else cared, so we poured all the effort into that. What was that? What was the glimmer? Because this is when Agile is taking over software development and at least the legend is Jira and Agile go hand in hand, is that a correct characterization? MCB: A little bit, but this is actually pre-Agile. So Jira comes out before Agile is even a thing. I think it was about two or three years before we had any version of marketing or feature sets that involved Agile. This was just a web-based, at the time, a bug tracker. So the interesting evolution part of the company obviously is it started as a bug tracker for software developers, it became an issue tracker for technology teams and now it’s like a business workflow for tens of millions of people every day across the world, most of whom have nothing to do with technology, so it’s gone on its own evolution. Would anything have been different if this was the plan from the beginning, or did it have to be this organic, “We’re figuring it out as we go along as we’re running away from Windows installations”, sort of story? MCB: I think, look, obviously, if we could choose to follow in our own footsteps, the Back to the Future skeptic in me would say it’s gone pretty well, so I’d follow every single footstep I took. (laughing) Yep, totally. MCB: And that would’ve become the plan. But look, we had two hunches really, which both turned out to be radically correct. Now, I would say we were following waves or whatever else, but one was that the Internet would change software distribution, which sounds ridiculous now and when I talk to graduates nowadays, I have to put them in the right time and place and say, “Look, when we started, software was distributed on a CD”, BEA WebLogic was the bee’s knees and you used to have to get it on a CD if you were lucky. If not, someone would come and install it for you and that’s how software was distributed. We made that CD into a ZIP file and put it on the Internet for people to download. You didn’t access it like a SaaS application, you literally download it from our website. Right. It’s funny that when you first say that, it’s like, “Oh, it’s completely transformative”, well, but you were an on-premises software story. But actually, no, there’s several steps to getting to SaaS, one of which is just downloading software. MCB: And we had people call us before they would download to check that we were real and stuff and I’m like, “Why don’t you just download the damn ZIP file?”, and I also date them, because, well, maybe I’ll get to the business model part, but the second innovation was that we thought open source would change software costs. So we had this big hunch, we were both writing a bunch of open source code at the time. Open source was a massive movement, especially in the Java space. Embarrassingly, I actually wrote a book called Open Source Java Programming that you can find with some mates. It’s still on Amazon and we sold a few thousand copies, I think, but I swore I’d never write a book again, it was a very painful experience. Thank you, you’re validating my life decisions . MCB: Yeah. Open source did bring the cost of building software down radically. We were writing a very small layer, 5% of the code at best on top of masses of amazing open source libraries and we contributed to those libraries, but we could deliver an amazing experience for a very low cost. We learned a lot, pricing and packaging. So what was the implication of that hunch though? Just that the market for developers, that would subsequently mean there was more software? MCB: A little bit that was the implication of the hunch. Largely for us, it was that the cost was going down. Pre-open source, you had to write everything so if Jira was back then, I don’t know, a million lines of code, if you added all the open source libraries together, it was 25, 30, 40 million lines of code. It was so big that it was so expensive, because you had to write all of that. To think of Windows, they wrote everything, the networking stack, there were no libraries, there was no open source involved in the original versions, it was all written by Microsoft. So the cost of that was very high, then you had to charge a lot of money. So we thought, look, if we could take all these amazing open source libraries, contribute back to them — we were a great open source citizen — and build a piece of proprietary software on top of them that solved customer’s problems, we could deliver that really cheaply. In fact, we sold the original versions of Jira, they were $800, unlimited users, unlimited use with no lifespan. So it was just 800 bucks, one-time fee forever and we learned a lot about pricing and packaging firstly, but secondly, it was very simple. Our goal in the early days, we had to sell one copy a week to stay alive, that was it. Some weeks, we’d sell two copies. $1,600 US would roll in and we’d be like, “Cool, we got a week off to survive”, and then one copy a week became two and two became five and five became ten, and now it’s hundreds of thousands. Well, isn’t the thing you just didn’t want to have a job? So I love this part of the story, because when I started Stratechery, I had a job from Microsoft that made, I think, $104,000 or something like that. I’m like, “I just want to make that, because I don’t want to work for a corporation, so if I could just get to there, it’ll be great”. MCB: We had exactly the same sets of goals. We had a few things we wanted to make somewhere that we wanted to go to work. I wanted to get up every day and think, “I want to go to work”, and weirdly, almost 24 years later, I love coming to work, so a tick achieved. We wanted to make it so we didn’t have to wear a suit, neither of us really like wearing suits at all — in fact, it’s a bit of an allergic reaction often and so tick, don’t turn up to work in a suit every day. And thirdly, most of our friends, so this is right where IBM bought PwC ironically, so out of the 30-odd kids in our class, maybe 10 went to IBM as consultants and 10 went to PwC and then they all end up going to the same shop and their grad salary there was $47,600. So our goal for year one was to end the year making at least a grad salary and convince ourselves we’re not crazy kind of thing and we smashed that goal, so that was good, but that was there. The Internet, the distribution part is important, knowing your favorite topics. Tell me about that and along with the business model, because again, this goes back so far, I don’t think people appreciate the extent to this entire idea of self-serve or bottoms up selling. This is really where it all started. MCB: Yes. And look, a few things. Firstly, if you come from Australia, we’re an exporting nation. “We’re built on the sheep’s back”, is a phrase, Australia’s built on the sheep’s back. What that really means is because we were this colony originally, then country on the far side of the world, anything we did to make money largely had to leave the country and go somewhere else. Originally, that was a struggle to find a product that could do that. “Built on a sheep’s back” is because wool was the first product that could do that, you could put it on a wooden boat, because it wasn’t very heavy and you could ship it a long distance, because it kept really well, so we could make sheep’s wool and make money as a country by shipping it back to Europe and it could survive the journey and so the country was built on the sheep’s back. We are a massive exporting nation. Trump brings in his tariffs, we’re the only country with a negative rate of return, we have a positive trade relationship with America and we’re like, “Wait a second, why did we get taxed?”, so obviously, it’s rocks, technology, we build and export everything as a country that we do. So our mentality was like, “Well, if we’re going to make money, it’s going to be overseas”, that was the first thing, is, “Okay, it’s going to be somewhere else, it’s not going to be Australians buying our software”, and so the Internet allowed us to do this. We put up a shopfront, early website and people could come to our website, download our software and then we just needed a way to get paid for it. The problem was in order to do that and the trust barriers of the Internet, we had to have a very low price and we had to have a fully installable offering. So we spent so much time on making it installable, documentation, “How would you get yourself up and running and try it?” — the software, as we put it, had to sell itself. Our software had to be bought, not sold. We didn’t have any salespeople, we couldn’t travel to your office in Sweden or London and help you out with it. For $800, we couldn’t have done that and secondly, it didn’t make any sense. So the evolution was, “Okay, this is the only possible path that we can go down is we have to figure out how to get people to do this”, now it turns out once you have figured out how to do that, it’s an incredibly powerful motor because you have lots of people coming, you have a very cheap piece of software for its relative performance, and you get people using it in all these big businesses all over the place. I would say 50% of the customers I go meet nowadays, probably meet a handful of customers, a couple a day on an average kind of thing, many of those have been a customer for 20 years, 22 years, 23 years. How many customers have been a customer 23 years? I’m like that’s crazy, we’re only 24 years old. That’s awesome. MCB: And so they downloaded very early, they didn’t download as all of , all of them are customers. Just one guy who’s like, “I need a way to track my issues”. MCB: Exactly. It was some guy in a backroom who needed to track it. I know the Cisco origin story, that was literally a guy, he’s still there, he’s been there 22, 23 years, he’s awesome. And they started with just, “I just needed a way to manage my issues for 10 people”, and now it’s hundreds of thousands of people, seats that we have there, it’s kind of grown over time. How did we know that business model was working? Again, it dates us a lot, this didn’t mean we didn’t answer questions, we were big on customer service and helping people, email was the way to do that. A bit of IRC back then, we had a channel you could log into and we’d help you. But the first customer, we used to walk into the office in the morning and we had a fax machine with literally rolls of paper. So if you wanted to pay for this distributed software, this says how old, there was no SSL keys, I heard you complaining about it the other day, totally agree with that era. You had to download a PDF off our website, which was pretty modern that it was a PDF, fill in your credit card details, and fax it to us, that is how you paid when we started. So we would walk in the morning and there’d be these rolls of paper on the ground, you be like, “Ah, sweet, someone bought something”, you know what I mean? It became a weird dopamine drug for us. The very first company was American Airlines… MCB: About six months in that we came in the morning and there was a fax on the ground with $800 and a credit card number written on it and we had never talked to American Airlines, they had never emailed us, they had never asked for customer service, they’d never gone on IRC, they had never talked to us in any way, shape or form. Man, this thing could work, we just made $800 out of the air. MCB: I mean, there was a lot of pre-work to get them there, but obviously that was kind of different. MCB: Then secondarily, as you wrote, I’m just trying to finish a very long answer here, we started Confluence in 2004, and those two became the jewel engines and both of those I think were probably major moments. I often say Confluence is a bigger moment, actually. The business model was kind of established, this is two years into the business. We made, I think, $800 grand in year one, $1.6 million in year two, maybe $5 million in year three, and $12 million in year four, if I remember the revenue numbers. So the thing was working really well. You’re the company that’s the Microsoft heir in some respects, which is the really just you took venture eventually, but didn’t really need to, just pure bottoms up. You and Scott, we’re able to keep a huge portion of the company because of that, it’s an amazing story that is, I think, under-told in some respects. MCB: Yeah, well, we actually did. I mean, we did and didn’t. So the venture story is one of my favorites because it describes how we think from first principles. Firstly, the first capital we put on the balance sheet, institutional capital to put on the balance sheet, I guess you could argue our initial, I don’t know, $10 grand each was some money, but was in the IPO . So in 2015, when we went public, that was the first capital that went into the business all time. We took two rounds of funding, one in 2010 and one in 2013, but both of which were to employees, the first was to the founders and the second was to large number of employees who bought in so both of those companies bought ordinary stock. Secondary shares basically, yeah. MCB: They bought ordinary stock, there were no preferences, there were no anything, that was kind of the way it is. And we love the Accel guys that invested, it’s kind of funny because their business model was wildly wrong, we now have their original spreadsheets and stuff. We’ve 15 years in, you know them really, really well, they wanted us to grow it. I think we had to grow at 30% for two years, 20% the year after and something like that to double or triple their money and at the time they put in $60 mil US , that was the largest investment I think Accel had ever made in anything software, digital kind of world and it was this massive bet. It was a one-page term sheet for ordinary stock, so credit to those two partners who took massive risk on us, had to fight, we know that GC, everybody else to do this unusual funding round and I think we did 50% growth the first year, and our CAGR since then is probably 40%. Yeah, it worked out pretty well. MCB: They did very well. I think their 2-3x was more like a 300x or something. You mentioned the Confluence moment. Why was that a big deal? Usually the story is you have one product and you need to focus and you’re two years old, you’re launching a completely new product. Is that the aspect you’re referring to? MCB: Yes, I think it comes down to being bootstrapped. Look, we spent nine years convinced we were going to die every day, there was just such a mentality that this thing was all going to fall over and we better work harder and keep going. The Confluence moment was important because I remember, I don’t know exactly, but sometime around then we understood venture capital. Firstly, on the venture capital side, because they do relate to each other, there was no VC available in 2001 and 2002 in Australia. We’re a nuclear winter, we’re two idiots with no credibility. Right. You could barely get funded in San Francisco, you’re not going to get funding in Sydney. MCB: No, because 2001, you weren’t even finding San Francisco funding because the whole dot-com boom had just happened, no one was getting funded anyway. We’re in Australia and we have no credibility, so we didn’t even bother. We literally, 2010 when we went to the Accel thing and we talked to five VCs, was the first time we’d ever pitched the business. It was just not a thing, people don’t understand, we used to say we were customer-funded when people would ask the also awkward question of, “Who’s your funding come from?”, we were like, “We’re customer-funded”, They go, “Oh, okay”. Lifestyle business! MCB: But we did understand venture capital, massive readers, I have an army full of technical books, books about technology and the industry and history and stuff from that magic era of airport bookstores. We read every episode of Red Herring and Industry Standard and Wired Magazine, I have just this huge library, so voracious readers. One thing you understood about venture capital is they put the portfolio theory on their side — and I’m a big fan of venture capital, I should say, I’m the chair of Australia’s biggest VC fund and that’s my other mate that I met in university, Niki Scevak . But we wanted portfolio theory on our side, we’d done finance and economics, we had one product, this was highly risky if you’re bootstrapped. So there was a little bit of the thinking that actually if we have two products, our chances of total failure are less, one of them can fail and we’ll be okay and so we started a second product. Yes, arguably it was hard, but our first one was going all right, it was like making, I don’t know, five million bucks a year and we had a handful of really awesome backpacker programmers. And the early people, it’s like a whole total band of misfits that somehow made this thing work and we’re having a lot of fun, we’re working really hard and so we made another internal tool that became Confluence and being adjacent, but very different, selling to different audiences, but having a lot — if you bought one, there was a good reason to have the other one, no matter which way you started, became a really good symbiotic loop of these two engines that powered us for a very long time. So it was more a case of reducing our risk actually than anything else. Wasn’t it risky to be splitting your resources or did that not even occur to you? MCB: I don’t think it occurred to us, no. It was more about splitting our risk and we were doing pretty well, but it changed the business because we moved from being the Jira company to a software company, and I say that’s probably the most under-understood moment because we had to learn about not how to market Jira, but how to market software, not how to build Jira, but how to build software. So now we have 20, 25 apps in 5 different categories that sell to all sorts of different teams who own a business, but we had to become a software company. Microsoft, I don’t know the analogy’s really that fair to them, to be honest, or fair to us, it seems massively over-glamorizing what they’ve achieved, which is amazing, I’m huge fan of Microsoft. The need to understand how to sell, in their case, like Minecraft, SQL Server, Azure, AI, you have to understand the building, the creation of technology, the selling of technology, the marketing of technology at a generic level, it really helped us generify the business. I think if we’d gone too much longer, everybody would’ve been on the Jira team, it would’ve been too hard to start a second thing and instead, we’ve always been a multi-product company. You just mentioned selling a lot. When did you finally realize or transition away from just being self-serve to actually, “We’ve got to grow beyond this”? Was it almost like a pivot that came too late because your identity was so wrapped up into the, “We’re the self-serve company”? MCB: Look, it’s never been a pivot, I get asked this by investors all the time. I would say our go to-market model and our process has kept evolving pretty much every year or two for 20 years and I say evolving because we’re very aware of the strengths of the model that we came up with and we’re very aware of what it takes to power that and we’ve been very careful when we’ve evolved, changed, added to it, not to destroy the original one. So nowadays, we have two amazing business models where we call them high-touch and low-touch. So we have the low-touch model, which is literally the same thing as it’s always been, hundreds of thousands of people show up every week, they try our software, we want them to have a great experience trying the software, we want to spread it as widely as possible and as many enterprises as we can, and some of those will stick, some of those will get working and we measure aggressively the rates of return and dollars and flows and funnels and everything else. This whole team whose job is to make sure that that’s working at now massive scale, right. But at the same time, what happened is as customers got more and more Atlassian software deployed, they wanted a different relationship with us, they wanted a bigger relationship. Those days they used to be spending, as soon as we were spending $20 grand, we were like, “Oh man, maybe we should talk to these people”, nowadays it’s more like around $50 to $100 grand is when we’ll talk to you. So the lines kept moving for different reasons and we actually have online sales, inside sales in between actually, the sort of classical someone gets on an airplane and goes to travel to you. So it’s just kept evolving. We talk about the IPO a lot, it’s our 10-year anniversary coming up this month, I’m off to New York next week to ring the bell and celebrate 10 years. When we went public, as an example, we had less than 10 companies paying a million dollars a year, now we’re well north of 500 in 10 years. So that doesn’t come without an amazing enterprise sales team and teams that go out and help customers and customer success and all the trappings of a really top flight enterprise sales organization, because for most of those customers, again, I think it’s north of 85% of the Fortune 500 are deep Atlassian customers. We become a strategic partner to these businesses that if we go down, rockets don’t take off, banks shut down, it’s a real critical importance to most of these customers. How big is your business outside of directly working with developer teams? As I recall, this was part of the consulting thing was you were wanting to do Jira for sales or Jira for all these different sort of functions, where and how did that evolve? MCB: So it’s been a continuum for a long time. So nowadays, less than half of our users are in technology teams, and probably a third of those are developers, less than half of them. So a portion of our audience, it’s a very important point of words. When I talk about this, all the engineers are like, “Hey, you don’t care about us anymore”, I’m like, “No, that’s not true”, that business is a great business, it’s just the rest of our business has grown massively around it. There are not enough developers in the world for our business. Our fundamental value has always been actually, and it took us one of these things, it took a decade to realize, firstly, we don’t solve technology problems, we never have, we’ve never had anything that’s like, “I care what code you write, which language the code is in, what the code does”. We solve collaboration and people problems, we always have solved people problems, even Agile was a people problem. It’s not a technology problem, actually, it’s a people problem. It’s, “How do we organize a group of people to build a piece of technology that best meets the customer’s needs and goes off track as little as possible?”, that is a collaborative people problem, we’ve always solved people problems. Our value actually came because there’s a lot of tools for technology teams and we never wanted to be in the dev tools business, that’s a road of bones, it’s very hard to build sustainable competitive advantage and dev tools, the history shows this. There’s just a different company every few years, developers tastes are fickle, our developers taste are fickle, this is not me sledging developers at all, we have a massive R&D arm and that group changes languages every couple of years, they change how they build software every couple of years, they’re constantly moving on, they change our analytics tools and everything else because they are tool builders and toolmakers, that makes sense, but that’s a hard place to build a business. Interestingly topical today, so we’ll see. But the easier place to build a business in the long term was the level above that, which is the collaboration problems that came, which started as, “How do we get engineers, designers, product managers, business analysts to all be on the same page about what it is that they’re building and have a repeatable process for that?”. It turned out that as the world has become technology-driven, as we say, our customers are technology-driven organizations. If you’re a large organization for whom technology is your key distinct advantage, it doesn’t matter whether you’re making chips and databases or whether you’re making rockets or cars or whether you’re making financial services or insurance or healthcare, I would argue for most of the businesses that are great, technology is their key competitive advantage, then you should be our customer, that is it. And what we help you do is we help your technology teams and your business teams collaborate across that boundary because that’s actually the hardest boundary. Building great technology is one set of problems, making it work for your customers usually means in different industries, a different amount of working with all sorts of business people and that’s what Jira did from the very start. Now that’s what our whole portfolio in service management, in strategy and leadership teams is about doing that at different scales and different amounts in different places. Does it bug you when you get complaints on the Internet of, “Jira’s so complicated”, “Hard to use”, blah, blah, blah? And are you speaking to, the problem is that the problem space we’re working in is not the single developer trying to track an issue, it’s trying to herd a bunch of cats and get them the same direction and muddling through that is a lot more difficult than it seems. MCB: It bothers me anytime people don’t like our software, sure. We’ve worked for the last 20 years to make it better every day. We’ll probably work for the next 20 years to make it better every day and people will still probably be dissatisfied and that is our fundamental core design challenge. There’s a few reasons they say that. Firstly, the on-premise business model and the cloud shift is really important because with the cloud shift, we update the software, with the on-premise business model, we don’t, so you would often be on older data versions, customers would upgrade once a year or every two years or something, and so we can’t control that. Secondly, the challenge of Jira is at our core, we solve a whole lot of what we say is structured and unstructured workflows. Confluence is an unstructured workflow, Jira’s a very structured workflow. You have a set of steps, you have permissioning and restrictions, you have fields, you have what’s happening in this process. The auditor will do something and pass it to the internal accounting team, the accounting team will do this and pass it to legal, legal will do this and pass it to these people. You’re defining a workflow and you’re having information flow back and forth and a Jira work item is, as we call it, it’s a human reference to work. That’s the best description of what Jira is work in the knowledge work era is this very ephemeral concept. Back to your development example, is the code the software? Is the idea the software? Is the designs in Figma — these are all parts of what it is, this thing that’s called this virtual thing that we’ve built. What we track is with a human reference to that, so someone can say it’s a new admin console. Cool, here’s the design for the admin console, there’s the spec for the admin console, there’s the code for the admin console, here’s where it’s been tested, here’s where it’s deployed. Did customers like it? We need a reference to this thing that is otherwise spread across hundreds of systems and virtualized. Once you’re building a workflow system, companies, ours included, love process, we love workflows, we love control, and that control usually comes with more data. “Hey, don’t fill in these three fields, fill in these 50 fields”, and they’re all required for some reason and our job to customers is to say, “Do you really need 50 fields?”, because you’re creating a user experience- You’re ruining it for us! MCB: Your users are going to have to fill in all 50 fields, and it feels like that’s going to take you a while. We have customers — I went back and checked, I think almost every single person you’ve interviewed on your podcast is a customer of ours. I don’t know if it’s 100%, but it’s definitely north of 95% out of the last 20 guests. Stratechery is a customer of yours, so there you go. MCB: Oh, really? Well, there you go. Thank you. One of my engineers adores Jira, so I get the opposite angle from what I asked about. MCB: That’s right. So look, it’s a challenge for sure, but at the same time, man, the value we’ve created, the business value, the number of customers that run on it, it’s ironic, we talk about the AI era and all these other things. Literally, no chips go out of any of the chip companies you love talking about, every single one of them, soup to nuts. So at what point did you realize that AI was going to impact you in a major way? Was there an “aha” moment or it’s just been in the air? Or is it a specific time you realized, “Look, this is going to completely change what we do?” MCB: Again, I’m one of these — I’ve realized I’ve become the old man in the room. We’ve done machine learning for a long time in lots of ways because of our online business model, so I’d say we’ve done AI for a long time. Obviously, LLMs are what people refer to nowadays by AI and agents and these words that have corrupted the entire thing, the meaning changes in technology when it means something else. The launch of various versions of ChatGPT were very instructive obviously, they were a moment for everybody. The optimism, and I would say we’re massive AI optimists, it is the best thing that’s happened to our business in 25 years. Why? Because people might look at you from the outside and say you’re still characterized as — even though your business expanded far beyond developers — “Oh, you have a lot of developers”, I’m skipping over the transition to the cloud just because we’re running out of time, but it’s an interesting story. You did announce you are finally ending the on-premises software, which I’m curious, it is a sentimental moment to come to that decision, but people might look at you from the outside and say, “Oh, there’s a company that’s going to have a problem with AI, AI is going to replace developers, it’s the decreased seats . What are they going to do?” MCB: There’s a few ways to take that. I’m trying to put it on a tee for you. I think I know what you want to say. MCB: There’s a few ways to look at it. Firstly, I think AI is a good example where people are very concrete about the negatives and the positives are upside. I think it’s a huge force multiplier personally for human creativity, problem solving, all sorts of things, it’s a massive positive for society. That doesn’t mean there aren’t any negatives, but the net effect is really high. And we spend a lot of time, you hear it in the media talking about the job loss, the efficiency gains, whichever way you want to put it, that’s the thing. Well, that’s because it’s really concrete in a spreadsheet, “I can do this process with half as many people”, “Wow, look at that, that’s great”, what’s never written in the spreadsheet is all the new processes that get created, all the new ways of doing things, the quality of the output is going to be twice as high. If software costs half as much to write, I can either do it with half as many people, but core competitive forces, I would argue, in the economy mean I will need the same number of people, I would just need to do a better job of making higher quality technology. So our view on AI overall is an accelerant, not a replacement to everything we do, and just the next era of technology change is really positive. We’ve loved technology, we love the cloud, we love all the tech changes we’ve been through, mobile. Look, us as a business, we are in the game of knowledge work. We solve human problems, workflows, business processes, this is what we do. These largely revolve around text, or if it’s video nowadays, that can be reduced to text in various ways. LLMs allow us to understand that text in a massively deeper way than we ever have been, and the problems we solve aren’t going away. 20 years time, there’ll be groups of people trying to solve some sort of problem as a team and working on a project, and so these things aren’t going to go. They’re going to need to talk to each other and collaborate of what work’s going on and how it’s working, so the textual aspect of it has been amazing. The features we’ve been able to ship, we never could have built five years ago, it was literally impossible, so the ability to solve customer problems is so much higher than it ever has been. Secondly, our software is incredibly valuable at the core of these workflows, but it’s also incredibly promiscuous. What I mean by that is we have always been very highly interlinked with everything else. If it’s a sales team, there are links to Salesforce and customer records, there are links to internal systems, there are links to maybe features that need to be built, there are links to some content and document. So any Jira, Confluence, or Loom , you don’t record a Loom unless you’re talking about something, you don’t have a Jira issue without pointing to all sorts of different resources, whether that’s a GitHub or Figma, whether it’s Salesforce or Workday. That gives us a really unique knowledge, which we’ve turned into the teamwork graph, that actually started pre-AI, so the irony is the Teamwork Graph is about 6 years old. Well, it started with Confluence. This is the whole thing where you look backwards, and to your point, if you had just been the Jira company, but because from the very beginning, you mentioned Confluence was different but it was adjacent and you had to build the links and stuff together, and as you build all these different tools, because everyone wants to be this point of integration. And I wanted you to tell me about Rovo and this idea of being able to search across all your documents. Who gets permission to do that? It’s someone that’s already there, and you made the critical decision to be there back in 2004 or whatever it was. MCB: That’s true. Certainly back in 2004, and then in I think 2019, the Teamwork Graph starts, which is trying to take all of those links and turn them into a graph. The connectivity, two things linked to this Figma thing, five things linked to this customer record — okay, cool, that means something, so we built this Graph. To be honest, it was a bit of a technology lark. We have a lot of these projects that are really cool and we’re like, “We’ll be able to use this somehow and it’s going to grown”, and now it’s a hundred billion objects and connections connecting all of the company’s knowledge. It becomes the organizational memory nowadays and context and all these things nobody knew in 2019 that’s what it was going to be, it just seemed we needed it for various process connections. That turns out to be because it’s got permissions and compliance and all of the enterprise stuff built in, which is incredibly difficult, the best resource to point AI at in various forms. You still have to be good at the AI parts to get the knowledge, the context for any area, so the Teamwork Graph is our data layer. It’s not only the best kind of enterprise search engine for your content from a 10 Blue Links kind of way of thinking. If you’re chatting through your content, you still need all your organizational knowledge. I actually obviously found your Article, I was like, “Hey, what has Ben Thompson written about us last year?”, and I asked Rovo in chat and it comes back to me with he wrote this, that and the other and pulls out some snippets. I’m like, “Tell me more, do you think we’ve hit that?”, I literally got a report written by Rovo on your report as to whether it had been accurate. “Go look at the last 10 years with deep research and web search and come back and tell me, was he right or wrong?”, and it gave me a really interesting analysis of whether you were right and wrong. It’s like most AI things, it’s like 90% correct, it’s pretty good. It solved a lot of the first problem and I would not have done that work otherwise. I would have read it quickly and so I wasn’t going to put an analyst on it internally to do this work, but I could send something to do work I never would’ve done. Who’s your competitor for this spot, for this Rovo position where you have all this context, you can actually search your company in a way that just wasn’t possible previously? MCB: Who are the competitors you say? Yeah, because everyone is claiming they’re in this spot, “We can be the central place that you go and we have visibility everywhere”, why is Atlassian the one that’s going to win that space? MCB: A few reasons why we will. I think we have a great chance to be a great player is maybe the easiest way to say it. I think everybody loves this absolute win position, we don’t believe in enterprise technology, you usually get these absolute wins, it’s not quite the same as in the consumer world. We have a lot of business processes and workflows, millions every day that run through us, those are human collaboration workflows, so they are cool. The auditing team hands off to the accounting team, hands off to the tax team, whatever it is, sales workflows, marketing workflows, and they span lots of our applications and many others. If you’re going to go and introduce agents, these autonomous AI-driven software programs, whatever you want to call an agent, you’re going to put them into existing processes to make those processes either more efficient, more accurate. When the human picks up a task, it’s got all the information they need because something’s gone out to find it, that is an incredibly powerful position, which is why we support our agents and everybody else’s. You can assign a Jira work item to a Cursor agent in terms of code, you can assign it to a Salesforce agent. If you have your agent technology choice, I don’t think you’re going to have one agent platform, I think you’re probably going to have multiples, there are going to be a handful of organizational knowledge graphs that are powerful enough to solve these problems across multiple tools, but we have access to all those tools. We already know the information to some level, and that becomes a very unique advantage. Do you see this as a way to expand even further how much of a company you cover? You started with developers, then you expand to adjacent teams, and you talk about it’s now just a fraction of your user base. Do you own entire companies or could you get there? It’s like, “Okay, we still have these teams over here that are not on Jira, but Rovo’s so good that we need to bring everyone in”? MCB: Look, again, it would be great. I think it is unrealistic, and we should say “Absolutely”, right? MCB: If [Salesforce CEO Marc] Benioff was here, he’d be like, “Absolutely, we’ll own the world”, we love him, that’s the way he is, I don’t think about it as owning a customer. Our mentality has always been — I always use the subway analogy versus we have some competitors, for example, that want to be the control tower, their whole thing is we’ll be the control tower, just give us control and we’ll go and control everybody else, we’ll move the planes around. I think in enterprise IT, that’s an unrealistic view. Every CIO has been sold this for decades, it doesn’t happen because the world changes too quickly. Our philosophy and our commitment to customers has always been we will be a great citizen on all sides, we will interact with all of the applications you need, the old ones and the new ones, and we will be a valuable point of exchange in your business workflows and processes, whether those are structured like in Jira, whether unstructured like in Loom or Talent or something else. The reason for that is you have lots of systems. We want to be a valuable station on your subway network, we don’t want to be at the end of one of the lines, we want to be one of the handful of hub stations that are about moving trains around, and that is the best way to get your knowledge moving in your organization, it’s the best way to deal with your processes. Therefore, we need to have amazing AI capabilities. We have a massive investment in R&D, we have thousands of people working on AI tooling at the moment, and we have a huge creation bent, which is one of the reasons I think — we’ve talked a bit about the data advantage we have, I think we have a huge design advantage, and I actually think design is one of the hardest parts of building great AI experiences because it’s real fundamental design for the first time. You had a great line, you did a podcast a couple of weeks ago that I’ll put a link to, but you mentioned basically, the customer should not need to understand the difference between deterministic and probabilistic in the context of design, that’s what you’re driving at here. MCB: They should not need to understand that, they should need to understand when outcomes, outputs may be wrong or may be creative. Again, you talk a lot about the fact that hallucination is the other side of creativity, right, you can’t have one without the other. Hallucinations are a miracle. We have computers making stuff up! MCB: Our job is to explain to a customer when that happens, so it’s like this might be something you want to do, and that requires a lot of design. We have a feature in Jira called Work Breakdown which is super popular, where I can take a Jira issue and say, “Make me a bunch of sub-issues, this task has to be broken into a set of steps”. I don’t believe in the magic button theory of AI, that I’ll just hit a button and it’ll do all the things, I believe deeply in the value from AI will come from human-AI collaboration in a loop. It’s me and the AI working back and forth. You talk about yourself and Daman quite a lot , and it’s you, Daman and ChatGPT working together, but it’s not like you ask one thing and it’s done. It’s an interaction, it’s a collaboration back and forth, and that’s going to happen everywhere. In Work Breakdown, what it does is it says, “Hey, based on these types of documents I’ve gone to find from your whole graph in Google Docs and Confluence, whatever, I think this piece breaks down into these, is that correct?”, and it goes, “No, actually, that one doesn’t make any difference, these two are really good, you forgot about this document”, “Cool, let me go do that for you again”, and come back and say, “Is it these?”, “That’s closer”, and then you’re like, “That’s good enough, it’s 90% of what I need”, and then I go add the two that I need myself. That is a huge productivity boost but it’s not magically correct, and it requires a lot of design to tell people, “These are not the answers, these are possible answers, help us refine them and get better at it so that you get the 90% upside and the 10% downside is managed”. Are all these people pursuing these full agents that act on their own, are they just totally misguided? MCB: No, because I think, well, agents will take — there’s a snake oil sales thing going on as there always is in any bubble, and the snake oil sales is not wrong, it’s just chronologically challenged. (laughing) That’s so good. MCB: Well, customers are struggling. When I talk to customers every day, they’re like, “Is everyone else using these things to just magically transform their business with this simple, it took them five minutes and it’s replaced entire armies of people?”, and I’m like, “No, nobody’s doing that”. What they’re actually doing is taking business processes that are really important to their business and saying, “Okay, can I make this step better? This is highly error-prone. It’s compliance in a large organization, how do I make this part of the process better?”, and we’re like, “Oh, we can totally do that”, and they will replace small bits of lots of processes so that in Ship of Theseus style, five years from now, the process will look radically different. Occasionally, they are replacing entire processes, but this is the 1% case, what they’re actually doing is they have whole machines that are running and they’re trying to fix this cog and fix that cog, and that’s super valuable for them. That’s not a downside, that’s really, really valuable. And often, it’s work they didn’t want to do, work that wasn’t getting done, it wasn’t done at a high quality, so we got to remember that, I say this quite a lot, people shouldn’t be afraid of AI taking their job, I fundamentally believe this, they should be afraid of someone who’s really good at AI taking their job. That’s actually what’s going to happen, is someone is going to come along, in a sales sense, they’re really good at using all these AI tools to give better customer outcomes or handle more customers at one time. Is this why you’re hiring so many young people? MCB: Yes, I guess so. Yes, they’re more AI-native, they come out understanding these tools and technologies. I find the biggest irony in universities is all these people who “cheat” their way through every assignment, I use cheat in quote marks, using ChatGPT to handle these assignments, and then they’re worried AI is going to take all these jobs. I’m like, “Wait, you literally took your own job of writing the assignment, but you’ve also trained yourself on how to use these tools to get the outcome required” — now one might argue the university degree should be different, but just like when Google came along and you could look up any fact, knowing facts became far less important than the ability to look it up. I still think AI, it doesn’t create anything, maybe slightly controversial, but I argue it synthesizes information, it’s really good at processing huge amounts of information, giving it back to you, changing its form, bringing it back. Humans are still the only source of fundamental knowledge creation. I point out one of the flaws in the one person billion dollar company argument, and this will happen but it’ll be an anomaly. That company doesn’t get created without that one person, so there’s not AI creating companies magically. It’s like can a company eternally buy back its stock? No, because at some point, someone is going to own the final share? MCB: That’s right and I think this is missed, right? This is where we say it’s about unlocking creativity and what we do for our customers is put Rovo and these amazing data capabilities that we have alongside all the enterprise compliance and data residency, and there’s a massive amount of making this work in the enterprise with trust and probity and security. It’s very difficult. And great design to say, “What do you hire us to do? How do you get these technology and business teams to work together? What workflows do you have in your projects and your service teams, and how can we make those workflows better with more data and make your teams more informed?” That will end up with us having more share of employees in a business that use our stuff every day. Awesome. You made two big acquisitions recently, the DX acquisition , I think, makes a ton of sense to me measuring engineering productivity, particularly in the area of AI. What actual ROI are we getting on this? MCB: And how much money am I spending? Because I’m spending suddenly a lot of money, right? This is not cheap at all, I have huge bills. Internally, we use Rovo Dev , we use Claude Code, we use GitHub Copilot, we use Cursor, we have them available to all. We have a huge R&D — again, I think we’re still number one on the NASDAQ for R&D spending as proportion of revenue. You can take that as a good thing in the AI era or a bad thing, everyone gets to choose their own view on that, but we’ve always been incredibly high on R&D spending since day one. The bills that we pay though are very high, so DX is simply saying, “Okay, cool, how do I measure what I’m getting for that? Should I pay twice as much money because these bills are worthwhile, or is there a lot of it that’s actually just it’s really fun and it’s not actually leading to productivity gains?”. This is going to be a hard problem because there’s a lot of money on the line at the moment that people are paying for these tools, which is not without value, but measuring exactly what the value is is really, really hard, and that team’s done a phenomenal job. And we now have an Atlassian office in Salt Lake City, Utah, where I already spend a lot of time. Totally by coincidence, but it’s really nice. So that purchase, love it, makes a ton of sense. In perfect alignment with you. How does The Browser Company fit in? MCB: A lot of ways. So I have believed for a long time that browsers are broken. We’ve built browsers for an era of software that we don’t live in today. And I don’t, in my browser, have a bunch of tabs that represent webpages, I don’t have that. I have a bunch of tasks, I have a bunch of applications, I have a bunch of documents, and the browser was fundamentally never built to do that. That’s what Arc, first product from The Browser Company — if you don’t use Arc every single day, you should be, it’ll increase your productivity instantly because it’s built for knowledge workers and the way that they have to actually work every day and how they manage all of these tabs and tasks and flows versus serving the New York Times or whatever. That is a browser built for knowledge workers, and there’s a lot more we can do in that era as software changes. Secondly, obviously AI has come along, and we now have chats and applications as a extra part of the browser experience, so I think we can change how enterprises use browsers, security being a big issue. I think AI in the browser is a really important thing, but I suspect it’s not in the basic way of just combining Chrome and ChatGPT, that’s not how it’s going to play out. I suspect it requires a massive amount of design, which The Browser Company is phenomenal at, and it requires changing how people use their day-to-day applications. From our point of view, and I’ve been an Arc fan since day one, [The Browser Company CEO] Josh [Miller] and I have known each other a long time, there’s a knowledge worker angle and there’s obviously a business angle to it in a huge way that our customers are knowledge workers. We can change the way they do their work in a meaningful way of productivity, that is exactly what we have been trying to do in a lot of different ways. The browser itself, being chromium-based, Edge being chromium-based, Chrome being chromium-based, the rendering of webpages is not the problem, it is the fundamental user experience of, “How do I take all of my SaaS applications, my agents, my chats, my tabs, my knowledge, and put it all together in ways that make my day quicker?” — that is what we are trying to do fundamentally at the start. The context that we have is incredibly important for that. And the browser has, if you think about it, my personal memory. We used to call it the browser history. Great, it shows what I’ve seen, it does not have my organizational memory, which we have a great example of in the Teamwork Graph. So if I can put these things together, I can make a much more productive browsing experience for customers fundamentally in that world. I think we have an amazing shot of doing that and of changing how knowledge workers use SaaS. We’re not trying to make a browser, as I’ve said, for my kids, we’re not trying to make a browser for my parents, we’re not trying to make a browser for shopping or for anything else. We’re trying to make a browser for people who spend all day living in Salesforce and Jira and Google Docs and Confluence and Figma and GitHub, and that is their life. The laptop warrior that sits in that experience, I believe we can use AI and design to make that a far better experience and build an amazing product. They’re well on the way to doing that, we can supercharge doing it. You look skeptical. No, I’m looking at the clock, I skipped over a huge section. Your whole shift to the cloud, all those sorts of things. However, there is one thing I wanted to get to: you are wearing an Atlassian Williams Racing hat , I am a big F1 fan, I was very excited about you doing this . How did that come about? How was the first year? Was this another hunch this is going to work out? I mean, Williams is looking like a pretty good bet. MCB: Yes, our world’s largest sports bet. Look, how did it come about? So how do I make a short answer? F1 is changing, I think, in a massive way. I know now being incredibly deep in the business of it, the fundamental change is that hardware is becoming less important and software is becoming more important, this is a trend that we are used to. JV, James Vowles , the Team Principal, was the first person that approached us a long while ago now to help them, and for a teeny, teeny sticker in the corner, to help them get more productive as a team. What people don’t realize about F1 is these are large organizations, right? There’s 1100 people that work for Atlassian Williams Racing. And Williams was really pared down and skinny, he was brought back in with new owners to actually rebuild the entire thing? MCB: Yes, they were in deep trouble. But in rebuilding it, he is a software engineer, software developer by trade, by history kind of thing. He’s a technically-minded person. He downloaded Jira himself in 2004 to install it, so he knows us quite well. So we were brought on for our ability to help them with their teamwork and their collaboration, they really needed a technical upgrade to a whole lot of their systems. Turns out they need us in almost every part of their business because the service workflow’s important. We’re now in the garage, we’re using tons of AI to try to make them better, so there’s a lot of things we can do to build to hopefully help them win, and it’s a mission you can fall in love with. Here is one of the most storied brands in Formula 1 that’s fallen on tough times, every sportsperson loves a recovery story. And I was sold early on the recovery story, I’m like, “Fuck it, let’s go help, let’s make this happen. Let’s get back to being a championship team”. So we fell in love with the mission, and JV is super compelling, he’s got a one-decade goal, and they’re very goal-driven, and we love that, but they needed a lot of help, so that’s what they asked us for help with is initially. The more we looked at it, the more we learned about Formula 1, yes, it’s becoming a software-driven sport. So as an example, Atlassian Williams, I believe have twice as many software developers as the next team on the grid. Because it’s cost-capped, you got to choose, “Do I hire a software developer or an aerodynamicist?” — it’s a very clear cost cap, you’re choosing where to put your resources. As virtualization and everything get better, it’s less, “How well can I draw a curve?” and, “How much can I help 1100 people work together, and how can we build great software”, which really is the core of the car, right? So that then comes to us, tiny sticker, probably a founder-ish moment where I’m like, “How much is the sticker on the top?”, and they didn’t have a sticker on the top and I’m like, well, “What would that get us?” So we ran the numbers on that and the reason is twofold. You talked about our GTM, our go-to-market transformation, we have an ability to build various things. Firstly, branding is obviously massive, top three teams get 10 times the branding as the bottom three teams. So if you’re going to make a sports bet, you pay for a long period of time with the bottom three team, you help make them a top three team, and your sport bet pays out really well just on a sheer TV time and etc — the number of staff, parents, and other things, have said to staff members, “Hey, that company you work for, it’s really great, I saw them on the TV on the weekend”, and the staff member will say, “Dude, I’ve worked there for 12 years, why do you suddenly know about it?”, “Oh, I saw them driving. Carlos [Sainz Jr.] is great”, or something. And he is! So obviously, there’s a huge marketing and branding angle that’s about their position being better. The really interesting part of what we’re doing there is we have customers all around the world, we have customers in 200-odd countries, and we can’t go and visit all of our biggest customers in a meaningful way. We certainly can’t take them to some of our best and most exciting customers, right? There are electric car companies that use our stuff that we’d love to take many customers to a factory, or rockets, or whoever, I can’t take many customers into some of your favorite chip companies and say, “Look how they use our stuff”, I can maybe get one or two customers a year into that customer and show them how they use our things. With Formula 1, what we’re building is a mobile EBC, so an executive briefing center. Formula 1 goes around the world. It goes to Melbourne, it goes to Singapore, it goes to Japan, it goes to England, it goes to various parts of Northern Europe, it goes to various parts of America and you’re like, “Hey, where are our customers?” — roughly distributed like that. It comes to town, we can invite a whole lot of customers into a great experience, we can tell them a lot about Atlassian software, we can also invite them into one of our best customers. They can sit in the garage, and I can tell them how our service collection is helping power the assets, that when that wing’s broken, it gets known here, and they start making a new one back in the factory in Oxford, and this one gets shipped around the world and another one will get moved. And, “Here, I can show you the asset management and the service that goes along with it, I can show you how the garage is getting more efficient because of us, I can show you how we’re helping them win races”. We don’t drive cars, we help them be more productive as a team and I can do that in an environment of it’s an exciting environment. They can drink a great latte or a champagne or whatever they want, and I can explain to them how we are transforming this business in a meaningful way with our tools no matter which way they want to look at it, which is the most powerful customer story that you can go and tell a couple-hundred customers a year in their city. We come to their city, right? I was in Montreal, I took a whole bunch of Canadian customers over the three days, they were like, “This changes my view of Atlassian”, and I’m like, “That’s exactly our goal”, that is at the enterprise end of enterprise sales though, right? But that’s the ironic thing, it’s as far away from where you started as you could be. MCB: Well, they didn’t get there. I met two Canadian banks we had in Montreal as an example, both of whom had been customers for over 20 years, they started spending $800 bucks or maybe $4800 as we moved our pricing to around five grand — now they spend a million, two million dollars a year, and they could be spending ten. We have the ability to give the massive business value across a far larger swath of their business. And I can say, “What do you use from our system of work today? What could you use? Let me show you how Williams uses that piece of the system of work”, which is just a very visceral and exciting customer example to show them how they’re winning. And it helps, again, culturally, super aligned. They’re an awesome group of people trying really hard to win in the most ridiculously competitive sport and the highs are highs, the lows are low. Any sporting fan, you’re well familiar with various different sports that we have in common, but this is technology built by a large business team that has to win a sport. That doesn’t happen anywhere else in the sporting world, I would claim. Giannis [Antetokounmpo] doesn’t make his own shoes and have a team of people making better shoes and a better basketball so he can win, that doesn’t happen in other sports. It’s all about the people on the floor in an NBA game as to who wins, and that’s great, don’t get me wrong, I love basketball. The work in Formula 1 is done by 1000 people back in Oxford. It’s a Constructor Championship . MCB: The constructor championship I do think should be more important, especially given the current exact week we’re in, which is an amazing week for Atlassian Williams Racing, second podium . You talk about that bet, I told JV at the start of the year, I thought that he’s like, “What do you think our five-year future is?”, and I said, “Look, I think, number one, we’ll get one podium this year, 2025; 2026, we’ll win a race; and by 2030, we will have won a championship, that is my OKRs [Objectives and Key Results]”, and he said, “Oh, wow, okay, yeah I think so”. It lines up, I know the team OKRs and other things. And we won two podiums this year, so I was wrong, and I think we have a great chance for 2026, and we are working hard to make the team better and the single-best customer example we have of every piece of software that we sell. Mike, I’d love to talk again. It was great talking to you again. And, hey, good luck. And I’m a Williams fan, so I’ll be cheering for you this weekend. MCB: Oh, yeah. Well, I’m not sure this weekend, but 2026, 2027- Okay. I’m kind of kissing up, I am dying for Max [Verstappen] to win is the honest truth. I need the McLarens to run into each other . But other than that, Williams is my second love. MCB: Do you think McLaren will issue team orders to switch them if Oscar is in second and Lando’s in fourth? Yes. And I don’t know what’s going to happen if that happens, and this will be fascinating. MCB: We will have to see. It’s going to be a huge week. But that’s what makes the sport exciting, right? The whole thing is amazing. Talk to you later. MCB: All right. Thanks, man. This Daily Update Interview is also available as a podcast. To receive it in your podcast player, visit Stratechery . The Daily Update is intended for a single recipient, but occasional forwarding is totally fine! If you would like to order multiple subscriptions for your team with a group discount (minimum 5), please contact me directly. Thanks for being a supporter, and have a great day!
Have you noticed that every issue and idea tracking tool you loved slowly morphed into boring, sluggish, corporate bloatware? Trello put on 40 pounds of cruft. Jira started charging by the migraine. Asana tried to become everything to everyone. GitHub Issues slipped into a steady state of decline. The whole category is a 20 car pileup of complexity. Time to route around that mess. Today we’re introducing Fizzy. Kanban as it should be, not as it has been. Fizzy is a fresh take on cards and columns, with a few twists, human-nature inspired defaults, and a vibrant interface that’s the opposite of the bland and boring software the industry has been flinging at you for years. Kanban has been around since the 1940s, and Trello brought it into the mainstream in 2011. Since then, some version of column-based kanban-style organization has found its way into any collaboration tool worth its salt. But most have over salted the dish. What was simple is now complicated. What was clear is now cluttered. What just worked now takes work. Fizzy presses reset, reconsiders what really matters, and presents a refreshing way to kanban that just feels right. It’s friendly, colorful, straightforward, and fast as hell. We still use Basecamp for our big, intensive projects, but lately we’ve been reaching for Fizzy to run the smaller ones. It’s perfect for tracking bugs, issues, and ideas, and it shines for lighter, self-contained workflows like podcasts or video production. We didn’t expect it, but Fizzy’s so good it might even cannibalize Basecamp on the lighter side of project management. We’d be thrilled. How much is it? It’s not much for so much. Everyone gets 1000 cards for free. Beyond that, we’ll host your account for just $20/month for unlimited cards and unlimited users. One price for all and everything. No tiers, no “contact us.” No pricing chart at all — just a price tag, like on a pair of jeans. And here’s a surprise... Fizzy is open source! If you’d prefer not to pay us, or you want to customize Fizzy for your own use, you can run it yourself for free forever. Have a great idea? Submit a PR to contribute to the code base and improve the product for everyone. It’s the best of all worlds. No excuses. Every idea comes back around. It’s time for take two on kanban. Fizzy’s our hat in the ring. Let’s make this platform insanely great, together. Come on in! Visit fizzy.do to check it out and sign up for free! -Jason
Quick announcement: I'll be visiting Japan in April, 2026 for about a month and will be on Honshu for most of the trip. Please email me recommendations. If you live nearby, let's have coffee? I've always been fascinated by old, multi-generational Japanese businesses. My leisure-watching on YouTube is usually a long video of a Japanese craftsman—sometimes a 10th or 11th generation—making iron tea kettles, or soy sauce, or pottery, or furniture. Their dedication to craft—and acknowledgment that perfection is unattainable—resonates with me deeply. Improving in their craft is an almost spiritual endeavour, and it inspires me to engage in my crafts with a similar passion and focus. Slow, consistent investment over many years is how beautiful things are made, learnt, or grown. As a society we forget this truth—especially with the rise of social media and the proliferation of instant gratification. Good things take time. Dedication to craft in this manner comes with incredible longevity (survivorship bias plays a role, but the density of long-lived businesses in Japan is an outlier). So many of these small businesses have been around for hundreds, and sometimes over a thousand years, passed from generation to generation. Modern companies have a hard time retaining employees for 2 years, let alone a lifetime. This longevity stems from a counter-intuitive idea of growing slowly (or not at all) and choosing to stay small. In most modern economies if you were to start a bakery, the goal would be to set it up, hire and train a bunch of staff, and expand operations to a second location. Potentially, if you play your cards right, you could create a national (or international) chain or franchise. Corporatise the shit out of it, go public or sell, make bank. While this is a potential path to becoming filthy rich, the odds of achieving this become vanishingly small. The organisation becomes brittle due to thinly-spread resources and care, hiring becomes risky, and leverage, whether in the form of loans or investors, imposes unwanted directionality. There's a well known parable of the fisherman and the businessman that goes something like this: A businessman meets a fisherman who is selling fish at his stall one morning. The businessman enquires of the fisherman what he does after he finishes selling his fish for the day. The fisherman responds that he spends time with his friends and family, cooks good food, and watches the sunset with his wife. Then in the morning he wakes up early, takes his boat out on the ocean, and catches some fish. The businessman, shocked that the fisherman was wasting so much time encourages him fish for longer in the morning, increasing his yield and maximising the utility of his boat. Then he should sell those extra fish in the afternoon and save up until he has enough money to buy a second fishing boat and potentially employ some other fishermen. Focus on the selling side of the business, set up a permanent store, and possibly, if he does everything correctly, get a loan to expand the operation even further. In 10 to 20 years he could own an entire fishing fleet, make a lot of money, and finally retire. The fisherman then asks the businessman what he would do with his days once retired, to which the businessman responds: "Well, you could spend more time with your friends and family, cook good food, watch the sunset with your wife, and wake up early in the morning and go fishing, if you want." I love this parable, even if it is a bit of an oversimplification. There is something to be said about affording comforts and financial stability that a fisherman may not have access to. But I think it illustrates the point that when it comes to running a business, bigger is not always better. This is especially true for consultancies or agencies which suffer from bad horizontal scaling economics. The trick is figuring out what is "enough". At what point are we chasing status instead of contentment? A smaller, slower growing company is less risky, less fragile, less stressful, and still a rewarding endeavour. This is how I run Bear. The project covers its own expenses and compensates me enough to have a decent quality of life. It grows slowly and sustainably. It isn't leveraged and I control its direction and fate. The most important factor, however, is that I don't need it to be something grander. It affords me a life that I love, and provides me with a craft to practise.
OpenAI is declaring code red and doubling down on ChatGPT, highlighting the company's bear case. Then, AWS makes it easier to run AI workloads on other clouds.
Listen to this post : A common explanation as to why Star Wars was such a hit, and continues to resonate nearly half a century on from its release, is that it is a nearly perfect representation of the hero’s journey. You have Luke, bored on Tatooine, called to adventure by a mysterious message borne by R2-D2, that he initially refuses; a mentor in Obi-Wan Kenobi leads him across the threshold of leaving Tatooine and facing tests while finding new enemies and allies. He enters the cave — the Death Star — escapes after the ordeal of Obi-Wan’s death, and carries the battle station’s plans to the rebels while preparing for the road back to the Death Star. He trusts the force in his final test and returns transformed. And, when you zoom out to the entire original trilogy, it’s simply an expanded version of the story: this time, however, the ordeal is the entire second movie: the Empire Strikes Back. The heroes of the AI story over the last three years have been two companies: OpenAI and Nvidia. The first is a startup called, with the release of ChatGPT, to be the next great consumer tech company ; the other was best known as a gaming chip company characterized by boom-and-bust cycles driven by their visionary and endlessly optimistic founder, transformed into the most essential infrastructure provider for the AI revolution. Over the last two weeks, however, both have entered the cave and are facing their greatest ordeal: the Google empire is very much striking back. The first Google blow was Gemini 3, which scored better than OpenAI’s state of the art model on a host of benchmarks (even if actual real-world usage was a bit more uneven). Gemini 3’s biggest advantage is its sheer size and the vast amount of compute that went into creating it; this is notable because OpenAI has had difficulty creating the next generation of models beyond the GPT-4 level of size and complexity. What has carried the company is a genuine breakthrough in reasoning that produces better results in many cases, but at the cost of time and money. Gemini 3’s success seemed like good news for Nvidia, who I listed as a winner from the release : This is maybe the most interesting one. Nvidia, which reports earnings later today, is on one hand a loser, because the best model in the world was not trained on their chips, proving once and for all that it is possible to be competitive without paying Nvidia’s premiums. On the other hand, there are two reasons for Nvidia optimism. The first is that everyone needs to respond to Gemini, and they need to respond now, not at some future date when their chips are good enough. Google started its work on TPUs a decade ago; everyone else is better off sticking with Nvidia, at least if they want to catch up. Secondly, and relatedly, Gemini re-affirms that the most important factor in catching up — or moving ahead — is more compute. This analysis, however, missed one important point: what if Google sold its TPUs as an alternative to Nvidia? That’s exactly what the search giant is doing, first with a deal with Anthropic, then a rumored deal with Meta, and third with the second wave of neoclouds, many of which started as crypto miners and are leveraging their access to power to move into AI. Suddenly it is Nvidia that is in the crosshairs, with fresh questions about their long term growth, particularly at their sky-high margins, if there were in fact a legitimate competitor to their chips . This does, needless to say, raise the pressure on OpenAI’s next pre-training, run on Nvidia’s Blackwell chips: the base model still matters, and OpenAI needs a better one, and Nvidia needs evidence one can be created on their chips. What is interesting to consider is which company is more at risk from Google, and why? On one hand Nvidia is making tons of money, and if Blackwell is good, Vera Rubin promises to be even better; moreover, while Meta might be a natural Google partner , the other hyperscalers are not. OpenAI, meanwhile, is losing more money than ever, and is spread thinner than ever, even as the startup agrees to buy ever more compute with revenue that doesn’t yet exist. And yet, despite all that — and while still being quite bullish on Nvidia — I still like OpenAI’s chances more. Indeed, if anything my biggest concern is that I seem to like OpenAI’s chances better than OpenAI itself. If you go back a year or two, you might make the case that Nvidia had three moats relative to TPUs: superior performance, significantly more flexibility due to GPUs being more general purpose than TPUs, and CUDA and the associated developer ecosystem surrounding it. OpenAI, meanwhile, had the best model, extensive usage of their API, and the massive number of consumers using ChatGPT. The question, then, is what happens if the first differentiator for each company goes away? That, in a nutshell, is the question that has been raised over the last two weeks: does Nvidia preserve its advantages if TPUs are as good as GPUs, and is OpenAI viable in the long run if they don’t have the unquestioned best model? Nvidia’s flexibility advantage is a real thing; it’s not an accident that the fungibility of GPUs across workloads was focused on as a justification for increased capital expenditures by both Microsoft and Meta. TPUs are more specialized at the hardware level, and more difficult to program for at the software level; to that end, to the extent that customers care about flexibility, then Nvidia remains the obvious choice. CUDA, meanwhile, has long been a critical source of Nvidia lock-in, both because of the low level access it gives developers, and also because there is a developer network effect: you’re just more likely to be able to hire low level engineers if your stack is on Nvidia. The challenge for Nvidia, however, is that the “big company” effect could play out with CUDA in the opposite way to the flexibility argument. While big companies like the hyperscalers have the diversity of workloads to benefit from the flexibility of GPUs, they also have the wherewithal to build an alternative software stack. That they did not do so for a long time is a function of it simply not being worth the time and trouble; when capital expenditure plans reach the hundreds of billions of dollars, however what is “worth” the time and trouble changes. A useful analogy here is the rise of AMD in the datacenter. That rise has not occurred in on-premises installations or the government, which is still dominated by Intel; rather, large hyperscalers found it worth their time and effort to rewrite extremely low level software to be truly agnostic between AMD and Intel, allowing the former’s lead in performance to win the battle. In this case, the challenge Nvidia faces is that its market is a relatively small number of highly concentrated customers, with the resources — mostly as yet unutilized — to break down the CUDA wall, as they already did in terms of Intel’s differentiation. It’s clear that Nvidia has been concerned about this for a long time; this is from Nvidia Waves and Moats , written at the absolute top of the Nvidia hype cycle after the 2024 introduction of Blackwell: This takes this Article full circle: in the before-times, i.e. before the release of ChatGPT, Nvidia was building quite the (free) software moat around its GPUs; the challenge is that it wasn’t entirely clear who was going to use all of that software. Today, meanwhile, the use cases for those GPUs is very clear, and those use cases are happening at a much higher level than CUDA frameworks (i.e. on top of models); that, combined with the massive incentives towards finding cheaper alternatives to Nvidia, means both the pressure to and the possibility of escaping CUDA is higher than it has ever been (even if it is still distant for lower level work, particularly when it comes to training). Nvidia has already started responding: I think that one way to understand DGX Cloud is that it is Nvidia’s attempt to capture the same market that is still buying Intel server chips in a world where AMD chips are better (because they already standardized on them); NIM’s are another attempt to build lock-in. In the meantime, though, it remains noteworthy that Nvidia appears to not be taking as much margin with Blackwell as many may have expected; the question as to whether they will have to give back more in future generations will depend on not just their chips’ performance, but also on re-digging a software moat increasingly threatened by the very wave that made GTC such a spectacle. Blackwell margins are doing just fine, I should note, as they should be in a world where everyone is starved for compute. Indeed, that may make this entire debate somewhat pointless: implicit in the assumption that TPUs might take share from GPUs is that for one to win the other must lose; the real decision maker may be TSMC, which makes both chips, and is positioned to be the real brake on the AI bubble . ChatGPT, in contrast to Nvidia, sells into two much larger markets. The first is developers using their API, and — according to OpenAI, anyways — this market is much stickier and reticent to change. Which makes sense: developers using a particular model’s API are seeking to make a good product, and while everyone talks about the importance of avoiding lock-in, most companies are going to see more gains from building on and expanding from what they already know, and for a lot of companies that is OpenAI. Winning business one app by one will be a lot harder for Google than simply making a spreadsheet presentation to the top of a company about upfront costs and total cost of ownership. Still, API costs will matter, and here Google almost certainly has a structural advantage. The biggest market of all, however, is consumer, Google’s bread-and-butter. What makes Google so dominant in search, impervious to both competition and regulation, is that billions of consumers choose to use Google every day — multiple times a day, in fact. Yes, Google helps this process along with its payments to its friends , but that’s downstream from its control of demand, not the driver . What is paradoxical to many about this reality is that the seeming fragility of Google’s position — competition really is a click away! — is in fact its source of strength. From United States v. Google : Increased digitization leads to increased centralization (the opposite of what many originally assumed about the Internet). It also provides a lot of consumer benefit — again, Aggregators win by building ever better products for consumers — which is why Aggregators are broadly popular in a way that traditional monopolists are not. Unfortunately, too many antitrust-focused critiques of tech have missed this essential difference… There is certainly an argument to be made that Google, not only in Shopping but also in verticals like local search, is choking off the websites on which Search relies by increasingly offering its own results. At the same time, there is absolutely nothing stopping customers from visiting those websites directly, or downloading their apps, bypassing Google completely. That consumers choose not to is not because Google is somehow restricting them — that is impossible! — but because they don’t want to. Is it really the purview of regulators to correct consumer choices willingly made? Not only is that answer “no” for philosophical reasons, it should be “no” for pragmatic reasons, as the ongoing Google Shopping saga in Europe demonstrates. As I noted last December , the European Commission keeps changing its mind about remedies in that case, not because Google is being impertinent, but because seeking to undo an Aggregator by changing consumer preferences is like pushing on a string. The CEO of a hyperscaler can issue a decree to work around CUDA; an app developer can decide that Google’s cost structure is worth the pain of changing the model undergirding their app; changing the habits of 800 million+ people who use ChatGPT every week, however, is a battle that can only be fought individual by individual. This is ChatGPT’s true difference from Nvidia in their fight against Google. This is, I think, a broader point: the naive approach to moats focuses on the cost of switching; in fact, however, the more important correlation to the strength of a moat is the number of unique purchasers/users. This is certainly one of the simpler charts I’ve made, but it’s not the first in the moat genre; in 2018’s The Moat Map I argued that you could map large tech companies across two spectrums. First, the degree of supplier differentiation: Second, the extent to which a company’s network effects were externalized: Putting this together gave you the Moat Map: What you see in the upper right are platforms; the lower left are Aggregators. Platforms like the App Store enable differentiated suppliers, which lets them profitably take a cut of purchases driven by those differentiated suppliers; Aggregators, meanwhile, have totally commoditized their suppliers, but have done so in the service of maximizing attention, which they can monetize through advertising. It’s the bottom left that I’m describing with the simplistic graph above: the way to commoditize suppliers and internalize network effects is by having a huge number of unique users. And, by extension, the best way to monetize that user base — and to achieve a massive user base in the first place — is through advertising. It’s so obvious the bottom left is where ChatGPT sits. At one point it didn’t seem possible to commoditize content more than Google or Facebook did, but that’s exactly what LLMs do: the answers are a statistical synthesis of all of the knowledge the model makers can get their hands on, and are completely unique to every individual; at the same time, every individual user’s usage should, at least in theory, make the model better over time. It follows, then, that ChatGPT should obviously have an advertising model. This isn’t just a function of needing to make money: advertising would make ChatGPT a better product. It would have more users using it more, providing more feedback; capturing purchase signals — not from affiliate links, but from personalized ads — would create a richer understanding of individual users, enabling better responses. And, as an added bonus — and one that is very pertinent to this Article — it would dramatically deepen OpenAI’s moat. It’s not out of the question that Google can win the fight for consumer attention. The company has a clear lead in image and video generation, which is one reason why I wrote about The YouTube Tip of the Google Spear : In short, while everyone immediately saw how AI could be disruptive to Search , AI is very much a sustaining innovation for YouTube: it increases the amount of compelling content in absolute terms, and it does so with better margins, at least in the long run. Here’s the million billion trillion dollar question: what is going to matter more in the long run, text or video? Sure, Google would like to dominate everything, but if it had to choose, is it better to dominate video or dominate text? The history of social networking that I documented above suggests that video is, in the long run, much more compelling to many more people. To put it another way, the things that people in tech and media are interested in has not historically been aligned with what actually makes for the largest service or makes the most money: people like me, or those reading me, care about text and ideas; the services that matter specialize in videos and entertainment, and to the extent that AI matters for the latter YouTube is primed to be the biggest winner, even as the same people who couldn’t understand why Twitter didn’t measure up to Facebook go ga-ga over text generation and coding capabilities. Google is also obviously capable of monetizing users, even if they haven’t turned on ads in Gemini yet (although they have in AI Overviews). It’s also worth pointing out, as Eric Seufert did in a recent Stratechery Interview , that Google started monetizing Search less than two years after its public launch; it is search revenue, far more than venture capital money, that has undergirded all of Google’s innovation over the years, and is what makes them a behemoth today. In that light OpenAI’s refusal to launch and iterate an ads product for ChatGPT — now three years old — is a dereliction of business duty, particularly as the company signs deals for over a trillion dollars of compute. And, on the flip side, it means that Google has the resources to take on ChatGPT’s consumer lead with a World War I style war of attrition; OpenAI’s lead should be unassailable, but the company’s insistence on monetizing solely via subscriptions, with a degraded user experience for most users and price elasticity challenges in terms of revenue maximization, is very much opening up the door to a company that actually cares about making money. To put it another way, the long-term threat to Nvidia from TPUs is margin dilution; the challenge of physical products is you do have to actually charge the people who buy them, which invites potentially unfavorable comparisons to cheaper alternatives, particularly as buyers get bigger and more price sensitive. The reason to be more optimistic about OpenAI is that an advertising model flips this on its head: because users don’t pay, there is no ceiling on how much you can make from them, which, by extension, means that the bigger you get the better your margins have the potential to be, and thus the total size of your investments. Again, however, the problem is that the advertising model doesn’t yet exist. I started this Article recounting the hero’s journey, in part to make the easy leap to “The Empire Strikes Back”; however, there was a personal angle as well. The hero of this site has been Aggregation Theory and the belief that controlling demand trumps everything else; there Google was my ultimate protagonist . Moreover, I do believe in the innovation and velocity that comes from a founder-led company like Nvidia, and I do still worry about Google’s bureaucracy and disruption potential making the company less nimble and aggressive than OpenAI. More than anything, though, I believe in the market power and defensibility of 800 million users, which is why I think ChatGPT still has a meaningful moat. At the same time, I understand why the market is freaking out about Google: their structural advantages in everything from monetization to data to infrastructure to R&D is so substantial that you understand why OpenAI’s founding was motivated by the fear of Google winning AI. It’s very easy to imagine an outcome where Google’s inputs simply matter more than anything else, which is to say one of my most important theories is being put to the ultimate test (which, perhaps, is why I’m so frustrated at OpenAI’s avoidance of advertising). Google is now my antagonist! Google has already done this once: Search was the ultimate example of a company winning an open market with nothing more than a better product. Aggregators win new markets by being better; the open question now is whether one that has already reached scale can be dethroned by the overwhelming application of resources, especially when its inherent advantages are diminished by refusing to adopt an Aggregator’s optimal business model. I’m nervous — and excited — to see how far Aggregation Theory really goes.
There are three editions of my book in circulation: two English editions of Mastering Vim (the second being a complete rewrite), and a Japanese translation by the amazing Masafumi Okura. And I don’t really mind if my book gets pirated. Yeah, piracy isn’t legal, yada-yada. But if $30 is too much right now and your library doesn’t have a copy to spare - I won’t blame you for torrenting it. Here’s my full permission, I hope you enjoy the result of my sweat, tears, and deadline anxiety. Amazon is convenient, but you don’t own your Kindle books. They can be deleted or changed at any moment. Amazon has literally changes book covers after purchase, especially when books get a movie release. Ugh. Meet DRM (Digital Rights Management) - the technology that ensures you’re renting, not buying. Packt, publisher of Mastering Vim, does offer DRM-free PDFs. But there’s no good PDF syndication ecosystem. No convenient library management. No sync across devices. You’re not really missing out. Growing up in Russia, I pirated video games. Not out of principle, but pragmatism. Fan translations arrived six months before official ones. They were better too - localizers who actually played the games versus outsourced rush jobs. This was different kind of piracy, too - a guy on a corner selling pirated CDs at a market-appropriate rate. I stopped in 2011 - already after I moved to the United States. Not because of some moral awakening, but because I learned about Steam. Cloud saves, achievements, automatic updates. The service became worth paying for. Music followed the same path - Spotify and YouTube Music (which I like because our family pays for YouTube Premium) made piracy pointless. Video streaming went backwards. Netflix was the Steam moment for TV - everything in one place, reasonably priced. I loved our Netflix subscription, it felt oh-so-magical. Now? Eight subscriptions to watch your shows, content vanishing mid-season, regional restrictions. I watch a handful of shows or movies each month, and I once calculated how much I would have to take in subscription costs if I didn’t strategically sign up and cancel for periods when I want to watch my favorite shows. Over a $1,000 a year. Screw that. I, of course, don’t pirate, not do I condone piracy. But I do use Jellyfin to organize my legally owned media library. One interface, no disappearing content, works offline. It’s simply a better experience than juggling Disney+, Netflix, HBO Max, Paramount+, Apple TV+, and whatever new service launched this week. Piracy isn’t about price - it’s about service. Steam proved gamers will pay. Spotify proved music fans will pay. But fragment the market, add restrictions, remove content randomly, make legal options worse than illegal ones? Don’t be surprised when people choose the better experience.
Black Friday is usually when ecommerce sets new records. This has certainly been true for Shopify through most of its existence. So much so that the company spends months in advance preparing for The Big Day(s). You'd think after more than twenty years, though, that things would have leveled out. But you'd be wrong. This year, merchants sold an astounding $6.2 billion worth of wares through Shopify on Black Friday. That's up 25% from last year, when the record was ~$5 billion. Just crazy high growth on a crazy big base. The law of big numbers clearly hasn't found a way to apply itself here yet! That volume of orders means the Shopify monolith gets put through its paces. The backend API peaked at 31 million requests per minute. The databases carried 53 million reads and 2 million writes per second. Bonkers. It's this kind of frontier load and criticality that makes Shopify the ideal patron saint of the Rails framework and the Ruby programming language. Rarely do the stars align to shine so brightly that a single company is stewarded by a still-active programmer with a stellar pedigree of core contributions, saddled with such unceasing success, faced with a constant barrage of novel technical challenges, and willing to contribute everything they learn and build back into the open-source base pillars. But that's Shopify. Ultimately, this is all downstream from being a founder-led business. Tobi Lütke not only served on the Rails core team in the early days, but continues to steer the Shopify ship with a programmer's eye for detail and exploration. The latest release of Omarchy even features his new Try tool. How many CEOs of companies worth two hundred billion dollars still program like that? Despite all this, there's occasionally still some fringe consternation in the Ruby world about Shopify's dominance. In Rails, Shopify employs almost half the core contributors. In Ruby, they have several people on the core team too. Seeing this as anything but a blessing is silly, though. We wouldn't have such battle-tested releases of Rails without Shopify running production on the framework's edge. We wouldn't have gotten YJIT without the years of effort they sunk into improving Ruby's core performance. And we wouldn't have seen the recent production-proving of Ractors without them either. Any programming community should be so lucky as to have a Shopify! Now I'm obviously biased here. Not only have I been friends with Tobi for over twenty years, but I also serve on the board of directors for the company. I'm both socially and economically incentivized to cheer for this extraordinary company. But that doesn't mean it isn't all true too! Shopify is indeed the patron saint of Ruby on Rails. Its infrastructure team is the backbone of our ecosystem, and its continued success the best case study of how far you can take this framework and language. They deserve a gawd damn parade for all they do. So on this Cyber Monday, I say cheers to Tobi, cheers to the thousands of Shopifolk. You're killing it for merchants, shoppers, and all of us working with Ruby on Rails. Bravo.
I sometimes wonder if we have too uncritically accepted the marketing narrative of social media companies about how connectivity is always good and preferable, and that they as the mediators always need to be the ones facilitating it in their own way. I’ll have to narrow it down: Of course having friends, family, a support network is good - even needed - and work connections get you further professionally, both offline and online. That’s not what I mean. What I see critically instead are tech companies continuing to advertise their services as facilitating these connections, when they actually do so less and less in favor of more sponsored content and AI bots, and that the best way for connection to happen is to have an endless supply, and on their platform. They were extremely successful in convincing many of us that merely having potential access to more people, and many more people having access to us directly, is an advantage and counts as “being connected” (meaning: more than the simple software connection between us). I just don’t believe that, at least for a private person who doesn’t need to win over customers or become a brand. We can see daily that most of us are just not equipped to handle thousands or more people coming at us online. There’s good reasons why famous people used to have a more filtered access to fans via fan mail, interviews, magazines and the occasional meet and greet, plus a PR team and media training. There is a sweet spot when we have relationships to just enough people to be happy without the attention becoming a burden. These companies have conflated a sort of passive consumption, access and surveillance with “connection” and “relationships”, using the image of keeping up with friends and family via a platform to imply that thousands consuming your posts without ever talking to you and more or less surveilling you as a stranger counts the same. They have facilitated a business model around parasocial behaviors with influencers via this exact narrative. They also want you to believe that you need their platforms for relationship maintenance, and they have succeeded, with many claiming they would not be able to get a hold of their inner circle or know about their lives if they deleted their account… which is sad. The idea that you cannot interact with family anymore without this platform, that you can go through millions of strangers to find your next best friend or partner or other opportunity, keeps you on it. The exchange of posts across millions of people keeps each other on the platform too, as you’re always looking for new posts and never run out. No one would use it if it was dead, and they’d use it less if a post couldn’t generate these juicy numbers. That reinforces itself. There’d be less posts to consume if most people limited their profiles and posts for privacy, and ragebait loses its teeth if everyone just blocks the poster or blocks each other too freely. People are also expected to make themselves available 24/7 and overshare, which helps mine additional data and creates more attractive and scandalous content round the clock for the other users to consume, as opposed to just using it for an hour a day. All of these factors have in common that huge masses of people need to be almost constantly available, active and not walled off to each other. That means no limits via settings, friend lists, block lists, feeds that only show who you follow, friction or time constraints, because then the free flow of “content” is disrupted and people spend less time on the app. That could also mean your friends and family drop off too, so you don’t stay there either, which means less eyes on ads and less data to harvest. So of course they’d want to counter this possible risk with the notion that the average Joe needs to open himself up to the eyes of millions because "connection is good!" and maybe you’ll even go viral and earn money. Don’t you wanna be “connected”? Why are you isolating yourself? You’re so weak for blocking that person, and you’re missing out by privating your profile or deactivating it, and you’re antisocial by not posting! Meta went notoriously hard on pushing its capability to be hyperconnective: In Careless People by Sarah Wynn-Williams , she describes again and again how Mark Zuckerberg met with lots of important authorities and key political figures to underline how the platform could connect, just to get more users 1 , without taking responsibility for what their platform would enable in some of the most heated regions - even hiding their role in the outcome of the 2016 US presidential election by pushing their narrative about openness and connection 2 . They also disregarded the setting not to import phone contacts and implemented the "People You May Know" feature 3 to make more "connection" happen, jeopardizing people's safety and privacy to do so. In general, bringing internet to other disadvantaged and cut-off countries is a good thing, and they did launch Internet.org (now: Free Basics) to allegedly aid with that 4 ; however, it quickly devolved into just providing rudimentary Facebook versions to these countries (Facebook Zero), becoming essentially the entirety of the internet in these places and therefore controlling it completely just to gain more users and influence 5 , and only pleaded with countries under the guise of connection to get unblocked, especially by China 6 . They even created a Connectivity Lab in 2014 7 , invested in a Connectivity Declaration and spent over 1 Million dollars on full-page advertisements for it. They even got positive press by CNN and Reuters about pleading with the UN that connectivity over their platform could eradicate "global ills" like extreme poverty 8 . Not only that, but as many probably already know, Meta has been pushing chatbots and fake AI profiles on their platforms (especially Instagram) for a year or so now. The goal is to keep you there still, as less and less people actually talk to each other while just passively consuming content. As the net gets taken over by bots, what’s the advantage of connecting with them? Connection at all costs huh, even if there's no human involved? That is where the idea of it starts to crumble and fall apart. Which is why the need for connectivity in the way these companies mean it and push it is a big lie just to further their financial interests and has nothing to do with how humans actually pursue, facilitate and experience true connection, and we need to question it. Discussions around isolation and viewership online are a bit skewed for me for that reason, especially when they happen outside of the mega-platforms and are about blogging, because they apply the marketing we internalized on social media to other spaces who don’t depend on this lie. My friend Suliman said something very sweet recently about discoverability in the indie web: “But what's the point of a home on the internet if you're living it alone? There's a saying in Arabic that says "a Heaven without people is no Heaven" and I think it's truer in our modern day than ever. We're already so isolated, so why isolate ourselves even further?” I think this is true for the offline context, but I am not convinced about how well it works for the online world. I am concerned this view on connection uncritically applied to online spaces is playing too much into the financial interests of Meta and others and is, at least partially, learned behavior from growing up on their platforms, and growing up in a capitalistic era that urges you to use everyone you know for professional networking, extracting favors and all to attain better work, housing, and donations. I can only speak for myself, but the reason why I would be able to be completely alone, unread and ignored online is because I already get all the connection I need offline. Online is a bonus, or a fallback. Not to mention that it could overlap and only my offline relationships could read my blog. Would that not be enough? Connections I have offline are people I can visit flea markets with, play board games with, we share beds and food and I can rely on them when I’m sick. Meanwhile, the online people I am supposed to crave being connected to en masse can give me an upvote, and an email - which is very appreciated, but it is just not on the same level. Online people absolutely can become offline people, as I met my wife online and have had good internet friends. But that, as shown above, has nothing to do with the widespread passive consumption and access that is presented under the guise of connection by these giants who abuse it. I don’t feel connected by simply witnessing someone exist; neither on social media, nor around the blogosphere. To me, saying I need people online to notice me to not be isolated is like telling me I need to go to Times Square on New Years Eve to not be lonely. All that will happen is that I’d feel lonely while surrounded by other people and noise. We should not value quantity over quality, and I don't want to pretend that the attention economy that these companies have instilled to further their own power is my way to find true connection. Reply via email Published 30 Nov, 2025 Small selection: Pages 81 (Myanmar Junta), 108 and 168 (Colombia), 181 (panel of several presidents in Panama), 186 (President Roussef) ↩ Page 256 ↩ Page 62 ↩ Pages 106-108 ↩ Page 203 ↩ Pages 144-145 ↩ Page 107 ↩ Page 194-195 ↩ Small selection: Pages 81 (Myanmar Junta), 108 and 168 (Colombia), 181 (panel of several presidents in Panama), 186 (President Roussef) ↩ Page 256 ↩ Page 62 ↩ Pages 106-108 ↩ Page 203 ↩ Pages 144-145 ↩ Page 107 ↩ Page 194-195 ↩
As of today, I run three different newsletters, all powered by Buttondown: there’s my recently announced Dealgorithmed , my outdoors-focused From the Summit , and the People and Blogs series. I also send my blog posts via email , if you prefer to consume content that way. They all require double opt-in. Which means that if you signed up for one of them, you should have received a second email, asking you to click a link to confirm your subscription. Sometimes those land in the spam folder, sometimes they don’t arrive at all. That’s just the unfortunate reality of emails in 2025. I just checked, and a solid 10% of the people who have signed up for Dealgorithmed have not confirmed their address. This is a reminder to check your inbox and click the confirmation link otherwise, you will not receive the first edition when it goes out on January 1st. Thank you for keeping RSS alive. You're awesome. Email me :: Sign my guestbook :: Support for 1$/month :: See my generous supporters :: Subscribe to People and Blogs
Business
Journalism
Media
Entertainment
Culture
Career
AI
Go
Sql
Java
Open Source
Cloud
Programming
Ruby
Marketing