favorite re:publica 26 talks - part 1
Sadly couldn't attend re:publica 26, but the talks have been recorded and uploaded on YouTube. If you don't know: It's a conference in Germany with workshops and talks that deal with internet topics, specifically digital rights, media culture, online presences like blogs or social media, and our general information society. So I watched many of the uploaded talks by now, and here is part 1 of my favorites! Initially, I wanted to do one post, but it's getting way too much/long, and I'm dragging my feet watching the rest I planned to. Most of these talks are in German, some are in English, but I guess you can also use YT's auto dub feature (which I find horrible, but has become a bit better lately, as far as I can tell from when it was suddenly turned on). Let's start! OpenClaw - Anatomy of the Wave - great talk about AI agents, how to set one up, and interesting use cases. Kinda sold me on at least going through the setup process, especially as one reader once inspired me to give open source local models a try. Interesting tidbits: The idea of letting an employee close to retirement use an agent so it will save all the knowledge that will otherwise be lost or won't be covered in onboarding of a replacement; anti-distill tech, that masks your skills and outputs so that models cannot be trained to replace you; the theory that as more and more people will talk with AI, we will go back to a predominately oral culture, worse grammar and writing (as AI will understand anyway, even with mistakes), and potentially affecting how we talk to each other; the appeal to build your own agent before the big players in the AI space barter over you and your data, and so that you are independent from platform capitalism. On Counter-Power (with Arne Semsrott) - absolutely banger talk, my favorite. I love Arne, I also bought his latest book. Interesting tidbits: A lot of INSA studies about public opinion/voting that show off the AfD's success in Germany are paid for by BILD and NIUS, and the INSA CEO has a past of supporting the AfD and other far-right and conservative groups; the %s of AfD in those polls looks scary, but when compared to the last election and how many groups did vote for a Bürokratieabbau for power and money, but the most vulnerable ( Bürgergeld/Grundsicherung , Immigration), get more and more of it; political and civic participation needs to be fun and enjoyable! Young, blonde, right-wing... & AI-generated is a talk about how many prominent rightwing personalities online (that rightwing politicians even interact with and platform on their account) are AI-generated, especially young, blonde white women. AI makes it easier to produce more polished content; even when we consider it slop, it is polished in a specific way that resembles high quality marketing stuff that AI is trained on. That makes it easier for rightwingers to churn out hard-hitting images and content of their fabricated reality that their audience laps up, and they especially focus on decline porn , which generates attention, money and power through the fantasy that we are close to a complete collapse etc. Ordinary media and their contribution to the comeback of fascism talks about how false neutrality in the biggest reputable/respectable news media has contributed to the rise of fascism. Everything is "debatable" or "controversial", "critics say" what they don't directly want to say, and therefore outsource Machtkritik onto third parties. One great example: NYT wrote "Critics complain that Italy's Government is interfering in the arts" but it actually was, as was explained in the text, so the title should have just said Italy's government is interfering in the arts. This trick gets used like a shield to blame a view on critics instead of standing by what they write. Tagesschau platforms BILD and NIUS like they are respectable platforms just to play a "both sides equally" game, false dichotomies get created for outrage and clicks when there is a more nuanced view they do not offer because it would actually involve journalistic work of contextualizing what politicians say with facts. Trumps unhinged tweets get sanitized, normalized and softened by news by breaking them down into "giving Iran more time to open the Strait" as if it was a respectable position. The Authoritarian Stack with Francesca Bria. She's showing stuff from the Authoritarian Stack and Euro Stack websites, explaining how much tech oligarchs are building and controlling the essential systems of our modern society; social media, office software, chips, internet, raw materials etc. while they are building a post-democratic world. They embed themselves into states and governments through procurement, venture capital, and personnel pipelines; supply chains and dependencies are weaponized, bottlenecks become goldmines (land, power, chips, mines, data centers) and political alignment with Washington becomes a requirement to access advanced compute. Interesting tidbits: Ideology steers the Venture Capital, which funds the companies, which send people off into government positions (State Capture) which leads to Regulation/Deregulation in their favor, and they get Gov Contracts which lets them build Infrastructure. We all pay for it via taxes, because those get invested when government buys their products (esp around immigration control, war, etc.); Marc Andreessen spent more on the election than Soros and Musk; General Matter reactors are the first privately owned uranium enrichment plant in America, Thiel sits on the board, 900 million contract; every year over 2 trillion euros of public EU procurement go to fund our dependency. Digital arms race (with Michael Kolain) covered whether we are in a digital arms race, the issues in it, and how we might get out. Old terms and concepts like disarmament don't work anymore, as these AI systems are dual use; spending in AI is also investing into AI warfare because of this. This means at some point, since systems and countries can no longer be disarmed, there is always the threat of attack, a silent war that never ends, a hyperwar that is unparalleled in speed and scale. Mirjam Walser – Unheard of! AI, animals, and the question of whose interests count is a great presentation on the animal rights aspects of it all. AI is promised to increase animal welfare, but it is a ruse of the big players in animal agriculture to optimize productivity and profitability. The goal is completely autonomous mass killing in giant meat plants, no humans needed. Interesting to note: Anthropic is, so far, the only company whose guidance includes respecting animals and all sentient beings. Trust in the AI Era: How does digital information remain credible? says that we are going through a paradigm shift in which we do not need to ask whether a picture is real, but instead asking ourselves where it comes from, who made it and what got changed. Also, interesting facts around content credentials; some cameras in Samsung and Pixel phones, as well as some Lyca model, cryptographically sign an image during the shot so it is known it was actually physically taken ("real"). Different approaches taken/different orgs around it are C2PA, CAI, CR. There are different approaches to both integrity and identity layers of this confirmation. Manipulation will be a daily thing, but we need visible signs and control so we can choose how to engage with it. High time for sexual media literacy covers digital sexuality (Digisexualität, Mediensexualität) for everyone, but especially teens. It makes clear that it's not only always about porn or porn websites; people also use the internet for sexual education, flirting and sexting with each other (or with bots) over social media platforms and games, exploring sexuality and finding validation from other sexual minorities, finding each other to have sex in real life, seeing thirst traps on the feed, etc. and unfortunately, they also get sexually harassed, groomed, deepfaked, and there is (new term for me!) Sharegewalt ('share violence', non-consensual sharing of nudes to other parties). Nowadays, children come into contact with sexual media at 11 years old on average; analog consensual sexual experiences tend to happen between 14-19, where most people report having their first intercourse at 19. This used to be earlier for past generations. Digital sex becomes a practice range for analog sex nowadays, which makes sense, as digital no's are easier than real life no's, and digital experiences can be paused or stopped easily. The big point is: It is important for teens to be able to find ways to cope with their changing, sexualized bodies, to experience themselves as a sexual being and find out what they like, see their effect on others, test their desires and how it is to generate sexual attention, and find out where they are on the gender binary (or outside of it). It's especially important for queer people, disabled people, abused people, socially anxious people or people who are surveilled a lot at home. We should not think of teens as passive victims to sexual content, but as consumers, and continue improvements in sex education, and expand it to teaching media competence around porn and sexting (how to keep yourself safe, how to draw boundaries, detect cybergrooming, knowing you have legal rights etc.). Classism in Digital Spaces - I really struggled not to cry watching this one because it hits close to home. You're not only confronted with different economic realities offline, but also online, where everyone shares their highlight reel and makes their life look more luxurious. It can quickly feel like you're the only one struggling, or the only one with an upbringing in poverty. As a poor person, you are shamed, and people who are more well-off than you refuse to believe your stories, and then when they feel guilty for earning more, they project that onto you, and shame you for making them feel bad. They act like having money is so hard and a burden that was placed on them, or as if they should not pay more of a collective bill if they also earn more. It is also staggering just how many people do poverty cosplay both online and in real life; their family has been comfortably middle class all their life, yet they somehow co-opt poverty discourse as if it affected them. I get mad when I interact with someone in that way and then later find out their family has multiple houses, for example. People like that don't know how good they have it; despite Germany being one of the richest countries, every fourth child is growing up in poverty by now. So many of my friends struggle to find jobs, and also rely on Bürgergeld/Grundsicherung , and for them it's like being in an abusive relationship. This huge, powerful entity gives you a very limited amount of money while degrading you, insulting you, and making you jump through hoops trying to prove you are not guilty of something , and there is nothing you can do about it. You can't evade it, or can't break out of it, you can't defend yourself. You are forced to endure it until you can find a job, which could be next week or in 5 years. You're living under constant threat of losing everything, and are expected not to be a fucking nervous wreck during it? Next time you read about how a high percentage of people getting benefits isn't looking for a job, remember that this includes minors, people close to retirement who don't get hired due to their age, disabled people, people doing care work for the children and elderly in their family, people who already work a low-income job and get the benefits as supplement, and more. Defeating digital corporations with class action lawsuits - collective legal action like that is new in the EU (since 2020), and this video explains the different kind of options and what they are used for. I will not bore you with legal details, but it was very enlightening for me! Includes real life example of class actions against X, TikTok, Amazon and more. Sorry, not sorry – The Art of (Not) Apologizing in Public - very helpful for me as a person who struggles to understand apologies and what makes a good one, and why social media apologies are the way they are. This may sound odd, but for most of my life so far, I just never understood what was expected of me when I was made to apologize, and what it is for. I often don't regret any behavior I know I am supposed to apologize for, and in many contexts while growing up, just did it because I knew it was the "right" thing to do, without actually feeling remorseful or wrong about what I did (nowadays, I apologize when I feel genuinely remorseful). Apologies by others unfortunately give me absolutely nothing and don't resolve anything for me emotionally (the result of growing up with people who apologized to me doing the thing they apologized for again and again, or abusers not apologizing to me ever, and also me being forced to apologize when I didn't want to, I guess, which made it all feel like a fake performance and scam), so I have a very limited view of what others actually feel about it. I appreciate this deep dive into what an apology is, what it is good for, and why some people apologize while others don't. The Future of Human-Machine Relationships was cool to watch because it's led by an actual researcher summarizing her studies on how humans interact with AI, what it means to them, what influences it, and the range of answers to some of the research questions. She makes clear that AI should be a bridge into the real world, not a wall, and not an escape. Related talk: When AI simulates consciousness , about how pareidolia and the frictionless, sycophantic design of the big GenAI aids the humanization of them. LLMs as they are publicly available right now are a mass social experiment. ... but there are many, many, many, maaaaany more videos and talks that happened, so feel free to check out the rest and see if you find anything you like! Part II hopefully coming soon. Reply via email Published 19 Jun, 2026 OpenClaw - Anatomy of the Wave - great talk about AI agents, how to set one up, and interesting use cases. Kinda sold me on at least going through the setup process, especially as one reader once inspired me to give open source local models a try. Interesting tidbits: The idea of letting an employee close to retirement use an agent so it will save all the knowledge that will otherwise be lost or won't be covered in onboarding of a replacement; anti-distill tech, that masks your skills and outputs so that models cannot be trained to replace you; the theory that as more and more people will talk with AI, we will go back to a predominately oral culture, worse grammar and writing (as AI will understand anyway, even with mistakes), and potentially affecting how we talk to each other; the appeal to build your own agent before the big players in the AI space barter over you and your data, and so that you are independent from platform capitalism. On Counter-Power (with Arne Semsrott) - absolutely banger talk, my favorite. I love Arne, I also bought his latest book. Interesting tidbits: A lot of INSA studies about public opinion/voting that show off the AfD's success in Germany are paid for by BILD and NIUS, and the INSA CEO has a past of supporting the AfD and other far-right and conservative groups; the %s of AfD in those polls looks scary, but when compared to the last election and how many groups did vote for a Bürokratieabbau for power and money, but the most vulnerable ( Bürgergeld/Grundsicherung , Immigration), get more and more of it; political and civic participation needs to be fun and enjoyable! Young, blonde, right-wing... & AI-generated is a talk about how many prominent rightwing personalities online (that rightwing politicians even interact with and platform on their account) are AI-generated, especially young, blonde white women. AI makes it easier to produce more polished content; even when we consider it slop, it is polished in a specific way that resembles high quality marketing stuff that AI is trained on. That makes it easier for rightwingers to churn out hard-hitting images and content of their fabricated reality that their audience laps up, and they especially focus on decline porn , which generates attention, money and power through the fantasy that we are close to a complete collapse etc. Ordinary media and their contribution to the comeback of fascism talks about how false neutrality in the biggest reputable/respectable news media has contributed to the rise of fascism. Everything is "debatable" or "controversial", "critics say" what they don't directly want to say, and therefore outsource Machtkritik onto third parties. One great example: NYT wrote "Critics complain that Italy's Government is interfering in the arts" but it actually was, as was explained in the text, so the title should have just said Italy's government is interfering in the arts. This trick gets used like a shield to blame a view on critics instead of standing by what they write. Tagesschau platforms BILD and NIUS like they are respectable platforms just to play a "both sides equally" game, false dichotomies get created for outrage and clicks when there is a more nuanced view they do not offer because it would actually involve journalistic work of contextualizing what politicians say with facts. Trumps unhinged tweets get sanitized, normalized and softened by news by breaking them down into "giving Iran more time to open the Strait" as if it was a respectable position. The Authoritarian Stack with Francesca Bria. She's showing stuff from the Authoritarian Stack and Euro Stack websites, explaining how much tech oligarchs are building and controlling the essential systems of our modern society; social media, office software, chips, internet, raw materials etc. while they are building a post-democratic world. They embed themselves into states and governments through procurement, venture capital, and personnel pipelines; supply chains and dependencies are weaponized, bottlenecks become goldmines (land, power, chips, mines, data centers) and political alignment with Washington becomes a requirement to access advanced compute. Interesting tidbits: Ideology steers the Venture Capital, which funds the companies, which send people off into government positions (State Capture) which leads to Regulation/Deregulation in their favor, and they get Gov Contracts which lets them build Infrastructure. We all pay for it via taxes, because those get invested when government buys their products (esp around immigration control, war, etc.); Marc Andreessen spent more on the election than Soros and Musk; General Matter reactors are the first privately owned uranium enrichment plant in America, Thiel sits on the board, 900 million contract; every year over 2 trillion euros of public EU procurement go to fund our dependency. Digital arms race (with Michael Kolain) covered whether we are in a digital arms race, the issues in it, and how we might get out. Old terms and concepts like disarmament don't work anymore, as these AI systems are dual use; spending in AI is also investing into AI warfare because of this. This means at some point, since systems and countries can no longer be disarmed, there is always the threat of attack, a silent war that never ends, a hyperwar that is unparalleled in speed and scale. Mirjam Walser – Unheard of! AI, animals, and the question of whose interests count is a great presentation on the animal rights aspects of it all. AI is promised to increase animal welfare, but it is a ruse of the big players in animal agriculture to optimize productivity and profitability. The goal is completely autonomous mass killing in giant meat plants, no humans needed. Interesting to note: Anthropic is, so far, the only company whose guidance includes respecting animals and all sentient beings. Trust in the AI Era: How does digital information remain credible? says that we are going through a paradigm shift in which we do not need to ask whether a picture is real, but instead asking ourselves where it comes from, who made it and what got changed. Also, interesting facts around content credentials; some cameras in Samsung and Pixel phones, as well as some Lyca model, cryptographically sign an image during the shot so it is known it was actually physically taken ("real"). Different approaches taken/different orgs around it are C2PA, CAI, CR. There are different approaches to both integrity and identity layers of this confirmation. Manipulation will be a daily thing, but we need visible signs and control so we can choose how to engage with it. High time for sexual media literacy covers digital sexuality (Digisexualität, Mediensexualität) for everyone, but especially teens. It makes clear that it's not only always about porn or porn websites; people also use the internet for sexual education, flirting and sexting with each other (or with bots) over social media platforms and games, exploring sexuality and finding validation from other sexual minorities, finding each other to have sex in real life, seeing thirst traps on the feed, etc. and unfortunately, they also get sexually harassed, groomed, deepfaked, and there is (new term for me!) Sharegewalt ('share violence', non-consensual sharing of nudes to other parties). Nowadays, children come into contact with sexual media at 11 years old on average; analog consensual sexual experiences tend to happen between 14-19, where most people report having their first intercourse at 19. This used to be earlier for past generations. Digital sex becomes a practice range for analog sex nowadays, which makes sense, as digital no's are easier than real life no's, and digital experiences can be paused or stopped easily. The big point is: It is important for teens to be able to find ways to cope with their changing, sexualized bodies, to experience themselves as a sexual being and find out what they like, see their effect on others, test their desires and how it is to generate sexual attention, and find out where they are on the gender binary (or outside of it). It's especially important for queer people, disabled people, abused people, socially anxious people or people who are surveilled a lot at home. We should not think of teens as passive victims to sexual content, but as consumers, and continue improvements in sex education, and expand it to teaching media competence around porn and sexting (how to keep yourself safe, how to draw boundaries, detect cybergrooming, knowing you have legal rights etc.). Classism in Digital Spaces - I really struggled not to cry watching this one because it hits close to home. You're not only confronted with different economic realities offline, but also online, where everyone shares their highlight reel and makes their life look more luxurious. It can quickly feel like you're the only one struggling, or the only one with an upbringing in poverty. As a poor person, you are shamed, and people who are more well-off than you refuse to believe your stories, and then when they feel guilty for earning more, they project that onto you, and shame you for making them feel bad. They act like having money is so hard and a burden that was placed on them, or as if they should not pay more of a collective bill if they also earn more. It is also staggering just how many people do poverty cosplay both online and in real life; their family has been comfortably middle class all their life, yet they somehow co-opt poverty discourse as if it affected them. I get mad when I interact with someone in that way and then later find out their family has multiple houses, for example. People like that don't know how good they have it; despite Germany being one of the richest countries, every fourth child is growing up in poverty by now. So many of my friends struggle to find jobs, and also rely on Bürgergeld/Grundsicherung , and for them it's like being in an abusive relationship. This huge, powerful entity gives you a very limited amount of money while degrading you, insulting you, and making you jump through hoops trying to prove you are not guilty of something , and there is nothing you can do about it. You can't evade it, or can't break out of it, you can't defend yourself. You are forced to endure it until you can find a job, which could be next week or in 5 years. You're living under constant threat of losing everything, and are expected not to be a fucking nervous wreck during it? Next time you read about how a high percentage of people getting benefits isn't looking for a job, remember that this includes minors, people close to retirement who don't get hired due to their age, disabled people, people doing care work for the children and elderly in their family, people who already work a low-income job and get the benefits as supplement, and more. Defeating digital corporations with class action lawsuits - collective legal action like that is new in the EU (since 2020), and this video explains the different kind of options and what they are used for. I will not bore you with legal details, but it was very enlightening for me! Includes real life example of class actions against X, TikTok, Amazon and more. Sorry, not sorry – The Art of (Not) Apologizing in Public - very helpful for me as a person who struggles to understand apologies and what makes a good one, and why social media apologies are the way they are. This may sound odd, but for most of my life so far, I just never understood what was expected of me when I was made to apologize, and what it is for. I often don't regret any behavior I know I am supposed to apologize for, and in many contexts while growing up, just did it because I knew it was the "right" thing to do, without actually feeling remorseful or wrong about what I did (nowadays, I apologize when I feel genuinely remorseful). Apologies by others unfortunately give me absolutely nothing and don't resolve anything for me emotionally (the result of growing up with people who apologized to me doing the thing they apologized for again and again, or abusers not apologizing to me ever, and also me being forced to apologize when I didn't want to, I guess, which made it all feel like a fake performance and scam), so I have a very limited view of what others actually feel about it. I appreciate this deep dive into what an apology is, what it is good for, and why some people apologize while others don't. The Future of Human-Machine Relationships was cool to watch because it's led by an actual researcher summarizing her studies on how humans interact with AI, what it means to them, what influences it, and the range of answers to some of the research questions. She makes clear that AI should be a bridge into the real world, not a wall, and not an escape. Related talk: When AI simulates consciousness , about how pareidolia and the frictionless, sycophantic design of the big GenAI aids the humanization of them. LLMs as they are publicly available right now are a mass social experiment.
Gaming
Culture
AI
Data
Business
Books
Writing
Programming
UI
Media
Security
Career
Music
Ruby