AI Billionaires Want to Control EVERY Aspect of Your Life — Aaron Bastani Meets Karen Hao

[00:00:00] Karen Howe: These billionaires have their bunkers. I was just in New Zealand and that's the place where they've all decided to build their bunkers. And I actually learned for the first time why they all chose New Zealand. It's because there was an analysis of where, if nuclear bombs detonate, like the probability of where they would detonate. When you look at the map of the blast radius, New Zealand is almost always outside of that blast radius. So that's why they're building their bunkers there. You know, like they will be fine. However the world ends, they're going to exist outside of that. They're like, this is inevitable. There's nothing you can do about it. And so all that's left for us to do is to make sure that we are part of the winners, not the losers. [00:00:49] Speaker 2: At some point this year, Elon Musk will become the world's first trillionaire. That's right. He'll be worth more than a trillion US dollars. It's a benchmark nobody's got to before. Remember the first company wasn't worth a trillion dollars until 2018. That was Apple, by the way. How is he getting so rich so quickly? A big part of it is of course, SpaceX. It'll be floated on the stock market this year. And part of that flotation is XAI. So we are now entering a world, not just a trillion dollar companies, but trillion dollar individuals. And that goes hand in hand with an industry, artificial intelligence, which is remaking commerce, politics, and society more generally. Its implications are huge. And our concerns collectively about it are growing by the month. But here's the thing. The more worried we get, the more we use it. And there's obviously some hypocrisy going on there, but that's the world, isn't it? It's always replete with contradictions. The point is politically, how do you manage a technology this powerful, this disruptive, this prone to changing the status quo? Or on the other hand, is that all just marketing rubbish? Have I fallen for the ads and the PR from the likes of OpenAI, Elon Musk, and Meta? Those are really important questions. And they were first touched upon by today's guest, Karen Howe, in the hardback version of this book, Empire of AI. We talked a year ago, superb conversation, did really well, not least because it's a really important topic. I'm very happy to have Karen back today because in the last 12 months, the empire of AI, its borders and its generals have changed to a pretty substantial degree. Karen Howe, welcome back to [00:02:30] Karen Howe: Downstream. Thank you so much. So excited to have you back on. I'm also excited to be back on. [00:02:36] Speaker 2: Before we started, I'm going to put smoke up the backside of Novara Media. Before we started, you were kind enough to say, I'm just really excited to talk to you, talk about your book. The first conversation was so illuminating for me personally. And you said, I don't do second interviews. [00:02:51] Karen Howe: I really try to avoid it because I've had a really crazy schedule, but I just loved my Novara media interview. It was my favorite one of the entire tour. So I'm coming back. [00:03:03] Speaker 2: Oh, it's so nice to hear. It did so well. And it's got millions of views. And I think it's because you just, you struck the right chord at the right time. And I suppose that's the story of the book more generally, isn't it? You've catalyzed the conversation, but I mean, it wouldn't blow up in the way that it did if it had just catalyzed the conversation. You've also fed into real concerns and anxieties that are globally held at this point. [00:03:29] Karen Howe: Yeah. Yeah. I didn't realize, I think when we had our first interview and I was just beginning the journey of going on book tour, I didn't really know how the book would land yet, even then. And I didn't really understand how rapidly the global conversation around AI would change. I mean, it's literally been just a year between our last interview and today and the turnaround on how people talk about AI now has been remarkable. And I think a lot of it has been because people are paying attention. They are hungry for knowledge. They are actually educating themselves about what this technology is, who is building it, what their agenda is, and they're not standing for it. And they're actually contesting it in resistance movements around the world. [00:04:22] Speaker 2: So what's changed in the last year? Because the book is called, and I should get it up, it's out in paperback. It's very good. I'm sure you bought the hardback, but you can get the paperback now, if you haven't got that. I always say this as well, if you want to radicalize loved ones and friends, buy books. What's changed with the empire of AI? How have the frontiers changed? What new generals are on the scene? [00:04:42] Karen Howe: To take the metaphor. [00:04:46] Speaker 2: Yeah. What rebellious provinces are there and so on? You said a lot's changed. So what really stands out? [00:04:53] Karen Howe: Yeah. Well, I think the more that I've talked about the book, the more that it's solidified in my mind that ultimately the most important way to understand the AI story today as it's conceived of by Silicon Valley is that it is a political project. And the central feature of this political project is taking agency away from people so that the empire can be the monopoly on decision-making power in the world, on shaping everyone's future. And the thing that we've seen is people aren't actually accepting this narrative anymore. I mean, when I first started reporting the book, every community that I encountered would say the same thing. It was the sense of powerlessness, the sense of there's nothing that I can do in the face of this immense amount of control and controlling influence from these entities. And now there are so many people around the world, whether individuals, communities, or institutions that are like, wait a minute. We have our own power. We have our own agency. And even more so if we act as a collective. I recently launched this new project with a group of journalists, critical scholars, and AI researchers called the AI Resist List. It is this project that documents around 30 different resistance actions around the world from some of the most unexpected places in challenging the way that the empire is engaging in its imperial project. And we are seeing, you know, all of these different people engaged in political activism, also artistic resistance. One of my favorites is this community that I think we talked about last time on the pod, because I talk about it in my book, Jilicora Chile. It's a community that is dealing with a lot of data center development in Latin America. It's one of the most most data center development hubs, hotspots in the continent. And this community, I wrote about it because they had the spirit of resistance already when I was documenting things back then. And they were pushing back and really mobilizing not just their own locale, but, you know, international media attention to, to show how this technology that is often talked about as digital and ethereal and existing in the cloud is actually very physical and extracts the resources of impoverished communities like theirs. Since the publication of the book, they have now taken that further. And they developed this project called Keeley.ai where you go on this website and it looks like a chat bot and you can type into a box and then you get an answer. But instead of an AI model answering you, it's actually someone within their community that is answering you. And so they did essentially an artistic demonstration where for 12 hours, literally anyone could type into Keeley.ai. And if someone said, I want a picture of a dog, it was routed to their local artist named Benji. Benji would take a piece of paper out, draw a dog, snap a photo, send it back to the person. And it was their way of creating a global conversation, which they did because this went absolutely viral, creating a global conversation around why is it that we mindlessly ask chat bots these questions when, first of all, it takes resources from a community like theirs. And second of all, we could just be asking people, we could just be using it as an opportunity to rebuild our social relationships. And in that way, resisting the divide and conquer nature of empires themselves. [00:08:44] Speaker 2: I've got a lot of questions about chat bots because, and I think this gets to the heart of what's changed as well over the last 12 months is that people are really skeptical about the stuff and they're way more on it at a macro level, at a micro level, global north, global south. But I also think people are using it more, right? And so there's this sense probably of heightened hypocrisy. I don't think people should feel hypocritical about these kinds of things. A lot of that is just, you know, you can't not be on Facebook, for instance, these days, right? It doesn't mean you can't criticize the big tech. I'm not on Facebook. Well, generally, you know, if you're a young person, I know, certainly for me, I left Facebook and then I was looking for a house share in my late 20s. Right, right. You couldn't do it. Yeah, yeah. That kind of thing. But you know, so we're not here to talk about hypocrisy and point fingers, but I think many people will feel that way. So for me personally, I don't use ChatGPT. I use Google Gemini sometimes. And I'll give an example of a search I did recently. So my wife and I want to go to Northern France for a couple of weeks. We just want a cheap beach holiday with our daughter. And we want to not fly. We want to get the ferry. And the ferry goes to a place called Ka from Portsmouth. And so I don't know any good beaches near Ka or what good villages, where would be ideal for a toddler, blah, blah, blah, blah, blah. And Google Gemini gives me all of that really succinctly, really quickly, quite tailored. Is that kind of search unethical in your view? And I don't mind you saying it is. It's just, I'm trying to establish what do we think these kind of chatbots might be good for? [00:10:09] Karen Howe: Yeah. So I think you're right that there has been a lot more adoption since we last spoke. Part of it is because people are learning how to use these tools and getting more utility out of them. Part of it is companies are pushing their employees to adopt it. Part of it is also the engaging and somewhat addictive nature of these technologies. So all of these are trends that are pushing in the same direction of more growth of users on these different chatbot systems. The question of ethics is tricky here. To me, the appropriate analogy to draw is to the fashion industry and the way that the fashion industry also had deeply exploitative supply chains. There was a lot of labor exploitation. There was a lot of environmental degradation to produce clothes. And the answer was not no one wear clothes anymore because it's unethical to wear clothes. The answer was these clothes are being developed in a really deeply unethical way. But why should we stand for having clothes developed in that way? You know, there are so many other ways that we could produce clothes. And there was labor organizing. There was consumer pushback, public backlash, government regulation, and significant societal mobilization to then get fashion in the fashion industry to develop other supply chains that were simply better. Where they actually paid their workforce minimum wage. They had safety standards. They were taking a more sustainable approach to that development. Of course, we still today have brands that don't engage in that. But now we've created entire new markets where these fashion brands can actually use their sustainable ethical track record to charge more money and be able to afford that kind of supply chain. And so, I think we need to have the same thing happen with AI. We need to have the same kind of worker organizing, mobilizing, societal pushback, consumer pushback, government regulation, all of these different things to create these new markets, create alternatives. And we also need more people to be thinking about what those alternatives should look like so that ultimately all the people that are using chatbots today actually finding real utility out of them can continue to do that without actually also then engaging in or somehow endorsing the deeply corrosive supply chains that the technology [00:12:26] Speaker 2: currently rests on. We had a conversation recently with an author regarding his book on the Democratic Republic of Congo. Elements of Power, I think the book is called. Really good book. And with, you know, we're talking about coltan and tantalum and cobalt and all these, you know, these minerals and the story is just extraordinary. The DRC is an extraordinary place. And he had a line, the author of the book, he had a line, you know, nobody is ever more than in the West, nobody's more than a few feet away from a piece of the DRC. And it's just so striking, you know, I thought, Christ, and it's the same with AI, isn't it? We're so alienated from the reality of what this is doing. You know, it's just a glass interface and it's all very ethereal and immaterial, [00:13:11] Karen Howe: but the reality is the complete opposite. Yeah, absolutely. And literally the Congo is one of the places that is part of the AI development supply chain. So it's deeply interconnected. I mean, one of the organizations that we feature on the AI resist list is this organization called Friends of the Congo, which has been raising awareness and fighting for the rights of people within the Congo because of the very extractive nature of the AI supply chain, accelerating and exacerbating things like child slavery in the mining processes of these minerals and the environmental degradation and so on and so forth. And so, yeah, it is, it is interlinked directly with what that other author is talking about. The way that Silicon Valley is currently building its technologies is accelerating the extraction and exploitation that has long existed in certain territories and inflaming them, making them worse. You had a great line right at the top of this. [00:14:10] Speaker 2: You know, it's fantastic when you do an interview and somebody says something, you think, well, that's a social media clip. It suggests they know what they're talking about. You said that this is a political AI. The big tech AI is a political project to take power away from people. I mean, that's a really extraordinary pithy encapsulation of what's happening. Your book says that lots of people are saying that it's being increasingly recognized. How have people in the industry responded? Are there some people leaning into that? Like every time I hear about Peter Thiel, for instance, it does kind of feel like he's saying, yeah, that's what I want to do. No shit. Or are they trying to dress it up in different ways, new ways to how you found it when you were writing this book [00:14:49] Karen Howe: before it came out in hardback? Naomi Klein has this really amazing critique of what's happening right now in Silicon Valley, which she describes as end times fascism. And she connects it with all of these other different phenomena that are happening, not just within Silicon Valley, but I think it really encapsulates exactly how some people within the Silicon Valley establishment think about what is happening. I've actually heard people in the upper echelons of Silicon Valley acknowledge that AI as it's currently conceived will exacerbate inequality. And what's so fascinating is when I listen to them talk, I'm like, wait a minute, that's my analysis. We agree with each other on the first part of our analysis of what AI is currently doing to society and what it could continue to do if we continue on this trajectory. And the difference between me and them is the next step of what we do with that. For me, I'm like, so let's stop doing that. You know, let's let's stop building AI in a way that exacerbates that inequality. And for them, they're like, this is inevitable. There's nothing you can do about it. And so all that's left for us to do is to make sure that we are part of the winners, not the losers. And that's what Naomi Klein describes with this, this kind of end times fascism ideology. It's this idea that the world as we know it right now is ending. There's, there's no point in trying to say, you know, climate change is coming. The religious end timers are also saying like the end is nigh and civilization as we know it for the AI boomers and doomers. They're also saying, this is going to be cataclysmic. The shift is on the horizon. And they don't have any desire to save what we have currently. They're just worried about getting to the next place and making sure that they are okay. And so, yeah, I mean, I, they, they, in response, you know, when they see critiques like this, some of them are like, oh yeah, yeah, that sounds about right. That is, you know, the analysis of, of what is happening currently. What they disagree with is what we do about it. [00:17:09] Speaker 2: Here in Navara Media, we have 1.2 million YouTube subscribers, millions of people across social media, hundreds of thousands of you watch these downstreams every week. Why not support our work? Just a tiny fraction of you need to get on board the Navara train and we can take these conversations to an even broader audience. It's never been more important. And I really mean that. It's all up for grabs ahead of the next general election in this country, at least. If you agree, support our work. Go to NavaraMedia.com/support. The link is in the description below. Help us build a new media for different politics. So where do you think this kind of learned helplessness comes from? Because you mentioned climate change and it's the exact same thing, right? We are sitting in London on a very hot day. Maybe people can tell from our, you know, Tans. You've been in the UK for a few days. You're in hay on wire. I presume it was, you know, scorching heat. Maybe not many people from North America associate Britain with that. But it's very hot. We've had the hottest day on record in May, two days in a row. Very strange. We're still in spring. And what's interesting is that people that were saying climate change wasn't real, even five years ago, some of those people now, we covered it on the show recently, are saying, "Well, Australia has 45 degrees heat. Why can't Britain?" And you think, well, that's slightly challenging because Australia is also 18% desert, you know, 70% of it's arid or semi-arid. And they also lost, I think, 2 billion animals a few years ago. So it's this strange, as I say, learned helplessness, which seems to define the ruling class of our age, which is so strange, because on the one hand, they had this Promethean rhetoric about master of the universe. We're going to change everything. But fundamentally, actually, they think we can't really change anything. We're victims of circumstance. All of us are victims of circumstance. What do you think explains that strange juxtaposition? [00:18:59] Karen Howe: I think that they have to tell themselves this story in order to be okay with not taking the burden of responsibility to actually make the world better. If they don't acknowledge their own agency in creating these problems, and therefore the burden of responsibility to fix these problems, and they instead situate themselves in the broad sweep of history, you know, marching forward without them, and without all of us, apparently, participating in it, then they get away with it. They can get away with it. And it is not only useful in washing them of that responsibility, but it's also they can hold that position because they are the ones that will be okay if nothing is done about it, right? These billionaires have their bunkers. I was just in New Zealand, and that's the place where they've all decided to build their bunkers. And I actually learned for the first time why they all chose New Zealand. It's because there was an analysis of where that if nuclear bombs detonate, like the probability of where they would detonate, when you look at the map of the blast radius, New Zealand is almost always outside of that blast radius. So that's why they're building their bunkers there, you know, like that they will be fine. However the world ends, they're going to exist outside of that. And that's why they can exist in this space where they neither have any kind of moral responsibility for actually ushering in a better world, and nor do they have any ramifications for not taking on that responsibility. You know, there's Neil Postman who talks about [00:20:59] Speaker 2: technopoly. I think we mentioned this the last time we had a conversation, which is to say, he writes a great book called Technopoly. He says we move in three stages from tool using societies to technocracies to technopoly, where increasingly people default to the needs of technology. People serve technology rather than the other way around in technopolis. That's his view. And I subscribe to that, I think it's true. But I also, I think what your point there a moment ago is that the fatalism is is politically convenient because of the material interests of these people. So these aren't ideological presumptions they have, which are an outgrowth of the age. This is the spirit of the age, which is kind of what Neil Postman says. Maybe that is also true, but I think you're right. They subscribe to the fatalism precisely because it allows you to create the world's first trillionaires, which is what Elon Musk is going to become in the next couple of months with the IPO, the flotation of SpaceX, which also includes an AI company, XAI. Yes. That is a new development. You know, I don't think we had the first trillion dollar company until 2018. We're now going to have a trillionaire. [00:21:56] Karen Howe: Yeah. [00:21:58] Speaker 2: You've had lots of conversations in the last 12 months with people at the level of activists and journalists, but also I presume amongst the elite. Do they not find this stuff kind of weird? Like we, when we're now going to have trillionaires? Is that not kind of a bit like. [00:22:12] Karen Howe: If you mean elite broadly defined as not just the tech elite, but just other elites like business elites or political elites around the world. Absolutely. They find it weird. I think part of it is because they're envious. Because they're not the ones that are being minted as trillionaires, but yeah, they, they, they, they are just baffled by the degree to which inequality has been allowed to continue to corrode society. And I think, you know, I was, I was in this conversation with a group of business elites from Europe and Asia and they were very, very cognizant of the fact that inequality is one of the most corrosive forces that destabilizes the environments in which they can ultimately do business, but also just corrodes democratic society in general and a healthy functioning society. And I was really struck by how much, once again, there was this learned helplessness. Like they recognize the analysis, they recognize this problem, but they felt that still there were people more powerful than them, AKA the US government, these US Silicon Valley business executives that somehow have more power. Therefore they just must operate in the system that was designed and constructed for them. It's, it's just bizarre. And then, and then I talk with the activists that literally have nothing, you know, that, that I was, I was driving around in New Mexico with an activist who, her car has, you know, damaged windshields and it looks like it's 20 years old because she just can't afford to get a new one. And she was talking about how she's purposely put her kids in a public school in a way that she knows they're getting a shitty education from in part because it's really important for her to continue being committed to the improvement of public institutions. And those are the types of people that have the most agency in the world and are like, we can change things even though they exist in this power structure at the bottom. And so it's just been, you know, as I've traveled through these circles up and down the power ladder, it's just been so remarkable to me how people craft stories around themselves to protect themselves from their own moral failings. [00:24:37] Speaker 2: Yeah. A good analog of that at the level of geopolitics for me anyway, is you take a country like Britain or France, which are middle great powers or Canada, although they're sort of changing their tune a bit with Mark Carney and they had the same learned helplessness, right? Yeah. And then you have a country like Iran, which has been isolated 45 years, but it still has a sense, whether or not it can do anything, it still has a sense of agency. Yeah. And it's fascinating where that comes from. I presume it's just an outgrowth of ideology and not just subordinating yourself to this kind of, this idea of the end of history. And I think there is this weird confluence of the ideological hangover of the end of history, the end of ideology. And then like you say, some people got really rich out of that and they kind of want to pretend it's still the case. Yeah. What's the strangest interaction you've had since the book was published? Oh, strangest. And also how's your life changed? You know, you've become something of a figurehead. I don't want to sort of go OTT because like I said, it's all about the activists, but you've become something of a figure out of this movement, at the very least AI skeptic. So have you been sort of accosted by, you know, a vice president of Meta or something? You have no idea what you're talking about. [00:25:51] Karen Howe: Um, I think this, maybe it's not like one particular interaction, but one of the things that has been a little bit bizarre for me is the degree to which people seek me now for therapy, essentially of a form of therapy. Um, which on one hand is an enormous privilege because they are opening themselves and being their most vulnerable selves in front of me. Um, and on the other hand is really concerning because I, I have, I've realized as I've gone from community to community that people are often not doing okay. You know, they're dealing with so many intersecting forces. They're dealing with the poly crisis, you know, at the most intimate levels of their lives. And they reach out for people like me to say that things will be okay. And I, I find it, um, yeah, I find it just to be like a really intense responsibility to bear when I have these interactions where in, you know, a one minute span, I am trying to convey to them things will be okay. And you have agency and you have hope and we can do this together. And it's just a role that I, I, I really did not expect to play. Um, and so I think that kind of answers also the second part of your question, which is that, that there has been this really, um, this really weird feeling of, I, I really dislike when people are put on pedestals and I've really disliked the fact that people now put me on a pedestal and it has been a really, yeah, it's been a really intense experience. Like I, I, I am trying to figure out how to make, how to, how to not be an avatar for this message and how to make everyone feel co-ownership over these ideas that I, I presented in the book because they ultimately weren't mine either, right? They were, they were from so many other academics and journalists and critical scholars that had been doing this work for a very long time. And I'd sort of just collated all of it and put, um, put, put some words to it and some detail and texture to it by documenting communities that were experiencing the very phenomenon that these people were studying. Um, and I just, I, I hope that people can continue to feel co-ownership over these ideas and this resistance movement and make it their own. [00:28:31] Speaker 2: Can capitalism be fixed? That's the kind of question you have really strong opinions about, isn't it? No, yes, of course it can. Capitalism is the greatest system ever. It's the source of all evil. It makes people have very strong feelings. It's a question that we will be putting to Mariana Mazzucato at a forthcoming event on June the 22nd at Earth Theatre in Hackney. Before that conversation with Mariana about the future of capitalism, the pitfalls, the possibilities, my colleague Ash Sarkar will be talking to the former Shadow Chancellor of the Exchequer, John McDonnell. Like I say, June 22nd in Hackney, the sun will be out. Huge, important, intriguing question. Grab a beer. That's my idea of fun. How are you doing? Because you're saying that people are coming to you saying, are you okay? I mean, your disposition seems very happy and cheerful. I mean, why wouldn't it be a beautiful day? Climate change is real. I'm not suggesting otherwise. The weather is lovely. We're built to be cheerful under such circumstances. So how are you doing? You seem fine, but it must be quite a weight to bear. And people do have this heaviness, right? You said with the poly crisis. I mean, I say there's something on the left. We're a left-wing media organization, you know, and sometimes we'll do a week's worth of shows. And I think, Christ, there is some good news out there. You know, there is some good. There's actually lots of good news. It's just, we haven't happened to be relaying all the bad stuff. Do you ever feel that you're doing something similar? I mean, yes, but I don't really think it matters [00:30:04] Karen Howe: how I'm doing because, um, yeah, like this is, this is, there are, there are so many mobilizations and resistance movements that are happening around the world that it doesn't really matter what one individual is feeling. Like the beauty of the fact that there are so many intersecting movements now means that we can all take care of each other and people can step in and out of these movements and resist and hold these companies accountable when they have the capacity to do so. [00:30:39] Speaker 2: I suppose it's just a very big weight to bear. And this, I mean, this is true, right? AI is probably the most disruptive technological force since the steam engine, maybe more, maybe, maybe more so. Sorry to be doing the advertising for open AI. I sound like Sam Altman. And it's obviously going to have these huge political implications. And, you know, I think for most people, they tread in, they tread out and, you know, but you're, you're constantly in it. So does that in any way sort of shape how you feel? You think sometimes, oh God, actually, I'm being, I'm being a bit pessimistic about the world generally. I'm being a bit blackpilled generally because I'm focused so much on without being sort of hyperbolic. I mean, we're talking about some pretty uniquely malevolent people, people like Peter Thiel and whatnot. It doesn't seem to affect you, but I thought I'd ask. [00:31:22] Karen Howe: I guess what I'll say is when I finished writing the book, I was deeply depressed. I was really in despair. Like I was like, how are we going to get out of this situation? And I had to go on a journey because I was like, that is not, you know, the message that I can deliver to the world and be like everyone to spare. You know, it's, it's, it's really bad out there. I had to go on a journey myself of, okay, what do we do about it? One of the books that I read that was so profoundly helpful for me in this moment was Hope in the Dark by Rebecca Solnit. It's this really beautiful meditation on resistance movements, grassroots movements throughout history and the way that they actually are the sole forces of social and moral progress in society because that progress never comes from the elites, right? The elites have too much to lose. They like the world the way that it is. That is what allowed them to be elite. And she talks about how it is people, everyday people engaging in the, even the smallest acts that actually help society move out of its darkest moments. That's literally how we've always done it. And your job every day when you wake up is to make sure that you are pushing in the right direction. And you might not see the effect of those actions, even in your lifetime, but you know that in the absence of you pushing in that direction, there are other things that would fill that space and could continue to push in the opposite direction and recede us backwards. And so you don't, you know, it is, it is, she, she says in, in this much more articulate way that it is almost selfish to expect that as you are trying to enact change, to expect to then enjoy the fruits of your labor. That's not the point. You're, you're moving forward in a direction for maybe you will actually get to see the outcomes, but maybe the outcomes will ultimately be enjoyed by future generations. And that doesn't mean that you should stop. And it doesn't mean that just because you're not seeing anything happen, that what you're doing is useless and futile either. And so that's all you can hope for. You just, you do the work every single day and you, in the, you find joy from the moments of connection with the other people that are doing the work and, and the moments of joy and the moments of shared, shared failures as well, because that is also part of what you are fighting for is that interaction and, and, and that beauty. And so when I read that, I was like, there's so much hope that we actually still have. There's so much optimism. I mean, we've, I know we are in a, in a, in a deep mess right now, but we've been in really deep messes in history and we've gotten out of them because of the very actions of individuals and grassroots movements and communities. That was part of the reason why I then started doing the AI Resist List project with other people that were thinking about these same ideas, because I wanted to see what are the actions that people are taking? You know, just, just some, we, we document like really small actions as well as really big actions because we wanted to recognize and acknowledge that no action is too small to have a difference as long as you are doing what Rebecca Solna is doing, like pushing in the right direction. And yeah, so now I am genuinely in a good place. Like I am, I'm really optimistic and really hopeful about where we are traveling because of the number of people that I've seen in the last year come and join as fellow travelers. The temporality you're talking about of, [00:35:25] Speaker 2: well actually this is bigger than me, is just the complete reversal of what we see from people like Musk, Altman, etc. And I think this is really interesting. Have you, have you heard of a book called Good Ancestors? No, I have not. Yeah, it's by a guy called Roman Krasnich, I think, that's his name. He's a very nice man. And you know, he, he says that basically we need to start thinking about politics through the prism of being good ancestors and through sort of deeper time. So we need to think back to our grandparents and forward to our grandchildren. This is quite a convenient time frame. It's about 200 years, if you think about it. And you know, that's, that's the prism through which you should probably understand change, good or bad. And I think that's a really nice way to, to, to view it, particularly through climate change, but also I think AI is another one as well. And the thing you said about, and this is a bit of a detail, I think it's really important because so many people come to these conversations about AI, climate change, inequality, and they just, it's so doomerish and they're so blackpilled. Yeah. I do think it's important to talk about it. You know, there's a book by a guy called Stefan Zweig, Memoir of a European. Have you heard of this guy? Really? I mean, I only read it a couple of years ago. He's a Jewish Austrian writer. You know, he's a best-selling novelist. He lives through the collapse of the Habsburg empire. He, you know, there's the first world war, there's the second world war, obviously has to flee Europe. And in 1943, he writes Memoir of a European looking back at the last 50 years of European civilization. He thinks it's over. He shoots himself in Brazil the day after finishing the manuscript to this book with his wife. Oh God. Yeah. Right. But it makes sense because he's a Jewish Austrian. You've got Anschluss. Hitler controls Europe. You probably have some inkling about what's happening to European Jewry. Yeah. He thinks it's over. And then of course, 60 years later, we're in a very different world, but I think significantly better. And it's, yeah, I think this is a really important point. It's never over. And I think, you know, that expanded sort of period through which to understand history is really important. You know, the people that came before us were so wonderful. They gave us such an amazing planet. Yeah. You know, all the people that cultivated grains and husbanded animals and built all the infrastructure. We just take it for granted. They were wonderful people. They were so, and then, you know, what are we giving our grandchildren? And that's the question. And I, you know, I think that's also the question you want to throw up Musk and Altman, et cetera. What are you giving to our grandchildren? I think they're screwing them over personally. On Sam Altman, I've already mentioned Musk. These companies, XAI, SpaceX, and OpenAI are going to be IPO'd this year. Yeah. You're the guru on OpenAI to some extent. What does that mean for the company in terms of how it's going to run? Is it going to become less utopian? Is it going to become more hands-on, more short-termist? What's the... It's a great question. [00:38:09] Karen Howe: I mean, I think that when it becomes a public company, there will be a certain level of professionalization that the company will have to undergo. I mean, you know, I don't remember if we talked about this, but one of the common descriptors that people would give to me when I was interviewing sources for the book was chaos to describe the inner workings of OpenAI. You know, decision-making often didn't have a process. There was just a lot of whiplash for employees about what projects they were ultimately pursuing, what deadlines they were ultimately pursuing, and what the priorities of the company even were. There was a lot of disagreement, a lot of clashing, so on and so forth. And so being a public company, I think it will be more difficult for OpenAI to have a lack of sound governance procedures. That said, I don't think it's going to fix any of the problems that I talk about in the book because, of course, Google, I also identify as an empire of AI, and Google is already a public company. And you could argue has even more tentacles and more leverage over different facets of society than OpenAI because not only is it an AI company, it's also a search company and is now using its power as a search monopoly to funnel people into its AI products and make it the portal through which everyone is interacting with the world. And so from that perspective, you know, I think OpenAI will shore up some things. There might be some transparency mechanisms that we gain from it being public, and yet it will still continue to plod along this really aggressive, reckless expansion unless everyone else [00:39:55] Speaker 2: does something about it. And with Sam Altman, I know we did cover this in the first conversation, but I think it's worth just going back to some of it. With Sam Altman, is this kind of shambling [00:40:04] Karen Howe: chaos, is that an intentional strategy to get what he wants? It's so interesting because there was one person that I interviewed that did say that there's a lot of theories about this chaos and Altman's relationship with it in particular. One person did say that they felt that they believed Altman engaged in chaos as a strategy. It was a kind of survival of the fittest type situation, create a lot of competition. And then the ones who rise to the top are the ones with the best ideas. And that's what will guide the company's direction. Another person that I spoke to had the theory that Altman, his skill is that he is able to persuade people to do things, whether that's investors to give him money or employees to join his cause. And because of that singular fundraising ability, he himself was an investor in startups, he actually never had to make compromises. He literally never had to engage in trade-offs because he just could raise more and more and more money and funnel it into more and more and more of those startups. And so that person had the theory that the chaos to rise from the fact that he just he's never exercised the muscle of actually making hard decisions. And so he has everyone work on whatever they want to work on and then tries to maintain all of the projects and has this cover the waterfront type strategy. But then, of course, that's not actually feasible in a company scenario. It's very different when you're talking about an investor investing in a portfolio of vastly different types of companies, but within one company where you need a unified strategy. That just doesn't work. So, yeah, you know, is one or the other true? Maybe it's both a little bit true. And they intertwine in these ways [00:42:11] Speaker 2: that create the instability. As a corporate leader, how has he changed since we last spoke? So, you know, Stephen Witt wrote this great book about NVIDIA and Jensen Huang, super interesting. And you talked to lots of people and say, look, we can slag off the Silicon Valley people, but Jensen Huang, he worked in a Denny's. He seems like he comes from a reasonably normalish background, even if he's lost lots of it. You know, his hobbies on the side are kind of like barbecuing steaks with his wife or whatever, you know, and that's not all just PR fluff. That's like a real thing. But you read about Jensen Huang and he seems like a very competent corporate leader. You know, he could have been a car salesman and he would have had, you know, 20 showrooms or something. He's very competent, very high competence guy. That isn't the picture you get from Sam Altman. But he's now leading this, you know, potentially trillion dollar company, potentially in the future. How has he changed? And is that a fair representation of him? [00:43:03] Karen Howe: I mean, he says himself that he's not the kind of leader that is good at operations and and at actually doing the daily business of businessing. You know, he's a vision setter. That's how he describes himself. He's good at setting a vision and then bringing a bunch of people in a bunch of capital together to make it happen. That's his skill set. And he said many times in conversations, he's not actually sure that he likes running a large company because OpenAI just grew so fast from a startup into, you know, it now has thousands of employees and it has offices all around the world. And yeah, he was like, I never really imagined myself being the kind of person that runs an organization like this. And so you could see, you can see in the past year, OpenAI has tried a lot of different strategies basically to continue keeping Altman as a figurehead. But distancing him a little bit from the day to day operations, they brought in other more experienced executives around him to try and do the actual business of operating. Many of them have also now left. So there's continuing to be this rotating cast around him. And I think in the public, there is this really strange change in how Altman has operated as a public figure where for the longest time he really benefited from his remarkable persuasive abilities. And that included being able to persuade the public about a particular narrative about him and also the company. He has now lost control of that narrative very much. [00:44:39] Speaker 2: Of his image, so to speak. [00:44:41] Karen Howe: Of not just his image and the company's image, but of just AI in general, you know, so much of his power and his image was predicated on the idea that AI, as he was designing it and his company was designing it would bring benefit, broad-based benefit to society. And he's lost that narrative in part because reality has hit. And people are now experiencing the impacts of AI in their lives, not just as consumers, but also as residents that live next to data centers, as workers that are dealing with the looming threat of potential layoffs. And people are realizing this, this, this is not actually what Altman said it would be, what any of the Silicon Valley elite said that it would be. And I think it has neutered his key lever of power. He, because he cannot control the narrative anymore, because he's unable to persuade people beyond the reality that they see. It has made him a lot more defensive in public. You see his engagements and he is, he, he's losing his rhetorical footing. [00:45:47] Speaker 2: Elon Musk, you know, we've said that he's going to be maybe the world's first trillionaire. He's talked about putting data centers in space. I remember when I first saw this story and I thought, I can't wait until you can now, because I can ask her about this, uh, because people have reached strong opinions one way or the other, right? As fanboys on X again, this is amazing. And then other people are saying, no, it's technically impossible or maybe it's somewhere in between it. Is that a thing? Could, could they do that with the present technology? I'm not about 20, 30 years in the future or, [00:46:18] Karen Howe: I mean, yes, could they do that? I mean, I'm, I'm sure that there are already plenty of computer servers in space because of the ISS and you know, the, every time you put a rocket up there, there's, it's running on all these computers. So how big is the data set? I mean, you could say that those are already data centers in space. And so if you imagine just scaling it a little bit bigger, yeah, you could have more data centers in space. But the problem is like, what is this actually solving? Because they're saying that if you put the data center in space, then it resolves the problem of needing the land and you don't need to cool it because space is cold. But you know, I, I'm not a physicist, but I have talked with, I've talked with people that have said that the, the physics actually doesn't really work out. It's putting it in space just to use the coldness of space is not quite actually sufficient for dealing with the complexities of cooling a data center. But the other thing is, I imagine that they're trying to put them in space to make it less expensive to develop these things. And how is that going to make it less expensive? You know, they, they, anytime, one of the challenges with data centers on earth is, you know, AI researchers tell me because they're so large now, they're running hundreds of thousands of chips and some of them are being currently built to eventually have a million chips, computer chips. [00:47:44] Speaker 2: In one data center? [00:47:45] Karen Howe: In one data center. So, so these data centers, they're actually campuses. They're not just one building. They're now networked buildings on a campus. And they're still called one super computing facility because they're connected underground through a bunch of wires that it's like high bandwidth connectivity. So you're able to transfer data really, really quickly between all of these facilities. And the, one of the challenges of having that many computer chips networked together is that one computer chip fails and everything starts to get a little bit wonky. And so sometimes, you know, AI researchers were literally telling me sometimes they get results after training a model. These models are trained for like months on end on these facilities. And it's just a weird, the result is weird. It's unexpected. And they don't know if it's because the AI model has suddenly developed new capabilities that might mean something or if there's a computer chip that failed. They just have no idea. And so the, the, you know, the people have, have talked about how AI research is not actually a science in the traditional sense because you're not able to replicate your results. And one of the reasons why they're not able to replicate their results is because the hardware that they're using is unstable and they don't have the tools to, to pinpoint exactly what led to a weird result. And so now imagine putting this already slightly unstable, faulty technology that underpins the very things that they're trying to produce into this, into space. The maintenance is going to be insane. And, and the ability to actually trace what failure modes are happening in the data center to result in strange outcomes with your model training will become impossible. So it's just, yeah, to me, it makes absolutely no sense for multiple reasons. [00:49:44] Speaker 2: That's a really good explanation. So even if it was cheaper, the maintenance costs would be much higher, at which point the Elon sort of fanboys say, well, there'll be some, you know, Mecca robot or something. God knows what they'll fix. Maybe, but you know, that sort of fine motor coordination is a long way off, 20 years away, let's say. But this is the kind of classic Elon thing, isn't it? He might put one in space. Yeah, right. And then, and then he'll say, look, XAI, SpaceX, we're building the future. We have, you know, off earth, off world data centers. And it's just kind of one crappy one that is poorly maintained. That's probably where it ends, right? For now. [00:50:20] Karen Howe: Yeah. I mean, it's like, it would be like, I think what's going to happen is similar to the boring company. [00:50:25] Speaker 2: Exactly. Exactly. Right? Yeah. I love that. In the US, do people say that seriously? [00:50:32] Karen Howe: The boring company? Yeah. I have not met anyone that takes it seriously, but I don't know. Maybe there are people out there that [00:50:40] Speaker 2: take it seriously. When I see that thing in LA where there's like a tunnel with cars and I'm just like, just metro systems. I mean, they've been around for 140 years. [00:50:48] Karen Howe: It's just like the privatized stuff. Yeah. [00:50:49] Speaker 2: It's crazy. Anyway. So we've talked about Elon. We've talked about, sorry, Elon Musk. It's like Boris. Elon Musk, Sam Altman. How about the good guys? Anthropic. They have this little halo of virtue around themselves, which is not normally ascribed to the other ones. Is that true? I mean, [00:51:08] Karen Howe: from your response, I presume not. To me, OpenAI is coal and Anthropic is clean coal. That's the way that I see it. I mean, if you look at the way that these two companies operate, they're fraternal twins. They engage in the same labor exploitation, environmental extraction, data extraction. They, I mean, literally just recently, one of the most infamous examples of, of the impact that data centers can have on a community is Elon Musk building Colossus supercomputers in Memphis, Tennessee. I think we might've even talked about this last time. And, and in order to power Colossus one, XAI wields 35 unlicensed methane gas turbines to the facility to make it run. That has huge climate implications, of course. That also has huge air quality implications. It's pumping tens of thousands of toxins into this community. And I, I actually recently met one of the community leaders that is leading the resistance against XAI now, Kashawn Pearson. And he was at this event that I was attending and, and, and he says to everyone, close your eyes, take two deep breaths. That is a human right that's being taken away from us. And in a community that's dealt with a history of environmental racism and, and now to this day has a cancer rate that's four times higher than the national average. Guess who recently partnered with XAI to use Colossus? Anthropic. So Anthropic is now going to be using this data center, the most infamous example of data center impacts on a local population and on the climate. And this is exactly why I say, you know, that they're the clean coal. It's like, they are engaging in the same ideologies, the same practices, scale at all costs as open AI. And it's not a coincidence because Dario Amadei was an executive at open AI. And he was one of the key people that set open AI down this scale at all costs path. And then he essentially just copy pasted it to Anthropic. Does it, does Anthropic on the margins iterate in different ways, the way that, you know, clean coal fixes things, tinkers things at the margins to try and clean up some of the air before it comes out of the coal plant? Yes. But is that actually addressing the root problem? No, it's just perpetuating it further by making people think that now it's okay. [00:53:55] Speaker 2: So we've done the three kings, the three wise men. That can be a chapter in your next book, the three wise men of AI. We've done Amadei, Musk, Sam Altman, who all know each other, often hate each other. [00:54:10] Karen Howe: They all hate it. Yes. Well, I guess Dario Amadei and Elon Musk are being buddy buddies now because they both hate Sam Altman. But, you know. [00:54:18] Speaker 2: Isn't this, this is another crazy thing, right? It's this extraordinary disruptive industry, multiple, I think ultimately, trillion dollar companies have come out of it, maybe multiple trillionaires. And they all, it's kind of like river, it's like, [00:54:31] Karen Howe: yeah, it's like a school girls. Yeah, yeah. Or school boys. I don't mean to be sexist. School yard. Yeah. School yard where people are having petty fights. Yeah. [00:54:39] Speaker 2: I mean, I remember seeing Elon Musk on the Joe Rogan podcast, and he was saying that Sam Altman doesn't really care about, because Sam Altman was saying, have you seen this clip? [00:54:47] Karen Howe: I've not seen this clip. [00:54:49] Speaker 2: So it was the one time where I thought Elon Musk actually showed quite high emotional intelligence, because he, because Joe Rogan was saying, but hold on, like, you know, Sam Altman is saying that we need this stuff to serve human need, and, you know, we need alignment and all these things. And Elon Musk was like, yes, all bullshit. And then he was saying, really? He goes, yeah, he just wants to get rich. He goes, really? He goes, he says, he says to the producer, he says, pull up Sam Altman and like the Konigsberg, you know, the world's like most expensive electric vehicle, I think. And he pulls it up and it's like a $600,000 car. He says, you think that car doesn't care about getting rich? And I was like, wow, Elon, this is like the one time, it's like, really savvy, wise, you know, thing. It's funny. They all know each other. There's shit talking to each other. So we've got these people, we've got the interest and the ideology that powers them. I want to talk about how this feeds through to politics. We've talked about activism a little bit. We'll come back to that. But sort of establishment politics, lobbying, campaigns. We've had Tony Blair in this country, you know, give another rare intervention, which he kind of does every month, talking about how bridge politics is going wrong. And invariably, he's talking about AI, right? Yeah. Which I just find so strange. We've got labor potentially going to double-digit MPs. We've got a complete collapse of two-part system. And this former prime minister is talking about why we need to build artificial intelligence in this country. It's kind of like surreal to me. Where does Tony Blair fit in all of this macabre world of political campaigns? [00:56:20] Karen Howe: And he just gave a speech, which I haven't actually watched, but one quote was relayed back to me, where he said, AI is going to change everything, literally everything. And we should stop debating whether it's going to be good or bad, because it's just going to happen. And there's nothing that you can do about it. I'm obviously paraphrasing. I find that so crazy that a former leader of a democracy is engaging, amplifying the very political project that, you know, we started this conversation off by by defining, which is taking people's agency away, telling them it's inevitable, telling them you literally can't do anything about it. And I don't know, I don't know enough about Tony Blair to really understand where this is coming from. All I know is that he's become buddy buddies with Larry Ellison because of some phenomenal investigative journalism that was done by Lighthouse Reports. Maybe is that sufficient enough to explain why he would engage in a statement like that? I'm not so sure, but I was really floored when that statement was read to me, that line was read to me. I was like, how could you not see that you are becoming part of this apparatus that these companies are [00:57:36] Speaker 2: constructing? So where does Larry Ellison fit in? Because of course, he's the chairman of Oracle, I think, right? He's not the CEO anymore. He's the chairman. He's quite old now, isn't he? He's like 80. [00:57:45] Karen Howe: Where does Oracle fit in all of this? I mean, Oracle was when Trump in early 2025, at the start of his administration announced the Stargate initiative, which is a $500 billion at the time. I don't know what's going on with Stargate now. It was a $500 billion investment into computing infrastructure for OpenAI. The three figures that were standing up at the podium with him were Sam Altman, Larry Ellison, and Masayoshi-san, the founder of SoftBank. And yeah, so Larry Ellison is deeply embedded within within this project, this political project and this Imperial project, because he is providing data centers to companies like OpenAI. It's also, Oracle's also not doing well financially anymore as a company because it's taking on an extraordinary amount of financial risk to be involved in this. And there are a lot of other data center developers that are now trying to eat its lunch as well. But that's where Larry Ellison is situated in this ecology. Okay, so that's Blair. And you might be thinking, [00:58:51] Speaker 2: why is he, this is so parochial of me to talk about the one British politician, but he does these stupid essays and everybody talks about him for days. So I thought I had to mention Blair. But it's particularly in US politics, isn't it, that AI companies now are just flooding certain campaigns. Tell me a little bit about the relationship between these interests in AI, elite big tech, PACs, and the democratic process right now in the US. Yeah, I mean, one of the things that has been [00:59:19] Karen Howe: happening is the AI industry has both learned from the crypto industry and recognizing that they need to play politics, they need to use their financial might to try to distort the democratic process in their favor. And they have also realized that in this particular moment, they need to play hardball because they are losing the narrative, and they are the public is against them. And so you've seen the spinning up of multiple super PACs, one of which is called leading the future, it has amassed, I think at this point, 100 million over 100 million dollars, and 25 million is from Mark Andreessen, 25 or from Andreessen Horowitz, 25 million is from Greg Brockman, the now leading product development at OpenAI. But, you know, generally, you could think of him as just like the number two at OpenAI. And they are pumping money, not just into campaigns and picking, trying to pick the winners and losers in political races. So one of the losers that they've tried to tank the campaign of is Alex Boris, which is a representative in New York. And one of the winners is this other representative in Texas that they're trying to funnel a bunch of money into. They're also engaging in dark money campaigns. So when you pump money into the political process, there's rules and there's transparency. You have to actually declare how much money to whom and, you know, and leading the future was actually overly transparent and even stated straight out exactly what they were trying to do. They were like, we, if you are against us, we will seek to destroy you. And if you are for us, we will give you a huge boost to your campaign. But with dark money campaigns, what they're doing, there was recently a story out of Wired written by Taylor Lorenz about how they're funneling money straight to social media influencers to try to shift the public narrative as well. So they have been trying to get influencers, and this is only one of the campaigns that we know about, there might be more. They're trying to get influencers to chat up this idea of how American AI is, first of all, really, really good. And second of all, we need the US to lead on this technology. And then they're trying to get influencers to talk about how Chinese AI is really big and scary. And that's part of the reason why we need more American AI. And some of the documents that they've been giving these influencers are incredibly detailed about saying, we would like you to project images of you sitting at your kitchen with your kids and getting them ready for school while talking about how much you love using AI. And so they're doing, and when Taylor Lorenz gave Leading the Future a request for a request for comment, they just said, we are engaging in a whole of communication approach because this is, we cannot let progress be stopped. We must continue to proceed. And there was another interview that, that Greg Brockman recently did, where he's also become a top political donor to Trump directly, not just through leading the future, through a different super PAC that Trump spun up. And when he was asked about it in an interview, he was like, it is the most moral thing that I can do because, and I, and I, my interpretation of it is all of these people within the AI industry, they see this progress, the advancing of AI capabilities as the fundamental good thing to do in the world. And anything that is going to stall that progress is immoral. And anyone that can assist in that progress is someone that they are obliged from a moral obligation to support. And that is what he, Brockman was getting at with saying that is the, is, you know, like the right thing is the, it is my moral responsibility to donate money to Trump because Trump has, of all of, you know, the political leaders been the most effective at supercharging the ability of Silicon Valley to do what it's doing. [01:03:44] Speaker 2: I mean, that's religion, isn't it? You know, rather than saying, you're right, I'm wrong. I'm right. You're wrong. We can disagree. And that's okay. Maybe one side can be persuaded. It's I'm good. You're evil. You're a sinner. We're the anointed. There's a high purpose we're serving, which is, which is quite terrifying. So they're funding these campaigns and they're trying to basically smash certain kinds of politicians. And you know, that's in the US. I mean, what's this, what's the story outside the US? Because obviously there are huge global markets for this stuff. And I come back to the UK, obviously Blair's a really high profile example, Oracle giving the Tony Blair Institute a ton of money, hundreds of millions of pounds. But recently I saw Claire Coutinho, who's a former government minister in this country. And she was talking about, this is going to sound so strange to you. She was talking about AC, right? We don't really have AC in this country very much. And you might have noticed. And she's saying that there's a de facto ban on AC, which is kind of true, which is kind of weird for an American. I'm sure it's particularly weird. Or even somebody who lives in China or actually most of the world, it's kind of weird. There's a de facto ban on AC and she's saying, we're going to end the ban on AC to help people and AI. And it was this weird, and I was thinking, is that because they need to call the data centers or is somebody, and I don't want to get sued, is why talk about AI? And there have been lots of these little moments where I'm seeing relatively obscure politicians just talking about AI. And I wonder what's going on. Are they kind of now like quasi-influencers? And is that money coming over here? So that super pack you're talking about, for instance, Marc Andreessen. I mean, all these companies, whether it's Meta, Alphabet, they have huge London campuses. London is obviously a major city now for AI alongside Toronto, New York, Beijing. Do you think that money's kind of seeping out to these kind of influence operations elsewhere? It is entirely possible. [01:05:42] Karen Howe: I don't know. I mean, also, it could be that some of these politicians that are feeling a decline in their political careers are trying to leverage AI as a way to slingshot themselves back into the center of the conversation. In that particular case, I have no idea. That's really bizarre. [01:06:03] Speaker 2: Search it. I swear to God. Claire Coutinho, AC and AI. [01:06:09] Karen Howe: Yeah. I mean, it's bizarre. But you are right that there's absolutely a lot of money lobbying influence getting pumped into places outside of the US. I mean, the EU has a huge target. Brazil is a huge target because these are two governments that have then cascading effects on governments around the world, particularly in their regions. And one of the things that has been happening recently with the EU is the EU is under this very strange self-doubt period, self-consciousness period, where it's wondering, why don't we have our own tech giants? And I find this really bizarre because I'm like, it's because your regulation worked. You know, like, do you want these really corrosive, exploitative tech giants on your soil eroding European rights the way that they're eroding American rights? I mean, that seems like a race to the bottom and something that you shouldn't be be envious of. But it is in this moment of self-doubt for Europeans, there is this ability now of the tech lobby to have quite an effect. And so there have been these stories showing that the EU commission has been discussing whether or not to change some of GDPR so that people's personal data can be used to train AI models. [01:07:34] Speaker 2: Oh, I didn't know that. Yes. Wow. [01:07:36] Karen Howe: So when was this? This was just a few months ago that it came out that this was under discussion. There's also a story that came out in LeMond from a collaboration with Investigate Europe that found that the tech lobby, Microsoft lobby and Brussels Europe, I believe, was the tech lobby organization, successfully introduced language wholesale into a draft with a bill that the EU commission is working on that, you know, that the EU is once again kind of becoming vulnerable to the ways that the tech lobby is functioning. But it is a product of the fact that Silicon Valley is investing an extraordinary amount of money in doing this and sending lobbyists there. I mean, I was talking with an MEP who mentioned that not only do they send them to MEP offices and lobby in the traditional sense, they also do this thing where they will plant people at every single gathering in Brussels that is talking about a particular issue, like within the audience, to then ask a particular question about that issue to mainstream a particular frame for how to think about the issue. Wow. Yeah, like that is the degree to which they deploy human resources across a landscape to seed certain ideas and shape the conversation and get onto the agenda of policymakers. It's like nuts. [01:09:20] Speaker 2: It's like a revolutionary political organization. Yeah. No, seriously, that's what the Bolsheviks would do before 1917. You know, they would, if there were public meetings and Soviets, they would try and sort of shape the [01:09:29] Karen Howe: Yeah, yeah, yeah, yeah, yeah, yeah. Planting questions. They, they, they do this thing. Yeah. And then, and then, you know, like there, I have some, um, there's some incredible investigative journalists in Brazil that were also telling me about the ways that Opening Eye has been operating in Brazil, where they've placed the very first policy person in Latin America was placed in Brazil. And part of the reason is because the Brazilian government has shown, has demonstrated their willingness and their appetite to straight out ban a platform like X, if it's not actually complying with Brazilian laws. And there's this facilitation of the lobbying by the US government itself, where the US government will then invite Brazilian policy makers to come tour in Silicon Valley and see all of the amazing innovation that's happening so that it helps to grease the wheels of the conversation of the, the lobbyists of OpenAI on the ground. So, so there's all of these like crazy, crazy things that are happening that these companies orchestrate in order to bend the political and geopolitical landscape to their will. [01:10:40] Speaker 2: Yeah. I'm in public affairs. So lobbying, political communications, strategic communications, blah, blah, blah, blah, blah. It's always been a, you know, a smaller gig in London as in Washington or Brussels, right? Obviously, but it's always been, you know, it's a decent circuit. And what's been really interesting in the last 20 years, it's become just, it's just become cannibalized by big tech. All of a sudden the, the most jobs, the most money is big tech and it's just gone on steroids with AI. So you've got lots of people, you know, former political advisor to the, the deputy prime minister or something. They, they can't, they want a job. They just want a job, right? Like anybody else got to pay their mortgage and they find 120,000 pound a year job with Google. And it means you have to go to Kent and tell a bunch of people that actually our data center in their local area is really good. Yeah. And it's like an Ouroboros of the more valuable these organizations are, the more political influence they can exert, the more value they get, et cetera. And, and, and the job that you're trying to do here is, is to break that chain and Ouroboros by the way is the snake that you don't tell. So you said, um, a while back about the AI resist list that you've been working on. Yeah. [01:11:40] Karen Howe: What is that? So we, we, we documented all of these different projects around the world of people actually resisting the AI empires because we first wanted to help people understand that resistance is, in fact, happening. I mean, one of the things that I started doing at my, at my book talks is I ask people, how many of you don't use AI? Because I'm so tired of people saying, well, everyone's using AI anyway, as though it's a foregone conclusion. And I'm like, actually, I'm not sure they are. And I was doing this in Toronto, which of course, as you mentioned is a huge AI hub and 30% of the audience raised their hand and was like, yeah, we're not using AI at all. At least these generative AI tools as, as developed by, by the empires. Um, and, and so one of the things that we wanted to do with AI resistance is to simply, to simply demonstrate to people there is resistance happening. Like this huge trajectory and all of these powerful forces that are pushing in a particular direction are actually being challenged by many people around the world. The second thing we wanted to do is then show how diverse that resistance actually is. I mean, we, you know, one of the examples is simply, uh, a, a little or organic campaign that cropped up where people started using the term microslop on social media to refer to AI slop from Microsoft. And this is just, even these small linguistic subversions begin to seed certain ideas in the public's mind about how to think about these technologies. And so we have all of these different types of examples from, from those things to what I was talking about with the Chilean project, Keeley.ai to, um, you know, technical interventions. We, there's this example of a university of Chicago professor named Ben Zao who developed this tool called nightshade. It helps to shield images that artists put online as part of their portfolio. And you can't see the difference with your, the human eye, the shielded images. But if a, an AI company then scrapes those images without that artist consent, and then tries to train, train an image generation tool, it begins to degrade that AI model. So it poisons the well of data and makes the entire AI model begin to collapse. And so this is a technical act of resistance. And so, yeah, we just, we wanted to help people recognize there's so much left to be done. There's so much much agency and going back to this idea from Rebecca Solnit, even the smallest action makes a difference because all of us are interconnected in a tapestry of resistance. And if every single one of us is moving, even in these small steps, and some of us who have greater capacity, greater energy to move in these big steps are working together, we create an enormous vector pushing in the exact opposite direction that the empire wants to go. And, and it's having an effect now because OpenAI recently shuttered its video generation tool, Sora. It originally announced Sora as the second best product since ChatGPT. Why did it shutter it? When you look at all of the reasons, every single one of them was influenced by grassroots action. The first one was they are enormously compute constrained. And in 2025, over a hundred billion dollars worth of data center projects were stalled because of data center protests, many of which were OpenAI data centers. So that's already a direct impact on the bottleneck that OpenAI is experiencing. Second thing, OpenAI is facing an uncertain financial future because it's about to IPO. It needs to shore up its balance sheet, which currently looks really, really bad. And investors in Wall Street are becoming much more concerned about whether or not AI companies can actually meet their promises because they're seeing this broad public backlash and a souring of the narrative. And so they're beginning to factor that in to their valuations. And that is what's also making OpenAI need to become much more efficient with their resources in anticipation of potentially not getting as much resources in the future. And the third factor, flatlining usage. That's collective consumer action. There's been a lot of collective consumer action recently protesting OpenAI in particular, sometimes not for the right reasons. But it is having an effect. This constellation of resistance is actually literally forcing a company as branded as OpenAI, which has been the avatar of the AI revolution to literally cut off an entire product line because of their uncertainty. [01:16:37] Speaker 2: I like this idea of like, there's a dividing line between getting productive value out of the stuff, like I said, getting recommendations of places to visit, which I think is good. There's some utility there. And things like Sora and image and video generation, which is just like you say, slop. And just the act of having that word in micro slop or slop, like you say, it's empowering in quite a surprising way, given it's so easy to do. Yeah. Karen, this has been fabulous. I want to get you on a third time. You said you do no second interviews. One more year. Yeah. Maybe the next book we'll do an event and we'll get a lot of people through, I'm sure. It's fabulous to see this in paperback. It's going to do very, very well. Thanks for joining us on Downstream. Thank you so much for having me. [01:17:28] Speaker ?: Thank you.