“If we don't have at least some sort of public consultation through LLMs right now, what we're going to get is the tyranny of Silicon Valley CEOs and a few experts and maybe a few government heads around the world. Is that really the best we can do? I don't think so.”

Well, I think for me, the next step is to think about the corporate reform, corporate governance reform. So it's about who has a say at the level of the corporation, at the level of the industry, et cetera. Because I think that another thing we're going to have to question is capitalism in its current format. I don't see how we can avoid a conversation about that. I have some thoughts that I never got the time to write down, but I do think that capitalism, as we define it, is too much of a 20th century construct where we lump together free markets and shareholder primacy. And my view is that we should finally cut the cord between those two things. You can have free markets and a form of multi-stakeholderism, where the ones who have a say are not just the shareholders, because I think it's a huge problem. I'm saying this also because you mentioned the voices of these young people around the world that are not heard, but this really was brought home to me at a conference in Oxford that I just recently attended. I met a young Kenyan man who worked for OpenAI or some of the other companies on labeling data for like a few cents an hour. And the job is so destroying. They're exposed to horrible things that they have to label. They're supposed to help train ChatGPT about how to skin alive a person so that then the model can learn how to not talk about that and not engage with that, but you still have to make a human being suffer through the horrible experience of envisioning what this looks like and what the steps would be. And this is not something we talk about in Silicon Valley enough. Why? Because the tech people who are there, and that goes back to my point about how experts are too homogeneous, they don't realize that's what they're asking people to do over there. The responsibility is diffused. So finding a way to connect all these dots and these questions and bring in the voice of the people, the young people who work in these data centers is really important. The other reality is the question of globalization. Like we're in the midst of a huge backlash against it. But it's still the constraint on what can be done to help these people, because that young man was telling me that if it doesn't do that, someone else will. And if it's not someone in Kenya, it will be someone in India. There are enough desperate people. There's enough of a global loom and proletariat, if you want, that there's no real pressure on companies to pay attention to these issues. So, yes, what kind of governments does that look like? I think there are models that are in place. You have so-called midbestimung co-determination in Germany, where the labor force has 50% of the seat on the board of representatives of the company. And it's not like unions. I heard a famous economist recently talk about how it's like unions are in power. No, they're not. The workers are there, qua workers, but also qua employees of the company who want the best for the company, right? So unions still exist outside with a distinct purpose of fighting for the interests of workers in the sense of like maximizing their wages. But it's different from being involved at the level of the corporation where you're an actor and not just an input in the production chain, right? So for me, this conversation also needs to be connected. And it's like it's an agenda for the next 50 years. I mean,

no, there's so much to go on there. And I too have been nerding out on corporate governance as a really fascinating angle for addressing some of these issues where, for example, I know a lot of my students are big fans of public benefit corporations as a model that they would like to see more companies embrace. Yet, if you go and you read the Delaware Code, who can actually contest the extent to which a PBC drifts from their charter? You have to have like 2% of the shares. It's not as if any Joe Schmo can suddenly contest whether a company is going in the wrong direction. And so looking at the actual incentives that are being created, as well as who can enforce those incentives, I think desires way more attention. Because when we look at a national government, let's say the U.S., we've got 400 million people. We know that we need a system of checks and balances. And yet, for companies that are having a billion plus users and may have a trillion dollar valuation, we, the external public, don't have any sense of the internal checks and balances that they have within their corporate structures. I was talking with a member of one lab who was explaining to me that they have this body that's responsible for making sure that certain uses and deployments of their tools adhere to their company's values. And I said, wow, that's really fantastic. So how many external independent auditors or OMSBAD persons do you have on that board? And the person said to me, oh, well, zero. And I go, huh, that seems a little problematic. Have you ever thought of inviting someone to sit on that board? And they said, oh, well, the meetings, you know, we schedule them in a rapid fashion. So I don't know if anyone has the free time to come and attend these meetings. And I told them, hey, give me a call anytime, and I will happily be on that Zoom call. And I know my dean will excuse me if I get that red phone ringing and I will go and I'll listen in. But we need to think about new ways to have independent voices on some of these critical decisions. And I'm not saying that board meetings should be broadcast via Zoom or YouTube or something like that. But there needs to be an opportunity such that there's friction in how and when some of the most consequential decisions are going to be made by these companies.

Yeah, and I think, too, there's so much to remembering first principles in these conversations. Again, I'm sure some listeners are at this point ready to throw their phone into the ocean. We've questioned capitalism. We've questioned Congress. They're just, wow, this is a radical episode. But if we go back to the founding and we ask the question about who is meant to be the ultimate sovereign, it's the people. And the people are meant to have control over how their delegated power is going to be exercised. And an often neglected fact of history is that we just didn't have a whole lot of corporations that were around in 1789. There were very few of them, and the few that did exist had to receive special authorization from their state legislature to perform very discrete tasks. And it's no longer the case that we have state legislators that are, for example, really paying close attention, really being watchdogs on the extent to which these corporations are adhering to their specified charters and purposes. And so this begs the question of what forms of new oversight can we introduce such that we do still retain that popular sovereignty and that popular control? Because as you and I have discussed ad nauseum, the people want to have mechanisms with which they can contest how and when this technology is going to be used. And something that I'm curious to get your two cents on is: I am increasingly worried that absent creating something like the global engagement you've called for, the global consultation you've called for, or even some variants of that in a smaller form, if we don't create those outlets for the expression of popular will and desires, the odds of far more political violence or upheaval or unrest are only going to grow because people will be frustrated by the absence of having a meaningful path to participating in these conversations. Are you at all worried about the sort of early signs of political violence we're already seen in countries around the world and especially with respect to certain tech leaders?

Well, yes, I think we've seen some attacks on like Sansa Maltman. It's looking like a sort of populist rebellion in the making. But I would say, even if we don't see that violence, I hate when we have to wait for violence to do anything. In France, too, we had to have the yellow vest movement and people breaking stuff for the government to finally pause and say, hey, maybe we should listen preemptively and start doing things like what we called then the great national debate or the citizens' convention on climate. Well, how about getting in front of the problem and opening up the system to those voices before you fed too many conspiracy theories, before you fed too many terrorist movements? Like, you know, the book by Stanley Kimley-Robinson, the Ministry of the Future, it's picturing this eco-terrorist movement. I could see the same thing with AI, like AI terrorist movement that start bombing data centers and taking down CEOs. I mean, it's just like we saw it for, you know, Luigi Mangione. It's there. It's like the Robin Woods of the future are going to be anti-AI for sure. So let's do something. We can do something. There are pieces in place that just need to be put together to create something that's more legitimate. You mentioned the sovereignty of the people. There's also the question of which people, because it cannot be just the American people, right? They themselves cannot constrain alone international corporations. It's too late. We've created these monsters that are more powerful than states, that can escape taxation now because they have different ways of moving around their assets, their investments, et cetera. So we have to reconquer that sovereignty as a global people. I don't see an alternative. It's the same thing. And it's nothing new. It's the same on climate. If you want Exxon to stop polluting, you cannot even count on American legislation anymore. You have to create some sort of global standard. So one idea that came and that is enforced, obviously, so it's very hard to imagine how that can be done. But when I mentioned this global assembly on the climate that took place in 2021, it was only 100 randomly selected citizens. What I liked about that pilot, which was flawed in many respects, but still very pioneering, they brought together 100 people selected at random on the basis of their geographic coordinates. So imagine they threw basically darts on a map and they picked people where they were a bit more complicated than that, but roughly that's the idea. So it wasn't based on nationality or anything. They brought them together. You had a Brazilian seamstress, an Afghan shepherd, illiterate, a bunch of people. And one idea that came out, which I thought was really interesting, was the idea of ecocide, creating and recognizing a crime of ecocide at the global scale. It's the same idea that the French had come up with during the French Citizens Convention on Climate in 2020. And of course, Macron, French President Macron, took down that idea and said it's unimplementable. If only France implements an ecocide, recognizes a crime of ecocide, which is when you kill the environment and you need to pay for it and clean up and everything, then we will be non-competitive compared to all the companies in the world who are polluting at will and at no cost, right? So that's the solution. You need to have a recognition of a crime of ecocide at the global level. And I would say not necessarily through international agreement and negotiations, which currently are being completely captured by the more powerful countries. I mean, I think that's more or less the diagnostic I'm reading from people who know more about this topic. So it needs to be done again through some sort of other process, other institutions, which would have true legitimacy. And then the question is: okay, how do you enforce these things? That I don't have a clear vision of how you do it, but we have to start somewhere.

Yeah, what a fascinating provocation because I was just reading before this some scholarship from Jillian Hadfield looking at how important it is for society and for cooperative action to have sanction and to have clear penalties. And it's compliance with those penalties and enforcement of those penalties that arguably do the most in terms of shaping norms. It's that fear that Adam Smith recognized as well as having that third party observing your behavior. And it's just that sense that you may be called out for perhaps violating societal rules that ultimately will shape your behavior and steer you in a more pro-social direction. And so naming what is The equivalent of something like ecocide in some of the AI domain questions is a really fascinating inquiry that I haven't seen a lot of people diving into.