The AI dividend starts with thinking about how we can give Americans a real right to an AI economy, and with the humility of not knowing what will actually happen. We don’t know how devastating it will be, but now is the time to plan for possible outcomes. And in my economy class on ILR, we always had this conversation. It was something like this: “Oh, every technological revolution has always created more jobs than it destroyed.” It may be debatable, but this is the first time someone has developed a technology and said that its goal is to replace all human labor. It is superior to humans in every way. And the criterion for understanding how good a technology is is how functionally it can imitate different forms of human labor. That’s exactly right. And exceed them. That’s exactly right. In other words, we are creating an alternative to the human labor machine. that’s right. And it’s my first attempt. That doesn’t mean it will be successful, but it does mean that the government needs to take it seriously. Here are some ideas for AI dividends: What would happen if we lived in a world where all human labor was replaced, or only a significant portion of it? So what would a truly functioning society look like? And we have to start talking about universal basic income. The idea is to make sure that we are building structures now that will protect Americans if we find ourselves in that situation in the future. And I have a lot of things about how we can prevent that future and changes and things like that. But AI dividends are more like insurance, and they can also be financed through something as boring as the much-talked-about wealth tax. Funding can also be raised through a token tax. So taxing the use of AI would probably be limited to commercial opportunities to replace human labor, and as long as investment in capital always leads to increased employment, it’s good policy. This is centuries-old economic theory. But maybe AI is changing that. So if that’s going to change, we need to change the tax system to tax AI and discount human employment, and token taxes start to go towards that. But the other funding mechanism that I’m talking about for AI dividends is actually acquiring warrants for these companies. Most of the warrants would give the government the right to buy shares at a set price if the value of the AI company increases significantly. Basically, you only make money if one or more companies are highly successful. Basically, it’s a case of replacing all human labor. And if you do it now, VCs will celebrate it and say you’re part of the upswing. And if either of them attempts to do so after they have succeeded, they will seize control of the means of production and control the wealth. So my idea is that you go down all of these paths and start finding ways to generate income to actually fund universal basic income, investment in job retraining, or just a broader safety net, but do it in a way that automatically scales and adjusts and starts at the speed of AI. That’s the concern I’ve always had about this set of policies, or set of answers to the question of AI and job loss. The trick to universal basic income to me is that you may be supporting it on its own merits, which is fine, but under the plausible scenario of job loss to AI, it’s happening to some people, not everyone. I know you’re being skeptical, but I can’t imagine a world where you wake up one day and find that everyone’s job is gone. It starts with some work. It starts with some work. So I wouldn’t worry if you think it’s going to be everyone’s job at once. Because then you just come up with a policy that compensates everyone. But you imagine you’re a Teamster and you drive a truck, right? And you make $80,000, $120,000 a year. And self-driving truck companies will put you and your fellow Teamsters out of work. And don’t worry, we actually passed universal basic income. No, not at all — and you’re now receiving $37,000 in universal basic income. 100%. And I get $37,000 from Universal Basic Income. And I’m still in the podcasting studio. It’s a mess. I received a check. What worries me most is the transition, even though I don’t believe we will go to a fully automated world, even if I believe we will. Some people will actually lose money, while others will be unaffected and make a profit. And I don’t hear any policy proposals that know what to do with the people who are losing out along the way, the people who are actually being displaced. While not everyone in the world has evacuated, the world is now more likely to graduate with a marketing degree. They are three times more likely to be unemployed than before, and programmers suddenly find themselves in less demand for their services. However, some programmers make huge amounts of money. Yeah. What do you think about the difference here? Universal basic income alone is not enough. And I’d love to understand why you think we’re not heading towards a fully automated world. Because once we start it’s hard to understand where it will stop. But we can come back to that. Either way, there will be a transition period. I don’t think it will happen all at once. So the idea isn’t just, “Oh, yeah, you’re right, we’re all going to get this basic income, people are going to be extreme over by it.” The idea is to do a lot of things at the same time, like actually charging for the use of AI and changing the tax code to discount the use of labor. And it’s a way to protect jobs and slow the transition itself. It’s not just an investment in universal basic income, but also in retraining programs and structures that help people get into new careers. Now, not surprisingly, their track record is very poor. This is my concern. That’s a really bad track record. But that doesn’t mean you shouldn’t invest in your community college and find ways to improve it as much as possible. But you’re right. It’s not enough to just say, “Oh, we just give a universal basic income.” We need to think of other ways to orchestrate that transition. This may include if someone has permits, training, or licenses that take years to obtain. Maybe you need it for 5 or 10 years to transition so people can turn that training into capital. And that’s another way they have a stake in the AI economy. Many policy solutions will be needed. That’s why the framework I put out includes 43 different ideas.
