Of all the things artificial intelligence has promised to revolutionize, OpenAI founder Sam Altman recently hinted at maybe the most audacious. On an episode of the “All In” podcast last weekend, Altman, during a conversation about AI regulation and social policy, proposed a “universal basic income” could be replaced by universal… access to his product. “I wonder if the future looks something more like universal basic compute than universal basic income,” Altman said. A quick backgrounder for anyone wondering what in the world he’s talking about: To a small but influential group of people thinking about the future of a super-automated society, a universal basic income is an obvious and simple way to solve the problem of a world with a robust, productive economy and very few jobs for actual humans to do. If companies are thriving but there aren’t many jobs, you can just pay out a stipend to everyone alive. Politically speaking, that idea is still mostly relegated to the realm of thought experiments and third-party candidates — but Altman is already taking it one step further. His idea, as briefly articulated on the podcast, is that when OpenAI builds a super-powerful “artificial general intelligence” (or AGI), everyone will get a small bit of deeded access to it. “[T]hey can use it, they can resell it, they can donate it to somebody to use for cancer research,” he suggested. “What you get is not dollars, but you own a part of the productivity.” The idea of replacing money, which is somewhere around 5,000 years old, with access to a large language model might seem utterly fantastical. But the fact that Altman so casually entertains it sends a crystal clear message to the rest of society about his, and his peers’, world-upending ambitions. To people watching closely, this kind of thing has been part of the rhetoric around generative AI from the start. OpenAI’s original investor agreement warned it “may be difficult to know what role money will play in a post-AGI world.” Altman now seems convinced it could fade out entirely. In fairness to Altman, when it comes to compute as a form of currency, some observers think we might already be there. “Today, AI startups rise or fall depending on whether Nvidia or one of the big tech companies folds GPUs or cloud credits into their seed round,” said Samuel Hammond, senior economist at the pro-tech Foundation for American Innovation. “Companies are literally raising capital in the form of compute, as if it were currency.” Each major AI product release is accompanied by an increasingly potent dose of hype that this outcome is right around the corner. Take the free-to-use GPT-4o voice assistant, announced yesterday just days after Altman’s big-swing interview — an all-in-one app that can talk to you, see and summarize the world around it, and even attempt to detect emotion. The bot has already led breathless critics to compare it to the humanoid AI in the 2013 Spike Jonze film “Her,” a work of near-future speculative fiction that posits a world where not just meaningful labor but emotional relationships are intermediated by AI. In Altman and his peers’ maximalist vision for AGI, tools like GPT-4o are the first step toward a future in which powerful AI models underpin and assist almost every element of the human experience. Assuming that future — which is assuming a lot — it becomes a lot easier to understand why access to computing resources might literally be as good as gold. “One other way to think about this is that if everybody gets ‘Her,’ then you would want nobody to not have an AI companion to speak with,” said Rohit Krishnan, a venture capitalist and author of the Strange Loop Canon Substack. Krishnan took a more modest view of Altman’s UBI pitch, describing it as one of many spacey rhetorical gambits from an entrepreneur who’s made a fortune on them. He pointed out that in a blog post accompanying the GPT-4o release, Altman said giving away access to the system for free was a small price to pay considering the returns it might bring to humanity and that OpenAI is “a business and will find plenty of things to charge for.” “I don't know if [“universal basic compute”] is a replacement [for UBI] as much as a necessary provision to ensure that everybody has access,” Krishnan said. “Giving your best model for free to everybody is a pretty powerful move.” In that light, Altman’s rhetorical flourish about UBI is a neat bit of sleight-of-hand by a master marketer — one that puts his latest product at the center of not just the economic universe, but the entire human experience. And it’s possible that given how central AI has become to global economics and policy in just a few short years, it’s worth taking his idea — or at least the existence of his idea — as important. When I spoke several months ago with Ethan Mollick, an author and University of Pennsylvania business professor who studies AI, he pointed out that all sorts of otherwise unconventional business decisions, like begging for regulation, or giving away your powerful, shiny new toy for free, make more sense if you take developers’ belief in a super-powerful AGI at face value. Altman is ultimately talking about a vision for what humans will actually need if this future all takes place. A universal basic income can give people money, which buys food, shelter, and all the other old-fashioned items on Maslow’s hierarchy of needs. A “universal basic compute” makes more sense in a world where it’s more effective simply to hand people access to a powerful army of machine assistants. Whether that will be empowering, or just another engine for economic inequality in a world where not everyone has access to broadband (or otherwise thinks of their computer as the center of their lives), remains to be seen. FAI’s Hammond points out, for example, that money hasn’t always been the most important currency. Japan once had a feudal system, the koku, whereby wealth was measured in the amount of rice necessary to feed an adult for a year. In that sense Altman’s re-imagining of UBI is a window into a human future where computing feels as central — either to the people who need it, or to their high-tech leaders — as literal food.
| 
Comments
Post a Comment