The Case for Nationalizing Artificial Intelligence

Our special summer issue is out now. Get a discounted subscription to our print magazine today. Your support keeps us publishing!

The Case for Nationalizing Artificial Intelligence<br>ByDustin Guastella<br>Artificial intelligence has been built by robbing the collective work of humanity. The public built AI — we should own it, not a handful of billionaires.

If taxpayers bail out the AI industry when its bubble bursts, as seems increasingly likely, we shouldn’t repeat the mistakes of the 2008 crisis. A public bailout should mean nationalizing AI. (Heather Diehl / Getty Images)<br>Our summer issue is out now. Get a discounted subscription to our print magazine today.

Israel’s Young Settler Vanguard<br>E. A. Halevi

The Making of the Teenager<br>Lauren Fadiman

Zohran Needs to Create Popular Assemblies<br>Gabriel HetlandBhaskar Sunkara

A Half-Century of Harry Braverman’s Labor and Monopoly Capital<br>Sophina ClarkDaniel Judt

Silicon Valley is still riding high on what seems like an unending wave of Wall Street enthusiasm. Investments in artificial intelligence continue to eclipse expectations, with Bloomberg reporting that financiers plan to plow some $700 billion into the industry this year alone. AI stocks now account for roughly a third of the entire stock market and 45 percent of the S&P 500’s total market capitalization.<br>Of course, the extraordinary investments in AI are driven by the technology’s promise to make large swaths of workers redundant. Tech leaders aren’t shy about saying as much. “It is not clear,” says Dario Amodei, chief of Anthropic, “where these people will go or what they will do, and I am concerned that they could form an unemployed or very-low-wage ‘underclass.’”<br>If that’s not anxiety inducing enough, the explosive growth of AI investment is suggestive of a speculative bubble, perhaps the biggest of all time. If it bursts — a very real possibility — it would take down the entire global economy. Anyway you slice it, the AI boom-and-bubble represents a tremendous social and economic liability, and one that the government must eventually reckon with. But how?<br>In America, the modal policy response to any sort of economic crisis is to bail out failing firms while leaving the workers to fend for themselves. President Barack Obama’s auto industry policy is a paradigmatic example. In a sense, the reasoning was sound, if woefully incomplete. Letting Detroit fall would’ve meant destitution for hundreds of thousands of autoworkers and millions of other workers who depend on that industry. However, after the checks were cashed and the dust settled, the government and the governed received no great return on our public investment despite, at one time, owning 61 percent of General Motors. No dividends or profit sharing resulted. A few years later, General Motors launched a massive stock buyback program to the tune of $10 billion in 2023, then $6 billion in 2024. The welfare all went to the top, while workers were asked to make “shared sacrifices” on the shop floor.<br>This won’t work with AI. Not only will bailouts be economically insufficient to rescue wildly overcapitalized firms, but there exists no political and moral justification for rescuing them. In fact, it’s not clear whether AI’s success or failure would kill more jobs.<br>If the bubble bursts this time, the response must be one that aims to correct, rather than exacerbate, the balance of power between capital and labor.Dealing with AI, then, needs to start from a different, bolder premise. There are those who wish to simply halt AI’s progress. As someone who blames the smartphone for so much of our social pathology, these Neo-Luddites have my tremendous sympathy. However, as satisfying as it may be to shout “No!,” it won’t do much. For every small town that throws up a barrier to building a data center, there will be a dozen investment-starved ones begging for the ribbon cutting. And even if we were to somehow jail the heads of Anthropic, OpenAI, and Gemini, and make AI development illegal stateside, new firms would shortly take their place abroad — the economic incentives are just too irresistible. The cat really is out of the bag.<br>Opposite the Neo-Luddites are the tech-friendly progressive reformers who adopt the logic of the meteorologist. They insist that we cannot stop or alter the path of AI anymore than we can stop or alter the path of a hurricane. Our only recourse is to build up figurative levees to slow the rising flood of unemployment claims; mitigate the destructive effects of mass unemployment through cash transfers; and later clean up the mess. Of course, there is nothing wrong with trying to help those displaced by AI-induced automation, just as we would those displaced by a hurricane. But the pitch for reform along these lines concedes a great and destructive lie: that technological progress and its effects are inevitable and uncontrollable — a force of nature. Yet the reason we can’t stop,...