Nobody Knows Anything
The fact that a piece of AI science fiction rocked the stock market this week is a clear indication that absolutely no one knows how the next few years will go.
It’s not every day that a piece of science-fiction gets treated as a news bulletin. In 1938, a radio broadcast of H. G. Wells’ novel The War of the Worlds, produced and read by the actor Orson Welles, sent listeners into a tizzy, as many believed that war-hungry Martians had set foot in New Jersey. Nine decades later, another science fiction story about the untoward effects of alien intelligence making contact with Earth has caused another national freakout.
If you know, you know: On Sunday, a viral article published by Citrini Research called “THE 2028 GLOBAL INTELLIGENCE CRISIS” described a future-world scenario where powerful AI leads to spending cuts across the economy, triggering a massive recession. On Monday, investors reacted as if the article were a news reel from the future. Many of the companies name-checked in the essay saw their stocks plunge. The Wall Street Journal reviewed the day’s carnage under the headline “Viral Doomsday Report Lays Bare Wall Street’s Deep Anxiety About AI Future.” Some details:
Software firms Datadog, CrowdStrike and Zscaler each plunged more than 9%. International Business Machines’ 13% decline was its worst one-day performance since 2000. American Express, KKR and Blackstone—all name-checked by Citrini—tumbled …
Shares in DoorDash also veered 6.6% lower Monday after Citrini’s Substack note called the delivery app a “poster child” for how new tools would upend companies that monetize interpersonal friction. In the research firm’s scenario, AI agents would help both drivers and customers navigate food deliveries at much lower costs.
Is the Citrini story … compelling? plausible? accurate? These are the questions colonizing all of financial and tech media at the moment1. But to me, those questions miss the deepest and most interesting feature of this strange episode: What does it say about the state of AI and AI anxiety that a literal science-fiction story had the power to move a trillion dollars?
AI’S MARKETPLACE OF FICTION
“Nobody knows anything,” the author and screenwriter William Goldman once wrote of the movie business. Goldman was talking about Hollywood’s famous inability to predict future hits. Surrounded by false confidence, Goldman counseled humility.
Today his motto applies forcefully to the discourse around artificial intelligence. I am lucky to have participated in conversations about the future of AI with executives and builders at frontier labs, economists at AI conferences, AI investors, and other bigwigs at off-the-record dinners where important truths can theoretically be bandied about without risk. And if I had to pick three words to summarize this collective expert view of the future, I could not in a million years, or with a trillion tokens, find three words more suitable than these: Nobody knows anything.
I do not mean that AI architects are stupid. I do not mean that their speculation is absurd or worthless. I certainly do not mean that they don’t have access to narrow truths, such as rising adoption of AI in general and autonomous “agents,” in particular. What I mean is that the frontier labs don’t really know what they’re building exactly2, and economists don’t know how to model the thing that they claim they’re building. As a result, nobody really knows what is going to happen with AI this year, or next year, or the year after. There is no secret cigar-filled room of elites who have unique access to some authentic postcard from the future. When you drill down underneath the bluster, the boosterism, the fear, and the anxiety, what’s present at the bottom of it all is a genuine uncertainty, a vacuum into which storytelling is flooding.
Even folks who you’d think would be experts about AI and the economy are mostly offering elaborate speculation by way of analysis. Here’s JP Morgan chief executive Jamie Dimon at an investor cocktail event on Monday evening (as quoted by Yahoo Finance executive editor Brian Sozzi):
What if, I think there are 2 million commercial truckers in the United States, and there are lots of other examples you can give. There’s a thought exercise, and you could push a button, eliminate all of them, and they make $120,000 on average. Save fuel, save lives, save time, a more efficient system, less disrupted highways, all that beautiful stuff. Would you do it if you put 2 million people on the street where even if there are jobs available, that next job is $25,000 a year, stocking shelves. I was saying, “That’s kind of really bad, kind of civilly, should we as society agree to that?” I don’t think so. I was talking about the business and government, and they should start thinking today, not when it happens, what would we do to deal with the [AI] issue? It’s got to be business and government.
Long block quote, sorry. But the most important words are the first two: “What if.” The conversation about AI really has become a marketplace of competing science fiction narratives. The level of uncertainty about AI’s economic effects is so high—and the quality and supply of real-world, real-time information about its economic effects so paltry—that even serious conversations about AI from otherwise analytical people often veer toward the genre of fiction rather than the category of empirical analysis. AI land is so full of science fiction precisely because the space is so bereft of official high-quality data.
Consider, for example, a topic that should be incredibly easy to talk about clearly: AI and the labor force. Many people predict that AI will soon destroy jobs, perhaps at catastrophic levels. This generation of AI—ChatGPT, Gemini, Claude, Copilot, etc—is already the fastest-growing technology class in modern history. So one might guess that it’d be trivially easy to say whether the technology ripping through the economy is reducing employment, yes?
Nope. Despite predictions of imminent labor market “bloodbath” by AI’s architects, economists have struggled to find any solid evidence of AI’s presence in the official data. A report from the Economic Innovation Group found “little evidence of AI’s impact on unemployment.” In the Financial Times, John Burn-Murdoch pointed out that “the much-discussed contraction in entry-level tech hiring appears to have reversed.” Last August, a team of academics from Stanford University published a paper claiming that young workers aged 22–25 in “highly AI-exposed” jobs, such as software developers and customer service agents, experienced a 13 percent decline in employment since the advent of ChatGPT. But when the Atlantic journalist Josh Tyrangiel asked other economists if the Stanford paper had settled the debate, his sources mostly wanted to “punch holes” in its methodology and interpretation. After spending dozens of hours talking to numerous economists about the issue, Tyrangiel concluded that “numerically speaking, nothing indicates that AI has had an impact on people’s jobs.” Nothing!
Artificial intelligence offers its obsessives a kind of Schrodinger’s apocalypse, which exists in a superposition between “the economy is about to change forever” and “from a macroeconomic standpoint, everything still looks eerily normal.” In the film Don’t Look Up, Adam McKay made a climate change parable where politicians and the public ignored an approaching comet despite the desperate warnings of scientists who saw the rock clearly through their telescopes. In the case of AI, the desperate warnings are with us, and they may be fiercely prescient, and I wish more people were taking the possibility of violent and calamitous change more seriously; but also, let’s face it, our telescopes kind of suck, and when we put our eye to the lens, nothing is clearer than the fact that little is clear.
THE SEVEN QUESTIONS
There are so many questions that I want answers to in the space of AI and economics, but for now I have to reconcile myself to the fact that we don’t have answers to at least seven questions that will determine the course of this decade and even this century. Those seven questions are:
