You know all that money and effort everyone’s putting into AI because they’re convinced it’s a technology that will change everything? What if they’re right?
That’s a kind of fair summary of a Substack post that went viral over the weekend, and then turned into a Wall Street phenomenon Monday, when lots of tech stocks lost value because of AI fears.
I say “kind of fair” because the “what if” post, published by Citrini Research, isn’t just a “what if” post, but a “if it happens, we’re in big trouble” post.
And now the blowback: AI boosters are calling the Citrini post “AI doomerism” and are poking at some of its doomsday scenarios.
I’m not here to argue whether any of Citrini’s individual points are reasonable — I have no idea how AI will affect the likes of, say, DoorDash.
But this also seems like a forest/trees issue. Maybe the details are wrong, but the big picture should be a sobering one: If the AI future we are being promised comes to pass, it could have a massive effect on our economy, and part of that effect could be meaningful unemployment.
Because that’s not just a wild-eyed anti-AI thesis, but one that’s core to the pro-AI thesis, too. It’s what accounts for the enormous investments tech companies are making in AI, and the crazy valuations we’re seeing for the likes of Anthropic ($380 billion) and OpenAI ($850 billion).
In short: The only way those companies are going to be worth anything close to those numbers* is if they end up having a massive impact. Yes, some of the value may be justified by amazing innovation AI could unlock, like wonder-drugs that get discovered in a fraction of the time. But a lot of the value will come from the fact that AI will let computers do lots of work humans do now.
You can’t have the upside without the downside. Full stop.
It’s easy to imagine all the ways this ripples through our economy. And every couple weeks or so, we get a new impetus to fire up our imaginations. Right now, for instance, the market is trying to get its head around the notion that AI might dramatically cut into the enterprise software industry, which is why we saw companies like IBM plummet on Monday. But you can do the same exercise for all kinds of work, in all kinds of industries: Lawyers. Consultants. Screenwriters. Truckers. Etc.
Nor do you have to be an AI “doomer” to think this way — it’s baked into the pitch of the biggest AI companies. “There are cases where entire classes of jobs will go away,” OpenAI’s Sam Altman said last year. His rival, Anthropic CEO Dario Amodei, is blunter and more pessimistic, predicting that AI will wipe out half of entry-level white collar jobs and send unemployment skyrocketing.
It is a little weird to see Wall Street whip-sawing as it thinks through how all of this might go. On Tuesday, tech stocks gained again, presumably because investors have now decided they overreacted on Monday. But I won’t be surprised to see the markets lurch again down the line when someone else makes a convincing case that computers will disrupt a different industry.
Maybe Citrini is wrong about DoorDash and right about lawyers. Maybe it’s the other way around. But the general shape of the problem stays the same: The more AI lives up to the hype, the more it will affect the way people work — or if they work at all. You can be optimistic about that prospect, or terrified. But you can’t ignore it.
*Bear in mind these are private companies, which someday expect to be worth much more when they go public.
