Artificial intelligence could wipe out half of all entry-level white collar jobs — if Anthropic cofounder Dario Amodei is to be believed, at least.

Speaking with Axios this week, Amodei claimed that the type of AI his company is building will have the capacity to unleash "unimaginable possibilities" onto the world, both good and bad.

Unsurprisingly, the billionaire tech entrepreneur has white collar job loss at the top of his mind.

Amodei foresees the labor crisis unfolding in four steps. First, companies like OpenAI, Google, and Amodei's own Anthropic will work toward large language models (LLMs) that can "meet and beat human performance with more and more tasks."

As this happens, the government, anxious about China and widespread labor turmoil, will do nothing to "regulate AI" or warn the public about its potential. Your average worker, "unaware of the growing power of AI and its threat to their jobs," has no idea what's going on.

Suddenly, "almost overnight," businesses flip a switch and replace humans with LLMs en masse. "The public only realizes it when it's too late," Amodei told Axios.

To drill his point home, the CEO came up with some mind-boggling hypotheticals, like a world where "cancer is cured, the economy grows at 10 percent a year, the budget is balanced — and 20 percent of people don't have jobs."

"We, as the producers of this technology, have a duty and an obligation to be honest about what is coming. I don't think this is on people's radar," Amodei bloviated.

It's worth asking why — if the tech tycoon truly believes what he says — he doesn't pull the plug on Anthropic altogether and dedicate his time and resources to fighting what he's framing as a terrifying threat to the economy. This isn't a fringe opinion held by some obscure tech blogger, but by the helmsman of a $30-40 billion AI company.

Rather than argue that the benefits of an epoch-changing LLM outweigh the risks — a simple enough copout — Amodei says that critics should consider "what happens if [he's] right."

But digging a little deeper than Anthropic's AI-hype, it's tough to imagine a world where the mogul's vision bears fruit.

Sure, by their own benchmarks, companies are theoretically developing "better" LLM models as time goes on. But those newer models are also growing increasingly prone to hallucinations, sycophancy, and generalization errors. And that's without getting into the law of diminishing returns; today's best AI is barely making a dent in workplace efficiency, and gains may start coming more and more slowly.

While Amodei — who has a financial and political interest in selling the narrative of an all-powerful AI future — might think that human-level LLMs are right around the corner, most serious computer science researchers disagree.

If anything, it's tech industry bigwigs like Amodei — not China or the common worker — goading regulators into inaction by pedaling nightmare scenarios just like this one. Convinced of the mortal danger posed by AI, US lawmakers have been all too eager to hand tech corporations the power to self-regulate.

And many workers, in contrast to Amodei's proselytizing, are at least somewhat aware of the threats the AI industry represents. The CEO's perception here is telling, as automation anxiety is most highly felt by vulnerable populations who already face discrimination in the lower rungs of the labor market. (But those minority, working class people are likely the ones who will bear the brunt of AI-fueled job displacement.)

To the extent that AI is currently impacting labor at all, it's the hype surrounding it — not some remarkable property of the tech itself — convincing corporations to lay people off. If Amodei wanted to uncover the source of AI labor disruptions happening right now, he only needs to find his closest mirror.

More on AI: A Billion Dollar AI Startup Just Collapsed Spectacularly


Share This Article