Dario Amodei thinks there’s a bloodbath for American workers on the horizon.

He should know. His company, Anthropic, a leader in artificial intelligence, is building the tools that will be used in the bloodletting.

Earlier this month, Anthropic introduced the next generation of its AI-powered coding model. It’s named Claude, after Claude E. Shannon, who is often referred to as the “father of information theory.” Shannon’s namesake is poised to end a range of entry-level job tasks, from coding to administrative duties.

As models like Claude become more sophisticated, bosses will have less incentive to hire, say, recent college grads. That’s good news for experienced professionals whose skills will, for now, remain in demand. But it’s bad news for anyone hoping to get their feet in the door and learn those skills from such professionals.

The perennial employment dilemma is already ugly enough, with many entry-level jobs requiring or giving preference to applicants with experience that cannot be obtained unless someone first hires them.

Now, things are about to get even worse. It’s already starting.

“Leading firms, including major banks and tech companies, are reassessing junior roles becuase [sic] AI can now perform much of their workload,” TheStreet reported.

According to Amodei, AI could eliminate half of all entry-level white-collar jobs in the next one to five years, spiking unemployment to anywhere between 10 percent and 20 percent. He envisioned one possible future in which “cancer is cured, the economy grows at 10% a year, the budget is balanced—and 20% of people don’t have jobs.”

It is hard to fathom all the economic, political, and social chaos this scenario would cause. So hard to fathom, in fact, that most people choose not to believe it can happen, according to Amodei. That includes policymakers, the only ones in a position to take action before the clock strikes midnight.

“Most of them are unaware that this is about to happen,” Amodei told Axios. “It sounds crazy, and people just don’t believe it.”

Apart from ignorance and self-interest, I think many people also don’t want to believe it. This is the impression I get from friends who work in the IT field. A few years ago, they insisted that, in the worst case, AI would just push out the chaff—and, apparently, entry-level programmers—because that’s all its capabilities would allow. Well, we’ve just about arrived at yesterday’s worst-case scenario, and the Claudes won’t stop there.

Americans may be looking down the barrel at the rapid restructuring of society being ushered in by the ascendant algocratic elite, who are occasionally called “Futurists,” though that label is something of an error.

Futurism emerged in the early 20^th century, primarily as an intellectual and artistic movement inspired by Italian poet Tommaso Marinetti. It emphasized technology, speed, violence, vitality, and youth. The past was the ultimate enemy to the Futurists, who sought to level and sever all connections to it. All traditions were suspect, all conventions the enemy of progress, the engine of which was war. In one of their manifestos, a group of Futurist painters declared:

We rebel against the spineless admiration for old canvases, old statues, and old objects, and against the enthusiasm for everything worm-eaten, grimy, or corroded by time; and we deem it unjust and criminal that people habitually disdain whatever is young, new, and trembling with life.

On second thought, maybe labeling AI advocates “Futurists” isn’t so off the mark after all. There are some clear similarities in outlook. Many advocates of AI accelerationism share an indifference or outright hostility toward what came before. When Marinetti talked about “a new beauty: the beauty of speed,” he had in mind the screaming race car that “seems to run on machine-gun fire.”

Now, the new beauty of speed is the internet and the all-consuming model that eats the past and the future. Marinetti could only have dreamed of wielding such destructive power.