Is AI’s Apocalypse a Sales Pitch?

Last Sunday, Anthropic CEO Dario Amodei joined President Donald Trump for a one-on-one private dinner at the White House. The timing is hard to ignore; this dinner came weeks after Amodei and other artificial intelligence “AI” pioneers issued statements warning the public of an anticipated take-off of AI and the threat it would pose to humanity. Amodei also called for government regulation to slow the pace of technological development. But if Amodei wants things to slow down so badly, why doesn’t he turn his attention towards the progression of his own company, over which he has executive control? In my view, it is because he wants to form a partnership with the Trump administration—a recipe for regulatory capture—and profit off investors who flock to the idea of an AI take-off.

On September 8th, former Anthropic employee and AI researcher  Jacob Coxon posted a blast to X that gathered just under 175 million views, stating, “Neither company [OpenAI nor Anthropic] is acting responsibly. They are straight to self-improving superintelligence and gambling with our lives.” Coxon left Anthropic the same day he posted the statement, and the public quickly pointed out that he left months before his equity vested, arguing that he had no ulterior economic motives. 

While this argument may ring true for Coxon, it does not seem to apply to those who remain at the company. Shortly after Coxon’s post, Evan Hubinger, Anthropic’s head of alignment stress testing, publicly agreed, saying he believes there is a greater than 10 percent chance that AI could cause human extinction within the next 10 years. This month, Amodei wrote, “we must slow the pace at which we improve the capabilities of AI models.” For Anthropic’s leadership, quickly falling in line with Coxon does exactly what is best for its bottom line: bring in money and attention.