Is AI apocalypse talk distracting us from the real threat?

News items heralding the AI apocalypse are ubiquitous. The panic started shortly after the ex-Anthropic, ex-OpenAI researcher-turned-whistleblower, Jacob Coxon sounded the alarm on X, writing that AI companies are "racing straight to self-improving superintelligence and gambling with our lives.”

Coxon’s diagnosis is startling. AI, he wrote, “could kill us all by the end of the decade.” In a shocking turn, Anthropic’s own Evan Hubinger confirmed growing fears when he wrote, “We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade."

It’s no surprise that dire remarks like this from AI experts have turned every news headline or cable news segment this week into a freak-out over the coming doomsday. But is all of this panic really warranted?

In the latest episode of Always at War, we attempt to parse what’s actually dangerous about AI from cynical industry hype. We focus aspects of the story often overlooked by Washington’s op-ed class: namely, that in the same breath that AI CEOs predict their own technology will end humanity, they’re also working closely with the Pentagon to build deadly weapons, expanding the surveillance state, and asking lawmakers to remove guardrails — like antitrust laws — that would limit their companies’ power.

And, in a linguistic minefield that would entrap the Bard himself, CEOs like Sam Altman, Elon Musk, and Dario Amodei claim that the only way to prevent AI Armageddon is to free AI companies from the rule of law so that they can – stick with me – use AI to help us win a war with China.

If this sounds insane, you’re not alone.

Join us as we try to understand why AI CEOs are so eager to push the "end of humanity” narrative. Are these guys defending working at the proverbial “boiling people alive factory?” Or is this just a convenient ploy to distract from AI’s immediate harms, like its use to target and kill civilians in Gaza and Iran and the massive build-up of AI surveillance technologies here at home?

As we discuss in the episode, whatever its upsides, we also know this technology is already being used sloppily and at the expense of people's lives. The dangers of AI are not simply hypothetical. So why not focus on the real and present dangers that are already affecting us?