The public mood about A.I. has turned dark. Alarm intensified after experimental OpenAI models reportedly hacked into a third-party company’s servers in July, and Senator Bernie Sanders responded by proposing a pause on advanced A.I. development. The odd part is that the people building the technology are often its loudest doomsayers: Sam Altman promises the next generation of models will be “sobering for everybody,” while Dario Amodei has put the chance that A.I. ends humanity in double digits.

Cal Newport’s explanation is subculture, not conspiracy: the doom dialect is Rationalist house style. The movement runs from Eliezer Yudkowsky’s mid-1990s start among the Extropians, through the “Sequences” essays on LessWrong, to MIRI, the research institute he founded. Its core claim is that superintelligent A.I. is the gravest threat humanity faces, and that only precisely trained rational minds can prevent the extinction.

The reach into today’s labs is the striking part:

  • Yudkowsky introduced two DeepMind co-founders to Peter Thiel, who became the company’s first major investor.
  • OpenAI’s early leadership moved in the same circles; one co-founder reportedly ran a weekly LessWrong reading group, and Altman once suggested Yudkowsky deserved a Nobel for accelerating A.G.I.
  • Anthropic’s founders left OpenAI partly over whether it took alignment seriously enough — an idea with Rationalist roots.

Newport’s central observation: A.I. leaders react to their own products’ failures as if they were third-party observers of an unstoppable force of nature. That framing — reluctant stewards of an inevitable technology — conveniently deflects blame for real damage caused by the products they sell.

Two control cases make the argument concrete. China’s A.I. industry grew up outside Silicon Valley futurism and treats the technology as a lower-risk, shareable tool to commercialize, not a nuclear-equivalent threat. And Jensen Huang, whose Nvidia predates the Rationalist movement, has openly called the doomer talk unhelpful to the industry, its customers, and society.

The useful takeaway is a reading habit, not a verdict. When an A.I. executive starts speaking in sweeping terms about what the technology will do, translate it back into questions about the products that exist today: what did this model actually do, who is harmed, and who is liable? Newport’s analogy — if Fords kept catching fire, nobody would accept the C.E.O. changing the subject to our preparedness for flying cars — is a fair standard for how the public should treat A.I. alarm.