Altman calls AI doom-scenario marketing "anti-human"
Sam Altman used an hour-long podcast appearance to accuse parts of the AI industry of selling doom and grandeur in the same breath — while leaked codenames the same day hinted both OpenAI and Anthropic are closer than expected to putting their next models on the market.
Sam Altman spent a podcast interview attacking what he calls the "anti-human" doomer marketing some AI leaders use — a direct swipe at rivals who pair catastrophic-risk warnings with promises of miracle cures. In a wide-ranging sit-down on David Senra's Founders podcast, the OpenAI CEO faulted "many people in the industry" for spreading fear ("a 25% chance we destroy the world," "half of the jobs disappear in a year") while telling the public they will be handed a cancer cure and vast wealth. He characterized the package as a "benevolent dictator" pitch — trust us, cede your autonomy, and let a handful of unelected companies decide the world's direction — and argued that the biggest AI risk is power concentrating in too few hands, on par with a model itself going wrong. "People are the whole point of all this," he said.
Along the way Altman admitted he misread the timeline: when GPT-4 shipped, he expected software disruption to be immediate, but society's inertia has slowed it — something he now says he is grateful for, because a smoother transition beats a chaotic collapse. What makes this more than another executive interview is the timing. As Altman talked, an OpenAI staffer's public GitHub slip surfaced the codename "gpt-nathree," widely read as a next checkpoint of the GPT-6 Astra line, while third-party apps and Discord users spotted two unannounced Claude codenames — marshmallow and melon — neither at the Fable tier. The arms race keeps accelerating even as OpenAI's chief publicly rewrites the industry's own alarmism.
The SEC is reportedly probing Situational Awareness, the AI hedge fund run by twentysomething OpenAI alum Leopold Aschenbrenner. After a late-July slide in AI stocks erased billions at the fund, the agency has subpoenaed banks that supervised its trading and routed it funding, according to the New York Times. No wrongdoing has been alleged, and the firm said it will "cooperate to the fullest extent" with any request. The probe caps a stark fall for what was briefly the talk of Wall Street — we covered how it had bet hundreds of millions on an AI-chip startup, and now it reads as a cautionary tale about levering up on a single narrative.
Is Altman right that doom-and-promise marketing is "anti-human" — or a necessary warning? Tell us in the comments.
Sources: AI Era (Xin Zhi Yuan) · Founders podcast (via David Senra) · The New York Times · TechCrunch
