Brad Gerstner Calls AI Extinction Warnings ‘Hyperbolic Scare Tactics’
The Altimeter Capital chief says companies are taking unusual precautions, while former researchers warn that fast-improving systems could become dangerously autonomous.
Listen to this story
The audio brief
Story brief
3 key pointsAltimeter Capital CEO Brad Gerstner has rejected recent AI-extinction warnings as politically motivated exaggeration, putting investor credibility into the debate because his firm backs both Anthropic and OpenAI. The dispute was triggered by former researchers’ warnings about rapid progress toward systems that could independently conduct AI research and improve themselves. Daniel Kokotajlo placed that possibility...
- 01
Gerstner said AI companies are taking safety more seriously than internet and social-media companies did during their early development.
- 02
Former Anthropic researcher Jacob Coxon said AI firms were “gambling with our lives,” prompting broader researcher criticism online.
- 03
Former OpenAI governance researcher Daniel Kokotajlo called current model progress “very concerning.”
A public dispute over AI’s worst-case risks has become a fight over whether warnings are a safety alarm or political messaging. Brad Gerstner, chief executive of Altimeter Capital, called recent discussion of AI causing human extinction “hyperbolic scare tactics” and said he believed it was hiding behind a political agenda.
Gerstner argued that the AI industry is already making extraordinary efforts to address harms, unlike the earlier development of the internet and social media. He said the week’s discussion created the false impression that companies were advancing with total disregard for safety.
CNBC reports that Altimeter invests in both Anthropic and OpenAI, two companies central to the debate over increasingly capable AI systems. The investment disclosure gives added context to Gerstner’s public defense of the industry’s safety efforts.
Gerstner’s comments followed an outcry sparked by Jacob Coxon, a former Anthropic researcher, who left his job and wrote on X that AI companies were “gambling with our lives” in their pursuit of superintelligence. Coxon’s post ignited a social-media frenzy, with more researchers joining the conversation about rapid AI progress.
Daniel Kokotajlo, a former OpenAI governance researcher, described the direction of model improvements as “very concerning.” He forecast that within one to three years, AI could conduct the full research process itself and become autonomously self-improving—an outcome he said could go very badly.
Kokotajlo was describing a future scenario, not a system he said already exists. His warning nevertheless identifies the capability at the center of the dispute: AI that can carry out research and improve itself without direct human direction.
The clash is unfolding as opposition to AI has widened across the U.S., including backlash tied to the buildout of large AI data centers. That broader response gives Gerstner’s charge of political motive a wider public setting than the researchers’ warnings alone.
Sources
- cnbc.comAltimeter's Gerstner blasts researchers voicing AI extinction warnings, questions 'political agenda'
Loading discussion...
Reader comments
Newest comments first. Replies stay oldest first.