www.lesswrong.com/posts/aD2RA3vtXs4p4r55b/on-not-feeling-the-agi
1 correction found
The downside risks of AI, let alone existential risks, are flat out not discussed
This is incorrect: the interview explicitly discusses AI safety, regulation, and even world-destroying risks.
Full reasoning
Ben Thompson’s March 20, 2025 interview with Sam Altman includes a dedicated section titled “Hallucinations and Regulation.” In that section, Thompson asks: “If we’re worried about safety and alignment, isn’t teaching AIs to lie a very big problem?” Altman also says “I still think there can be big risks in the future,” and, while discussing regulation, adds “hopefully, we’ll find our way through as best as we can and probably it’ll be fine. Not that many people want to destroy the world.”
Those lines show that downside risks—and even existential-style concerns—were discussed in the interview. One can argue they were not discussed enough, but saying they were “flat out not discussed” is contradicted by the transcript itself.
1 source
- An Interview with OpenAI CEO Sam Altman About Building a Consumer Tech Company - Stratechery by Ben Thompson
The transcript includes a section titled “Hallucinations and Regulation.” Thompson asks, “If we’re worried about safety and alignment, isn’t teaching AIs to lie a very big problem?” Altman says, “I still think there can be big risks in the future” and “Not that many people want to destroy the world.”