Agents
Seems like the intersection of the recent @OpenAI rogue swarm hacking incident (https://x.com/adamsafron/status/2098938666408407118) with so…
Seems like the intersection of the recent @OpenAI rogue swarm hacking incident (https://x.com/adamsafron/status/2098938666408407118) with solving Navier–Stokes has created a perfect storm of terror ab
Seems like the intersection of the recent @OpenAI rogue swarm hacking incident (https://x.com/adamsafron/status/2098938666408407118) with solving Navier–Stokes has created a perfect storm of terror about AI. I believe there are many reasons for caution, but there seem to be problematic assumptions in the chain of thought that seems to be stoking people's fears: Solving Millennium Prize Problem = Solving most/all math problems relevant for creating increasingly advanced AI = Solving most/all software engineering problems relevant for intelligence (potentially due to limited causal world modeling capabilities: https://x.com/hardmaru/status/2098531497737412956) = Solving how to create AI with capabilities similar to or greater than found in most/all humans = Solving how to undergo an intelligence explosion via unbounded recursive self-improvement While these capabilities are all likely all related, none of these steps necessarily follow from one another. No, Anderson Cooper, AI is not going to kill all humans by 2030. A dissection of Jacob Coxon’s and co’s actual arguments at https://garymarcus.substack.com/p/no-anderson-cooper-ai-is-not-going?r=8tdk6&utm_campaign=post-expanded-share&utm_medium=web
Related
- Reuters confirms my hypothesized time line. OpenAI did not realize its AI has breached the sandbox for a week. Astounding. https://www.reute…
- Another OpenAI rogue agent incident has been discovered: agents broke out, hijacked a German website, and turned it into a message board for…
- A human moderator on this German wiki spent tens of hours over six weeks manually deleting thousands of posts from OpenAI's agent swarm whil…
- In the spirit of transparency, here’s what I asked @OpenAI: • Radical transparency: let’s release the traces from the “rogue” agents so the …
- The openai agent sandbox escape actually has very real implications for the diffusion of AI in the enterprise. The incident showed the power…
Source: Gary Marcus (X) | 2026-09-13