Altman says AI finds bugs we can't patch; Goertzel proposes $1T formal verification push
bengoertzel · x · 2026-10-24
OpenAI CEO Sam Altman warned that AI models are finding bugs outside the normal threat model of what hackers can do — "hardware-level, insane stuff" — adding he no longer believes we can simply patch all the bugs. The same week, the PHP vulnerabilities dubbed PHP2Shell were reportedly discovered using OpenAI's model, enabling remote code execution and data exposure.
Ben Goertzel argues the solution is using coding agents to replace essentially all existing buggy code with formally verified, safety-proven code — a huge undertaking, but one the current pre-AGI yet superpowerful AI tech can enable. He suggests governments allocate $500B (eventually trillions) to open-source communities to roll out automated buggy-code replacement, comparing it to the 2008 bank bailouts.
More from AGI Musings
- Technologists have a duty to create, not just critique—quoting Blake — fkasummer · 2026-10-25
- Palantir CEO Alex Karp: Europe barely matters in the AI race between China and the US — 2C_ornot2C · 2026-10-25
- AI drug-discovery skeptics have a point: binding simulation accuracy is just 49%, data costs 30x longer — citrini · 2026-10-25
- Blogger claims hyperscaler-backed AI regulation aims to crush small labs, entrench monopoly — Ghost_Pilot_MD · 2026-10-25
- Toddler development as a metaphor: evolutionary pre-training and efficient post-training — paigeinsf · 2026-10-25
- "I do miss programming": dev says AI coding killed the flow state — generativist · 2026-10-25