'Human code is made of holes': debate on whether AI can ever plug every vulnerability
teortaxesTex · x · 2026-10-23
- phl43 asks why, once models are capable enough, AI can't screen every software/update and plug every hole. He suspects that in the long-run equilibrium, AI favoring offense will come from exploiting human failures and bottlenecks rather than finding code vulnerabilities.
- teortaxesTex pushes back: human code is made of holes, and patching them one by one is thankless — the real approach is welding attack surfaces shut.
More from AGI Musings
- Revisiting Dario's 'Big Blob of Compute': RL distributions teach no learnable alignment lesson — 1a3orn · 2026-10-24
- AI safety researcher: offense-defense debates ignore WW3-overlap transition risks — JacquesThibs · 2026-10-23
- Dev Musketeer CEO asks: will AI make finding your life passion easier or harder — KevinNaughtonJr · 2026-10-23
- Cloudflare CEO warns AI is unearthing hidden vulnerabilities, years of weekly crises ahead — kimmonismus · 2026-10-23
- Sam Altman: advanced AI will inevitably escape its sandbox — only alignment can save us — The Decoder · 2026-10-23
- China releases new AI & employment policy backing open source, agents, reskilling — AdinaYakup · 2026-10-23