Halvar Flake: AI worm threat plausible only if frontier models hit 1000s of tokens/s on a single CPU
halvarflake · x · 2026-09-14
Halvar Flake says most of his arguments are unchanged from two years ago: the AI worm threat picture becomes far more plausible only if frontier-capable models run at thousands of tokens/s on a single CPU. His skepticism centers on the extreme computational cost of these models, alongside his core point that 'conflicts such as an attempt to kill humanity have no zero-risk moves.'
Related event: Halvar Flake Dismisses AI Worm Threats, Citing Extreme Compute Costs(4 posts)→
More from Safety
- Frog and Toad meme nails the HSM paradox: keys stay safe, but compromised servers can still sign — rickasaurus · 2026-10-31
- Heads Of Security: paper pinpoints and rewires attention heads behind prompt injections — jonasgeiping · 2026-10-31
- House Democrats resolution: no bailouts for failing AI companies — your data goes to bankruptcy court — alex_verem · 2026-10-31
- Amazon Echo kills local voice storage: everything goes to the cloud from March 28 — alex_verem · 2026-10-31
- DOJ's first crackdown on AI-assisted "ghost students" seizes $750K+ in crypto — Ars Technica AI · 2026-10-31
- Researcher Finds 600+ Bugs in LLM Kernel Code; Claude Blocks Safety Analysis — pmddomingos · 2026-10-31