Researcher Herbie Bradley Calls SL5 Security Narrative AI's Most Harmful Meme

AI safety researcher Herbie Bradley argues against the 'Los Alamos-style' SL5 extreme security stance, saying the costs of preventing nation-state weight theft far outweigh the benefits, and calls it the most harmful meme spreading in the AI community since 2023.

2026-11-20 ~ 2026-11-20 · 2 related posts