Naval: hidden backdoors in open-weight models would be found by closed labs
naval · x · 2026-07-27
If open-weight models contained hidden backdoors or biases, Naval argues closed-weight labs would eventually detect and disclose them. The post is a short but pointed take on model trust and the incentives around open vs. closed releases.
More from Models
- Claude Opus 5 ranks third on VoxelBench, just 30 Elo behind the leader — legit_api · 2026-07-27
- Open source and open weights are not the same thing, a post argues — RisingSayak · 2026-07-27
- New benchmark adds per-instance trend tracking as models improve fast — jyangballin · 2026-07-27
- ProgramBench adds Pareto curves as models improve sharply just two months after launch — jyangballin · 2026-07-27
- Open-source model debates are turning tribal, and confidence is outrunning evidence — matt_slotnick · 2026-07-27
- Moonshot teases Kimi-K3 with a July 27, 2026 release countdown — teortaxesTex · 2026-07-27