Open models are not the opposite of safety, BlackHC argues in a new post
BlackHC · x · 2026-07-24
Open models and safety are not mutually exclusive, the post argues
BlackHC says they are both pro-open models and pro-safety, and that the two positions are not contradictory.
Key points:
- Closed models carry risks, but open models do too.
- Simply banning open models would be too reactionary.
- At the same time, the “YOLO” attitude that dismisses all risks from open models is also not acceptable.
- The right approach is to acknowledge increasingly capable model behaviors and try to minimize specific risks rather than treating openness as inherently safe or inherently dangerous.
More from AGI Musings
- Emad Mostaque says open-weights debates are really US-vs-China debates — emollick · 2026-07-24
- If LLMs solve existence but fail universal claims, math academia may barely change — JFPuget · 2026-07-24
- If an AI knows which memories you regret, can it use them as leverage? — PierceLilholt · 2026-07-24
- Allie K. Miller Says She’s Betting on Multiplayer and Proactive AI — timemagazine · 2026-07-24
- Gary Marcus’s 2022 AI predictions already look outdated by 2026 — dhadfieldmenell · 2026-07-24
- Gary Marcus says today’s systems are not close to AGI — GaryMarcus · 2026-07-24