dailyai.report

Open-source weights allow independent researchers to audit LLM safety without relying on corporate promises. Restricting this access creates a transparency vacuum. Nathan Lambert argues that banning open models hinders security research and concentrates power among a few firms. Practitioners lose the ability to verify model behavior and build specialized, private implementations.