Safety & independence
We love using ChatGPT. We also believe powerful AI deserves scrutiny, meaningful oversight, and public accountability.
We support a pause on developing superintelligence until there is broad scientific agreement that it can be built safely and kept under control, with meaningful public support. AI's benefits do not justify gambling with people's safety, livelihoods, or ability to shape their future.
This is our editorial position. The Statement on Superintelligence describes the approach we support.
Being a fan does not mean defending OpenAI’s every decision. We will link to substantive criticism of OpenAI and other AI developers, alongside research on present harms and potentially severe future risks.
Our safety library distinguishes research, practical guidance, and advocacy. Inclusion is an invitation to examine a source, not an endorsement of every claim it makes.