OpenAI’s Long-Term AI Risk Team Has Disbanded
The blog post announcing the superalignment team last summer stated: “Currently, we don’t have a solution for steering or controlling a potentially superintelligent AI, and preventing it from going rogue.”