I work on self-improving AI despite the risks

So because I think it will help us, others might do it first, don’t worry, open algos. Isn’t this everyone’s reason to do something controversial? Also looks like the OpenAI route no?