Morphism comments on Please stop publishing ideas/​insights/​research about AI

[Morphism]

I could see embedded agency being harmful though, since an actual implementation of it would be really useful for inner alignment