OpenAI @yong_zhengxin pushes back on Yudkowsky's 2008 claim that a Bayesian superintelligence could derive general relativity from 3 frames of a falling apple: "I actually don't think that would be the case. What is most likely to happen is that from those three frames, it would think of various hypotheses, and use what it has seen in other experiences, maybe other interactions of gravity." "From that onward, it will put together a bigger picture, a bigger explanation that can explain…