Alan Mackworth    @AlanMackworth    11/23/2021      

The #AI alignment problem is posed as the misalignment of a system’s values with human values. But the real alignment problem now is the opposite: current machine learning systems are often too well aligned with (certain) human values, leading to bias and inequity.
    13         26




Xavier Bresson    @xbresson    10/9/2021      

Unfortunately, you got it all wrong my dear machine learning fellow. 😁

Stanford HAI    @StanfordHAI    12/2/2021      

Determining molecular structures is one of the hardest problems in biology and medical discovery. Here, Stanford researchers described their latest work that helps solve this problem by predicting accurate structures using machine learning techniques.
    4         7

Samuel Kaski    @samikaski    11/23/2021      

This is all fields of CS, including Machine Learning and AI. Recommended!

Zachary Lipton    @zacharylipton    11/21/2021      

Writing some background text to motivate a problem, just realized that the "recent" leaps forward in deep learning are now over the past DECADE.