Is Evolution Needed for AGI?

And other important questions, such as: (1) does a meaningful form of AGI already exist? (2) is the field of ML a bubble and will there be another AI winter? (3) how are breakthroughs in the field of ML made?