4 ms·

Reductionist approach (aka deep neural networks) is not going to work at full complexity scale but it is our best shot right now.