One of the biggest challenges of the 21st century is to make computers more similar to the human brain. We want them to speak, understand and solve problems — and now we want them to see and recognize images.

For a long time, our smartest computers were blind. Now, they can see.

This is a revolution made possible by deep learning.