A spelled-out intro to neural networks and backpropagation [video]

I'm sure Karpathy did a good job at Tesla to make self-driving cars - as good as anyone could have, anyway. But it's better for the world to have him back as a teacher/blogger/YouTuber/popularizer.

I can't recommend enough this "backpropagation from scratch" from heremy howard (fastai) where he implementes backpropagation from scratch using python.

UPDATE: link https://www.youtube.com/watch?v=CJKnDu2dxOE

The poster of this link, hardmaru, is no slouch either. Their work with evolutionary strategies is eye opening.

otoro.net/ml or their twitter account @hardmaru if you want to see for yourself.

Thanks for the link hardmaru!

Can’t imagine the stress levels he had to endure at Tesla. Looking from the outside it seems totally worth it - he solved “the money problem” very young. Now he can do whatever the hell he wants and he keeps building because that’s what builders do. He’s such an inspiration.

Karpathy is awesome and I love that he's doing this and going down this route. (Hope he doesn't get that "offer he can't refuse" like carmack predicted and keeps doing educational stuff for us all :).)

This is a very nice video from what I've seen so far. Hoping to see more from Karpathy. He seems to have a talent in explaining things.

Wow — this was in fact the most easy to understand and yet from-first-principles intro to NN that I've seen. Glad Andrej is focusing on ML education.

Andrej, this is really good. I like how you clearly and succinctly describe how it works. I subscribed to your channel! Thanks for sharing.