Ask HN: What AI/DL courses would you recommend me?

I'd describe myself as someone who has beginner/intermediate skills in the field. I started exploring AI field at the beginning of the last year. Then I took a year off, exploring other fields trying to figure out what am I interested in. Now I can confidently say that AI/DL is a path I want to pursue in the future. - I have some basic experience across the DL field in general (GANs, convolutional nets, word2vec, LSTMs, math/reasoning behind these topics) - I have experience building simpler neural nets (mostly conv nets + GANs) from the scratch using NumPy - I was frequently working with Tensorflow but I can imagine that the library changed a lot since the past year - I feel that my probabilistic/statistics skills could be better - I feel that I have very little experience building AI in the real world - I feel like I need to learn how to be able to apply the latest scientific papers What I need now is to (1) refresh my old knowledge, (2) take more challenging courses/classes with more "real world" approach and (3) start reading (and applying) scientific papers. Do you have any recommendations regarding me coming back to this field and getting deeper into it?