The problem with current ai right now is that it's very complicated and recent work (meaning the tools required are not for everyone), so it's more research tha

The problem with current ai right now is that it's very complicated and recent work (meaning the tools required are not for everyone), so it's more research than anything else. Most programmers don't really have a clue of how machine learning works when it works well enough, so that's really few people. And so far, what applications in the real world are we seeing that are good enough and are actually useful to justify the investment? You cannot really trust self driving cars yet, and researchers don't have an insight on a deep neural network and how to use those results. I have the feeling deep learning is another excuse to justify selling hardware. I wish there was more research done at a higher level in Neuro sciences, to have a better definition of general intelligence.