Honestly if you can I would try to find something outside "AI" for now: even if you are doing real deep learning stuff the reality is that 99% of the people investing in AI at the moment can't differentiate obvious bs from real R&D at all, they are here just because of the hype and are throwing millions to anything with an AI tag on it, so when this bubble pops it will be ugly for everybody, indiscriminately. This is another dot com.