Did you create an account, just to ask this question?

I don't think neither CV nor NLP is going away. CV is yet to be solved to the same extent as NLP, but I agree it might just be a matter of time.

Research wise, there are still tons of problems around uncertainty, complexity, causality, 'real-world' problem solving (domain adaptation) and so forth.

Just don't compete on having the largest cluster of GPUs.