We need to invest in long term R&D to potentially achieve an ML breakthrough in one of the above fields instead of allocating enormous capital to ML unicorn businesses.

But to do so, we need to first openly admit the truth. ML is not working for the wide range the problems it is currently pitched for.