The high quality stock market data is guarded by data monopolies and hedge funds. Monopolies enjoy being just that, and hedge funds with an information edge aren’t about to part wi…
… to a gold standard, or to other AIs that it is learning to replicate, or has already internalized. You won’t have to design a loss function. It might take a while, but the AI will figure out its own evaluation rules.
…is “actor-critic” approach to learning, and whether this is really true or just a cute analogy. But it does appear that we learn more effectively when we are able to try something new, and get immediate feedback from an expert.