Something I learned from working alongside data scientists and financial analysts doing algo trading is that you can almost always find great fits for your criteria, nobody ever worries about that. Its coming up with the criteria that's what everyone frets over, and even more than that, you need to beat other people at doing so - just being good or event great isn't enough. Your profit is the delta between where you are compared to all the other sharks in your pool. So LLMs are useless there, getting token predicted answers is just going to get you the same as everyone else, which means zero alpha.

So - I dunno about uniquely human? But there's definitely something here where, short of AGI, there's always going to need to be someone sitting down and actually beating the market (whatever that metaphor means for your industry or use case).

If you're doing like, real work, solving problems in your domain actually adds value, and so the profits you get are from the value you provide.

But "finance" is very broad and covers very real and valuable work like making loans and insurance - be careful not to be too broad in your condemnation.

Also ignores capital gains - and small market moves are the very mechanism by which capital formation happens.

Put another way - capital was accrued long before we had a stock market, and even longer before we had computers deciding which stocks to sell or buy.

It’s a very rubbery, human oriented activity.

I’m sure this will be solved, but it won’t be solved by noodling with prompts and automation tools - the humans will have to organise themselves to externalise expert knowledge and develop an objective framework for making ‘subjective decisions about the relative value of things’.