I have written prompts for stock, ETF and royalty trust evaluation. Recently, I asked if my prompts were efficient. Have you measured how many tokens you use and do you evaluate your for efficiency? The response I got from several pedestrian AI LLM was that I could be more efficient if I attached annual reports and the like instead.
VanEck Uranium + Nuclear ETF ($NLR ): Nuclear pure-play (CEG, utilities); +43% YTD; uranium supply chain upside.
I’d add one more step. Keep a source log for every number and claim so it’s clear what came from the filing and what came from the model.
I have written prompts for stock, ETF and royalty trust evaluation. Recently, I asked if my prompts were efficient. Have you measured how many tokens you use and do you evaluate your for efficiency? The response I got from several pedestrian AI LLM was that I could be more efficient if I attached annual reports and the like instead.