Not so fast on AI in finance

 From what I have seen, school finance (as a sector) has not been immune from the FULL STEAM AHEAD on generative AI adoption. 

This would be alarming in any case, but this recent piece of research posted through the Public Enterprise LLM Benchmarks collection on the accuracy of generative AI in finances makes it even more so. I'd recommend reviewing the entire piece, of course, but the headline, so to speak, is this: 

We find that none of the existing AI models exceed 50% accuracy, indicating that models still have a long way to go before they can be deployed reliably and trusted in the finance industry.

And note, as the Times writes of today,  AI hallucinations are getting worse, and even the companies developing them don't know why.