Comment by amelius

Makes me wonder, how well could an LLM-based solution score on the Netflix prize?

https://en.wikipedia.org/wiki/Netflix_Prize

(Are people still trying to improve upon the original winning solution?)