Algorithmic ranking is unfairly maligned

I built a personalised reccomendation algo for a specific part of my RSS feed a little while back.

I was subbed to the RSS feed for arXiv, and used a small local LLM model to rank the research papers relevance to my research based on reading my research bio and the paper abstract.

Works pretty well, but im not sure how to generalize it beyond that niche application.

https://caffeineandlasers.neocities.org/blogs/ChiScraperIntroduction