saltsucker
Also just think about it, why would a model trained on the world’s corpus of text (that isnt formatted in xml) perform better with XML? It would be a better study if that post tested markdown, org, xml, json, etc. 10 times to see if their is a difference
If I understand some of it correctly, I was contemplating this when I started writing functions for "single functional concepts" like, "check for X; return true or false", then called each of those functions sequentially in a single "run" function. Is that what you mean?
I found that approach much easier to test the functions and catch bugs, but your comment seems to go against that.
Even music apps are disappointing. Sometimes they do well, but most of the time it seems not. I play a radio station for Mat Corby, which is a pretty chill downtempo vibe, and the app throws in stuff from my library that has no relation--like Kanye West's Jesus Walks. Literally did that multiple times. Those vibes could not be more different.
Maybe cataloging music is a difficult problem, but there was a time (maybe 2010?) when YouTube would efficiently suggest music that had a vibe to what I was listening to, and it helped me find many artists I listed to now.
Edit: And a time when Apple's Genius was not a bad house party DJ