Comment by karpathy
9 hours ago
I think many experiencing the same thing but we're way too small drop of a drop in the bucket that it doesn't even resolve in all of their measurements and A/B tests. The vast majority of the "gradient signal" is from normies and their patterns, desires and tastes are different. Basically I think we see a massive deterioration of quality because we are a rounding error.
I can’t reject this line of thought strongly enough.
The vast majority of signal is from a plurality of opinions, and the outcome is an average of those inputs, usually tweaked a bit to make them more addictive.
Your thought that “the algorithm doesn’t serve me personally because I have exceptional taste” is not a unique observation. Everyone who consumes these systems has the same thoughts. Everyone is a rounding error, whether you’re in a knitting community, or you’re a worlds leading expert in deep learning.
When everyone is unique, nobody is?