Y
Hacker News
new
|
ask
|
show
|
jobs
by
jelling
252 days ago
Same. We've all fooled ourselves into believing that an LLM / stochastic process was finally solved based on a good result. But the sample size is always to low to be meaningful.