> A human might perform better on a 100 entries than an AI, but on 10 thousand? There's also increased variance in human accuracy. You might train 100 but 10k

> A human might perform better on a 100 entries than an AI, but on 10 thousand? There's also increased variance in human accuracy. You might train 100 but 10k people? A model is consistent all the way.