factual
Human annotation trained the vision model
The vision model was initially trained by human annotators who watched student screen videos overnight and labeled learning anti-patterns, which became a reinforcement learning with human feedback loop that has since been mostly automated.
factualpending
Speaker
Joe LiemandtEvidence Quote
“the AI handles almost all of it. We still have some humans where we're still fine-tuning things.”
Created: 11/18/2026, 10:33:53 PM
Your notes
Loading notes...