Scale and Operations in AI Training

Why experts writing AI evals is creating the fastest-growing companies in history | Brendan Foody

Topic 6 of 8

Published Sep 18, 2025 · Clip 29:56–39:00 (9:04)

Post-training uses experts for reasoning accuracy via evals and RL, like radiologists scoring X-rays. Scale involves tens of thousands working on evals/RL environments. Competitive edge from high-skill shift; Merkor hires for projects at $95-$500/hour. Examples include Harvard Lampoon for humor, screenwriters for creativity. Top 10% drive most impact, emphasizing talent identification.