Extreme Harness Engineering for Token Billionaires: 1M LOC, 1B toks/day, 0% human code, 0% human review — Ryan Lopopolo, OpenAI Frontier & Symphony
Latent Space: The AI Engineer Podcast
00:00
Use Long Explicit Prompts To Avoid Babysitting
- Give agents long, strict instructions and explicit success criteria so humans do not have to babysit them while they work.
- Ryan Lopopolo links this to disposable code and reruns, arguing that reliability rises when deployment conditions are tightly specified upfront.
Play episode from 50:14
Transcript
