How I Bet on Agents Before Anyone Cared | Candid Cohort Conversations

This Friday is the 200th free session. A question from a student in class explains why Mahesh Yadav keeps doing them. The student asked how he positioned himself in agents years before they were popular. His answer: he didn’t. He made bets, and some of them were wrong. In this candid cohort conversation, Mahesh covers: → Why 10 to 11 years in one field lets you see roughly three years ahead → The bets that didn’t pan out, including IoT and knowledge graphs → Why agents paid off → World models, and why he is putting most of his time there next → Why he won’t build a course on it yet: people don’t pay for the future, they want answers for next week’s interview That is what 200 free sessions are: 200 small bets made in public. Each one is a room where I try an idea, get corrected, and teach what I see right now. You don’t do that 200 times unless you like the people in the room. I do. For the 200th, I picked a bet that is already paying off: JEV. 200th free session: Evaluate AI with JEV, when classification beats an LLM Friday, October 16 | 9:30 PM IST / 9:00 AM PST Sign up: https://maven.com/p/b5dd86/evaluate-ai-with-jev-when-classification-beats-an-llm