AI Came for the Smartest Jobs First

Two AIs score high enough to win gold at the International Mathematical Olympiad—while robots worth hundreds of thousands of dollars still struggle to fold a T-shirt. That strange contrast reveals something important: AI is not replacing work in the order we expected. The jobs many people thought were safest—programming, law, analysis, writing, and other knowledge work—are among the first to be affected, while many physical and caregiving tasks remain surprisingly difficult. Using Moravec’s paradox, AI training data, and recent labor-market research, this video examines why. It explores why knowledge work is easy for AI to imitate, why code and mathematics are easy to check, why entry-level workers are taking the biggest hit, and why the disappearance of junior roles could create a dangerous gap in future expertise. We also look at the caveats. Layoffs are influenced by interest rates and overhiring, companies may be “AI-washing” job cuts, and AI often replaces individual tasks rather than entire jobs. Are trade jobs, caregiving, and human judgment the new safe bets—or will robots eventually learn those skills too? The biggest question may not be whether AI takes every job overnight. It may be what happens when AI removes the first step of almost every career. Chapters: 00:00 AI Wins Math—But Can’t Fold a T-Shirt 01:13 The Old Automation Story 02:39 Moravec’s Paradox 03:56 Why Knowledge Work Is Easy to Automate 05:28 Who Is AI Actually Replacing? 06:22 The Data: Freelancers and Young Workers 08:04 Why Entry-Level Jobs Matter 09:15 AI-Washing and the Limits of the Data 11:02 Are Trade Jobs the New Safe Bet? 11:47 When Robots Enter the Physical World 12:10 What Counts as Intelligence? 13:13 What Happens Next? 14:32 The Smartest Jobs Were Hit First