What's this about?
swyx (Latent Space) described a stark split in AI hiring: a "huge bull market" for AI-native individual contributors and player-coaches, and a "huge bear market" for traditional "head of X" managers — summarized as "1 year experience managing 10 agents beats 10 years managing 10–100 people." Separately he declared per-token pricing a dead cost metric, urging people to measure $/task instead.
What exactly is an "AI-native IC" or "player-coach"?
An IC (individual contributor) is someone who does the work rather than managing people. "AI-native" means they get their leverage by directing fleets of AI agents, not by delegating to human reports. A "player-coach" both does the work and guides others. swyx's claim is that the person who can personally orchestrate 10 agents is now worth more than the manager who coordinates a large human team.
Why is this worth noting?
It's a concrete labor-market read on what agents do to org charts — flattening management value and rewarding hands-on agent operators. The $/task point is related: if one person plus agents can do a team's output, cost-per-token stops mattering and cost-per-finished-task becomes the number that counts.