Steve Yegge

The 18-month hiring freeze detail is the buried lede here. Yegge isn't just describing adoption curves; he's flagging that the normal mechanism for spreading institutional AI knowledge — onboarding new hires who already use these tools — has been severed across the sector simultaneously.

The 20/60/20 adoption split Yegge describes from inside Google maps almost exactly onto the fragmentation story MIT Technology Review and Stanford's AI Index surfaced on April 13 ('Want to understand the current state of AI? Check out these charts'). Stanford's data showed conflicting signals on AI's workforce impact, and Yegge's internal Google snapshot is a concrete data point behind that ambiguity. The 60% sitting in chat-interface middle ground also helps explain why AI traffic to retailers surged 393% in Q1 (per Adobe data covered April 16) without producing the kind of deep agentic integration that power users describe — most people are still in a prompt-and-respond pattern, not delegating multi-step work.

If Google lifts its hiring freeze in the next two quarters and new-hire AI fluency rates measurably exceed the current 20% power-user ceiling, that would confirm the freeze itself (not culture or tooling) is the primary adoption bottleneck. If the ceiling holds post-hiring, the problem is structural.