Three posts landed this week that each describe a different piece of the AI adoption puzzle. Zapier's hiring rubric. Ramp's company-wide playbook. Karpathy's explanation for why two groups of AI users are talking past each other.

I lined them up and saw the same pattern I keep seeing in enterprise: companies pick one of these three and ignore the other two.

→ They hire for fluency but never grow the team they have.

→ They build a proficiency ladder but do not change the hiring bar.

→ They do both, but leadership sets the budget from Group 1 while practitioners operate in Group 2.

The sharpest framing came from Zapier: "With AI, you can delegate the work, but not the accountability."

I wrote about what the three playbooks look like together, where each one breaks on its own, and the diagnostic I use to figure out which one a company is missing.

Full piece is on Substack. Link in comments.