Yesterday I had one of those calls that starts like small talk and ends like a strategy session you wish you’d recorded.

The leader I spoke with runs a large team at a consumer-facing tech company (keeping it anonymous on purpose). They’ve been tasked with driving AI adoption company-wide.

Not “run a pilot.”

Not “explore.”

Company-wide.

And they’ve actually made real progress: clear pillars (employee productivity, product innovation, customer service efficiency), early wins in customer support automation, and laying the foundation to ship AI features repeatedly, not just once.

Then the most interesting moment: they brought up a term I hadn’t heard before. “AI brain fry.”

The idea is simple and slightly uncomfortable: the people pushing hardest for AI (often executives) may be the ones least exposed to the day-to-day cognitive load of using it. AI creates an explosion of optionality. Optionality creates decisions. Decisions create context switching. And context switching, at scale, is exhausting.

This is why “training” alone doesn’t work. Especially for leadership teams.

Most enablement is designed for individual contributors: write better prompts, summarize docs, draft emails. Useful, but incomplete. Executives don’t need more tactics; they need a new operating rhythm for how decisions get shaped:

• Heartset: Connect AI to the mission and values, and normalize experimentation (including failure).
• Skillset: Identify repeatable leadership tasks (meeting prep, narrative building, stakeholder messaging, scenario planning) and build shared patterns people can reuse.
• Toolset: Yes, roll out the tools. But expect policy, permissions, and governance to lag behind what the tools can do (it always does).

The punchline from the call: adoption often slows the higher you go.

The leaders care about the org, but their days are meetings, judgment calls, and people-work. It's hard for them to make the time to use AI for themselves.

And, ironically, that’s exactly where AI can help most. But you need to design for it, make the time for it and iterate.

If you’re trying to “drive AI adoption,” consider this your reminder: you’re not deploying software. You’re redesigning how humans make decisions.

Alex