Translated Strategy · · 12 min read

Your Management Experience Is an AI Superpower

Mollick's research shows that management skills — scoping problems, evaluating output, knowing when something's off — are exactly what makes AI powerful. If you've been running a business, you're already qualified.

Source material Ethan Mollick — 'Management as AI Superpower'
AI in Crayon · ep
Why managers beat prompters

You’ve probably noticed something weird happening. Every week there’s a new AI tool, a new capability, a new reason to feel like you’re falling behind. And if you’re running a service business — HVAC, insurance, a medical practice — the message from the tech world feels pretty clear: learn to code or get left behind.

Here’s the thing. That message is exactly backwards.

The research that changes everything

Ethan Mollick, a professor at Wharton, ran a study with BCG consultants. They gave half the team access to GPT-4, half no access, and measured output on real client tasks. The group with AI crushed the control on every dimension that matters — speed, quality, creativity.

The consultants who used AI well weren’t the best prompters. They were the best managers. They scoped the work, evaluated the draft, pushed back, and shipped.

Read that back. The consultants who used AI well weren’t the best prompters. They were the best managers.

Why this matters for you

If you’ve been running a 15-person service company for ten years, you have done exactly one job, every day, for a decade: decide what good work looks like, hand it off to a human, evaluate what comes back, and ship it. That is the skill.

  • Scoping the problem. You already do this when you brief a new tech.
  • Evaluating output. You already do this when you inspect a job.
  • Knowing when something’s off. You already do this with employees.

The Monday test

This week, try this: take a task your team does repeatedly — scheduling, customer follow-up, report generation — and spend 30 minutes with ChatGPT trying to build a workflow for it. Don’t worry about getting it perfect. Just notice what happens.

You’ll find that the hardest part isn’t the AI. It’s scoping the problem correctly. And that? That’s the skill you’ve been building for years.

Keep going

AI in Crayon · ep
The 18-month squeeze
Up next
The Shift

Where to Bet on AI Tools (And Where Not To)

Most AI tools you're being sold today will get squeezed out of existence in 18 months. Here's the filter I use with my own clients to separate real leverage from tourist traps.