Something exciting is happening on empowered product teams adopting AI: each person on the team is getting more capable, and the overlap between roles is growing. Same team. Bigger reach. Richer overlap.
On every empowered product team I work with, the pattern is the same. The team doesn't change — what each person on it can do does. Product managers clarify faster. Engineers implement faster. Designers stay in the creative loop longer. And the ground the team can cover together grows.
The metaphor I keep coming back to is hockey. Imagine putting bigger, faster, stronger players on a junior team. Same positions on the ice. Same set plays. But suddenly the team can compete at a level it couldn't before. That's what AI is doing for product, engineering, and design — when the team is empowered to use it well.
AI didn't add a fourth circle. The team didn't grow. What grew is what each person on the team is capable of doing — and how much shared territory the team can cover together.
What changes for each role
- ProductClarifies faster. Drafts specs faster. Prototypes before pixels are touched. Spends more time with customers and less time formatting documents.
- EngineeringImplements faster. Pair-programs with an AI that already knows the codebase. A mid-junior engineer ships at senior speed because the senior is in the editor with them.
- DesignStays in the creative loop longer. AI handles the rote — first-pass screens, copy variants, asset prep — so design spends time on the parts that need human taste.
What doesn't change
The fundamentals. We still need to know why we're building something. Who we're building it for. What problem we're solving. AI doesn't answer those questions — it just helps the team execute on them faster.
That's why the multiplier only works when the team is already empowered. If a team is grinding through a backlog of feature tickets with no clear connection to outcomes, AI just makes them grind faster. If a team owns its outcomes — knows what it's trying to move and why — AI is the most capable colleague that team has ever had.
AI isn't your team's replacement. It's the most capable colleague your team has ever had. Lead empowered teams accordingly.
The empowered-team test
If you want to know whether your team is ready to compound with AI or just churn faster with it, ask three questions:
Does the team understand the why? Not the feature list. The actual outcomes we're trying to move and why they matter to the business and the user.
Does the team own outcomes, not output? When a ticket ships, is the question "did we close it" or "did it move the metric we thought it would"?
Can the team act without waiting in a telephone chain? When a designer has a question about a product decision, can they answer it themselves — or do they wait for a meeting?
If all three answers are yes, AI is going to multiply that team. If any answer is no, fix that first. The tool isn't the problem — the operating model is.
I lead empowered product teams that ship consumer products people come back to. AI is the most exciting thing to happen to that work in 20 years — not because it changes who's on the ice, but because it changes how fast they can play.
— Steve Black · steveandthedogs.com