AI didn’t fail design. The bar did.

"Design is the LEAST AI-productive function in most product orgs!"

The uncomfortable truth is that’s NOT a tooling problem.

In Lenny Rachitsky's AI productivity survey, design reported the weakest gains.

My take:

Design is typically <10% of the overall product cost,

poorly understood,

and lightly measured.

So it flies under the radar 😊

When scrutiny is low, urgency is low.

And when urgency is low, leverage doesn’t get built.

As a result, AI is often under-leveraged where it actually creates value:

Decision and craft.

Instead, it’s used safely:

– Basic prototypes

– Quick summaries

– More options

That’s convenience, not productivity.

For our team this year, the focus is on:

- Speed

- (And) Sharper decisions

– (And) Higher craft

If AI doesn’t change how we decide or what we ship, it doesn’t count.

AI didn’t fail design.

Design hasn’t been held to the same bar.