Every pitch deck right now says the same thing: AI will make design 10x faster. They're right. But they're solving the wrong problem.
Speed was never the bottleneck
In twenty years of leading design for complex products, no one has ever told me, "We failed because the mockups took too long." The failures that actually cost money — the ones that set teams back months — happened because people built the wrong thing and felt good about it the whole time.
AI makes production faster. It doesn't help you figure out what to build.
What judgment looks like in practice
Here's a real scenario. The VP of Engineering tells you the architecture can't handle your ideal flow. You have two choices: fight for it and delay the release, or find the version that's 80% as good and ships on the existing system. The right answer depends on a dozen things that aren't in any design file.
The most expensive design decision is the one that feels right in the review but falls apart when engineering tries to build it.
That kind of call comes from experience. From watching things go wrong enough times that you start to recognize the patterns early.
AI actually makes this more important
When anyone can generate a polished screen in a few minutes, the hard questions move upstream:
- Problem framing — Are we even solving the right problem?
- Constraints — What are the real limits? Technical, organizational, regulatory?
- Alignment — Can we actually ship this, given everyone who has to agree?
- Sequencing — What do we build first, and what do we leave out on purpose?
These are the questions that decide whether a product ships or just looks good in a demo.
What this means if you're hiring
Don't hire design leaders for speed. Hire the person who's been in the room when things went sideways and helped the team find a way forward.
That's the work that matters. AI can't do it yet. And honestly, when it can, we'll all have bigger things to think about.