Product Design in the Age of AI

You might have heard. Last week, Block laid off about 40% of its workforce. Over 4,000 people. Yep, I was one of them.

CEO, Jack Dorsey framed it as a shift to an "AI-first" model. Smaller teams, intelligence tools, more output per person. The stock surged 24%. The market loved it.

I'm not here give my opinion on whether that was the right call. I'm writing this because I think it's the start of something much bigger. And it's something that the design community - my community - needs to talk about. Clearly, without either panic or denial.

The new desktop publishing

When Gutenberg invented the printing press, the scribes weren't thrilled. After centuries of careful, hand-copied manuscripts, suddenly anyone with a press could produce a book. The craft didn't die, of course. But it changed completely. The monks who'd been copying Bibles by hand weren't the ones who invented the newspaper.

Fast foward to the 1980s, and desktop publishing put typesetting into the hands of anyone with a Mac and a copy of PageMaker. The design industry panicked. And yes - a lot of truly terrible pamplets were born. But the long-term result wasn't the death of graphic design. It was an explosion of visual communication. More people making things. More experiments. More chances for great work to emerge.

AI is doing the same thing to software. Cursor, Claude Code, V0, Replit, Loveable - they're all collapsing the distance between idea and working product. A designer can ship a native app in a weekend. A founder can prototype a SaaS tool without hiring an engineering team. A kid with a good idea and some persistence can build something real.

The barrier to making software has never been lower. And I think that's genuinely exciting - even from where I'm sitting right now.

Moving fast, sometimes in the wrong direction

Here's the tension, though. Speed without direction is just expensive wandering.

AI makes it incredibly easy to move fast. It also makes it incredibly easy to move fast in the wrong direction. You can scaffold an entire product in an afternoon and only realise a week later that you solved the wrong problem.

In the rush for speed, discovery activities are often sidelined. They feel too slow in an AI-paced world. When engineers can spin up seven coding agents and ship a working prototype before you've finished exploring options in Figma, it's tempting to skip straight to building. But if you ignore discovery - if you skip the part where you fall in love with the problem - there's a much greater chance you'll build the wrong solution. Beautifully. Quickly. And completely wrong.

Here's what also worries me. I don't think the current pace is sustainable. People are spreading themselves thin, across too many agents, too many concurrent streams of work. It's leading to a kind of shallow competence - fast outputs with a fragile understanding of what's underneath. When nobody truly understands the logic behind the products they're building, mistakes compound. And burnout is inevitable, unless we're deliberate about preventing it.

AI helps us move fast. But we also need to slow down for the deep thinking and direction-setting that makes speed worthwhile. The two-year or five-year vision deck might be dead, but the need for vision isn't. It's just compressed into shorter horizons - three to six months, maybe and shaped more like a working prototype than a polished presentation. Someone still has to decide what's worth building and why. AI compresses execution. It doesn't compress judgment. If anything, it puts more pressure on it.

The taste trap

I've seen a lot of designers reaching for "taste" as the thing that keeps us safe. AI can generate, but it can't curate. It can produce, but it can't feel. Taste is uniquely human. We're fine.

I think there's truth in this. But I also think we're holding onto taste a little too tightly as our life raft. AI's sense of taste will get better. It'll get better at judgment, at understanding context, at many of the qualities we've assumed were uniquely human. That doesn't mean we become irrelevant. But it does mean taste alone isn't the moat we think it is.

So what does keep us valuable? I think it's something broader than taste. It's instinct. The ability to make something that feels right. Knowing which details matter and which don't. It's sweating the details, once in Figma, but now in code. It's the last 10% that makes someone trust your product, come back to it, recommend it to a friend.

All of this comes from years of caring deeply about the people using what you build. And that's not something you can shortcut with a faster model.

Interfaces everywhere, owned by no one

Here's the bit that excites me most.

I think we're heading toward a future where interfaces are omnipresent but ephemeral. Bespoke. Disposable, even. Software that's generated on the fly, shaped to a specific context, and dissolved when it's done.

Think about it. If building an interface takes minutes instead of months, why wouldn't every team build their own tools? Why wouldn't every workflow get its own UI? We're already seeing it - people spinning up internal dashboards with Claude, creating one-off data tools with V0, building personal Swift UI apps that only they will ever use.

I think we're going to see an explosion in the amount of software being made. Not all of it polished. Not all of it good. But a lot of it genuinely useful in ways that mass-market software never could be. Home-grown interfaces, built for a party of one.

This is the desktop publishing moment for software. More people making things. More experiments. A lot of questionable typography along the way. But ultimately, more surface area for brilliant work to emerge from unexpected places.

And this is where designers have a massive opportunity. Someone still needs to understand how systems should feel, how information should flow, how trust gets built - through shipping, learning, responding to feedback, and shipping again.

The fear is real. The opportunity is bigger.

I won't pretend the fear isn't personal right now. I was laid off last week. The company I gave years to decided that AI means they need fewer people like me. The stock went up. That stings a bit.

But I keep coming back to this: the scribes didn't invent the newspaper. The typesetters didn't invent desktop publishing. The people who thrived in each revolution weren't the ones who clung to the old tools. They were the ones who understood the new material deeply enough to shape what came next.

I must admit - even through all of this, I've been having the most fun I've ever had in my career. I'm building native iOS apps on weeknights. I'm prototyping ideas that would've taken me months, in hours. The tools are genuinely magical if you let yourself play with them.

So yeah, it's a weird moment. Things are shifting fast. Some of the ground under our feet isn't solid yet. The industry needs to be honest about the human cost of moving this fast, even as we embrace what's possible.

But I've never been more excited about the future of product design. We're not just designers anymore. We're builders too. I want to focus on building what's next.