The artifact stopped proving seniority

Happy Tuesday.

Design Twitter hit the same nerve last week from three directions.

• Gokul Rajaram argued that product design is getting compressed as a standalone function.

• Josh Puckett made the more useful split: making mockups is separate from deciding what to build and why.

• Karri Saarinen named the mechanism underneath it: AI looks most impressive when you don't know the subject, don't care about the result, or don't have a clear target.

I heard the practical version in my conversation with Hannah Ahn at Superpower.

Their design team is about five people. Claude Code rolled out across the team in January. A video lead with no coding background built a Premiere Pro plugin for shared media access.

The people closest to the friction are solving it independently.

I want to dig into what that means for designers.

– Tommy (@designertom)

JOIN US AT OUR CONFIG AFTER PARTY

On Thursday, June 25, 2026, we're taking over 25 Lusk in San Francisco for the Detach After Party.

It's the night Config 2026 wraps, which means everyone's already in town and slightly fried from conference talks and activities.

I'm co-hosting with Jesse Showalter, Ridd, and Ioana Teleanu. ~400 designers, builders, and friends in one room.

I'm not going to be humble: this is hands-down the best networking opportunity in the design community, and we sell out every year. Don't wait.

RSVP → https://luma.com/usxsrlu1

The Artifact Used to Carry the Signal

For a long time, design seniority had a visible proxy.

You could look at the artifact.

The portfolio was clean and flows were polished. The prototype looked like someone had the reps. That was never the whole job, but it was a useful signal.

Now that signal is weaker.

A clean artifact used to prove reps. Now it mostly proves a tool got you to plausible. The work still matters, but it stopped carrying enough proof by itself.

The first rungs were artifact rungs: make the screen, polish the module, turn a messy brief into something presentable enough for critique. But a junior can now produce senior-looking work before they have senior judgment.

Not ideal.

Superpower makes this less theoretical because their healthcare product has real stakes. If the experience miscommunicates a lab result, someone can misunderstand something about their body. Average UX is not cute there.

So Hannah's AI adoption story is not a demo reel. Claude Code gave the team more surface area: dashboards, workflow tools, copy passes, review systems, little internal utilities. Suddenly the person closest to the annoying thing could build the fix.

Good. Also a managerial headache.

Because once everyone can make the artifact, the senior job moves upstream. It becomes deciding what should exist, who should make it, what standard it has to meet, and what gets killed before it wastes the team's attention.

That is the new proof.

Hannah's hiring lens points there. She is not trying to clone one AI-native designer across every seat. One person might build tools. Another might own brand, trust, psychology, or the last 1% of polish. The question is not just, "can this person make impressive work?" It is, "what does this person add to the composition of the team?"

If a junior on your team can produce senior-looking work tomorrow, what does your seniority actually rest on?

Clean craft, polished artifacts, and tool speed all got easier to fake.

The proof moves to judgment.

Can you name the constraint? Can you kill the tempting thing? Can you tell when average UX is not enough? Can you compose the team instead of just the output?

That is harder to fake.

The obvious reaction is to make juniors more technical. Fine. AI literacy is becoming baseline. But Hannah's stronger hiring signal was simpler: love of the game.

If you love the game, you stay close to the tools. You try things before they are required. You build, fail, and build again because you are obsessed with making the work better.

Good luck faking that.

So if you are hiring designers this year, spend less time asking whether the portfolio looks senior and more time asking how the person thinks before the artifact exists.

And if you are early in your career, do not wait for the ladder to be rebuilt for you. Pick a real problem. Build the artifact. Then show the judgment around it: what you tried, what you killed, and where AI only got you to average.

These are tools I actually use, so I asked them to sponsor the newsletter. They said yes. The best way to support us is to check them out 👇

• Framer → How I build websites without code

• Mobbin → How I find design patterns fast

• MagicPath → How I design in canvas

• Dscout → How I run user research

That's it for today.

Friday I'm covering the surface area of new media, and what it means to create work that stands out in the state of social distribution.

See you then,

Tommy