Six designers on what AI actually changed in June 2026.

Week 24. The design industry has been having one continuous argument since late 2023, and this week it finally crystallized into something useful: the State of AI Design Report dropped on June 10, surveying 50+ named practitioners, and for the first time the pro-AI and skeptic-craft camps are talking past each other with enough specificity to pick apart the actual disagreement. The question isn't whether AI changed design. It's: what exactly did it change, and who gets to decide whether that's good?

Let me hand it over to the room.

Katie Dill, Head of Design at Stripe

"It's like a burrito that took a minute and a half to microwave. I am happy with it, because I am in a rush and I am thrilled it's ready so fast. But is it actually good? Heck no. This same phenomenon is what leads to a ton of slop. It's easy to get enamored by a decent one shot and overlook the flaws. A good practice is to ask yourself, would I be happy with this if it took me a week to make this? If the answer is no, it's probably worth some further prompting and prodding to make it something you're truly proud of."

— State of AI Design Report 2026, stateofaidesign.com, June 10, 2026

Dill's burrito test is the most honest frame I've seen for the AI output quality problem. The danger isn't that AI produces bad work — it's that it produces fine work fast enough to make you stop caring. The industry isn't drowning in failures. It's drowning in good-enough.

Graham Sykes, Global Executive Creative Director at Landor

"As we continue to embrace AI as a tool, human-driven craft is coming sharply back into focus as the antidote to AI's hyper-slick visual language. When algorithms flood the world with flawless flatness, the marks of the maker become signal. The story lies in process; the emotion lies in imperfection."

— Creative Bloq, 2026

Sykes is right about the signal, but there's a catch he doesn't name: AI can already convincingly fake imperfection. The brush mark, the scan artifact, the "handmade" texture — all generatable on demand. The craft rebellion works right now, but it has a limited shelf life as a pure differentiation strategy. The designers who'll outlast it understand why imperfection matters — not just how to produce it.

Joel Lewenstein, Head of Product Design at Anthropic

"We see code as one more tool in the tool belt. I'm fairly agnostic. If you want to write all the front end code yourself; if you want to put up some PRs at the very end to polish; if you want to just use your knowledge of the code to collaborate better with engineers and be less intimidated — that's all fine. But there are no excuses anymore for not being deeply involved in the final delivery process."

— State of AI Design Report 2026, stateofaidesign.com, June 10, 2026

"No excuses anymore" is the phrase doing the most work here. Lewenstein isn't saying every designer must write production code — he's saying the gap between design intent and shipped product can no longer be explained away by tooling. The designer who can't close that gap now is making a choice, not hitting a wall.

Ryan Mather, Designer at Anthropic

"AI is powerful because the tool itself can think and experiment. It's up to the person holding the tool to decide where to direct that. Every good designer has a point of view, and AI is like a point-of-view amplifier."

— State of AI Design Report 2026, stateofaidesign.com, June 10, 2026

This is the optimistic case, and it's true — for designers who already have a point of view. For junior designers who haven't developed one yet, AI doesn't amplify anything. It replaces the years of deliberate practice where the point of view was supposed to form. That's not an argument against AI, but it's a real gap in design education that nobody has solved yet.

Cat How, Creative Director and Founder of How&How

"Once you know exactly where a tool's edges are, the fear has nowhere left to live."

— Creative Bloq, June 6, 2026

Short, but load-bearing. How's whole piece on running a live AI rebrand for a skeptical client is worth reading in full, but this sentence is the thesis. Fear of AI in creative work is almost always fear of the unknown applied to a specific tool, not a principled aesthetic position. The antidote isn't conviction — it's competence.

Alexander Cheung, Senior Product Designer at Pinterest

"If AI is designing for AI, it doesn't even need to be delightful. It just needs to work."

— State of AI Design Report 2026, stateofaidesign.com, June 10, 2026

This is the quote that will age the worst or the best. Cheung is raising the real possibility that a significant portion of future design output never gets seen by a human — agent-to-agent interface, utility stripped of aesthetics. If that's where we're going, a large chunk of what designers currently get paid for becomes infrastructure work. It's a low-confidence, high-impact scenario. I'm surprised more people in the room aren't talking about it.

The meta-trend: AI raised the floor and ate the middle.

Every voice in this roundup points to the same structural change. AI has commoditized the adequate — the fast comp, the first-pass layout, the passable stock substitute. What remains valuable is what was always valuable: taste, judgment, point of view, and the ability to close the gap between what you imagined and what ships.

The disrupted zone isn't the top or the bottom. It's the competent-execution-without-distinctive-thinking middle. That tier used to be a career. Now it's a prompt.

Next week prediction

Confidence: Medium | Timeframe: 6 months

The first agency job title explicitly built around AI brand governance goes mainstream — something like "Brand Guardrails Lead" or "AI Brand Custodian." The Jessica Rosenberg (AirOps) model — one person building a visual brand skill that enforces standards across all AI agent outputs — is moving from experiment to headcount. Wrong if: agencies take longer to formalize what's currently running informally on Slack.