AI is a craft problem, not a replacement problem

AI won't replace designers. It raises the bar for craft. When the cost of generating mediocre work drops to zero, taste and judgment become the differentiator.

Key takeaways

• AI makes output cheap. It does not make taste cheap.
• The bottleneck shifts from production to curation.
• Treat AI like a collaborator: brief, feedback, iteration.
• The baseline rises. There are fewer excuses to ship bad work.

---

The real question: quality when mediocrity is free

People keep asking if AI will replace designers. They're asking the wrong question.

A better one is: what does quality work look like when the cost of generating mediocre work drops to zero?

We've been using AI heavily in our workflow for over a year now — not as a novelty, but as a genuine collaborator. Sinyo and I use it to prototype quickly, pressure-test briefs, draft strategic frameworks, and generate first-pass copy that we inevitably rewrite. It's in the room on almost every project.

---

Why taste becomes the advantage

And what I've noticed is this: the value of taste has gone up, not down.

When anyone can generate 50 moodboards in an hour, the skill isn't generating the moodboard. It's knowing which one is right, why it's right, and how to explain that in a way that moves a client.

When anyone can scaffold a working prototype overnight, the skill isn't the scaffold. It's the judgment call that determines what to build, and the restraint to stop when it's done.

This connects to something my partner Arctic Fever keeps saying: "We become the curator." When generative tools flood a field, the raw material becomes cheap. The curation becomes the value.

---

The taste gap (definition)

The taste gap is the distance between what a tool can produce and what's actually worth shipping.

That gap exists in every medium — it's why not everyone with a camera is a photographer, why not everyone with a computer is a software engineer, and why not everyone with Figma is a designer.

AI widens the territory a single person can cover. It doesn't close the taste gap.

---

Vending-machine AI vs collaborator AI

The designers I've watched struggle with AI are the ones who treat it like a dispenser. You put a prompt in, you get an output, and you pick the best-looking one. That's not a design process. It's a vending machine relationship. And it produces vending machine work: technically fine, spiritually empty.

The designers who are thriving are the ones who treat AI the same way they'd treat any other collaborator: with a brief, feedback, iteration, and opinions. They argue with it. They reject things. They push back. They use it to find the edges of their own thinking, then do the genuinely hard work of synthesizing something coherent.

---

What this means for designers (and studios)

There's something BR-ND People — one of our close partners — embodies that I think about a lot. Their whole model is built on the idea that "the people make the business." Not the strategy deck. Not the brand system. The actual humans who carry it.

That principle applies here too. A studio's taste, its judgment, its ability to ask the right question at the right moment — that isn't a prompt. That's accumulated experience, and no tool replaces it.

AI makes the individual more capable. It doesn't make the individual less necessary.

I don't think "will AI replace designers?" is a meaningful question right now. It's too binary.

A more useful frame is this: AI is a material or tool, like code or print. Learning to work with it well is a craft skill. Ignoring it is a choice, but it's a choice with real costs.

For what it's worth, we're not worried about being replaced. We're more interested in what it makes possible — and what it makes inexcusable. There are fewer good reasons to ship bad work now. The baseline is higher.

That feels like the right kind of pressure.