I had a business owner tell me my company was obsolete because of AI

I had a business owner tell me to my face that Nordic Design Collective was obsolete. He said there was no more need for a design and development studio because AI could do all of it now. Web design, branding, copy, code. All of it.

He was wrong, and I have spent a lot of time since then thinking about why.

The version of the AI conversation that landed at his table is the same version that lands at most tables. AI is going to replace whole categories of work. Designers. Developers. Writers. Anyone whose job is to make a thing. The tools are getting good enough fast enough that the people who used to be paid to make the thing are not going to be paid much longer. He believed it. He was not the first person to tell me a version of it, and he will not be the last.

The argument has one true premise and one false conclusion. The true premise is that creating things has gotten dramatically more accessible. Anyone can now generate a website, a logo, a brand guide, a chunk of working code, a marketing video, in minutes. The barrier to producing something is lower than it has ever been. That part is real.

The false conclusion is that this eliminates the value of the people who used to do that work professionally. It does the opposite.

What actually happens when the barrier to creation drops is that the market gets flooded with output that looks right and is structurally broken in ways the person who shipped it cannot see. A website goes live with a contact form that does not send to anyone. A storefront launches with payment processing that does not actually charge the cards. Application code stores user passwords in plain text where the first person who pokes at the database can harvest them. A brand identity gets built with three fonts that already belong to three other companies in the same market. A marketing email goes out with no authentication, the inbox providers send it straight to spam, and the business never figures out why their conversion rate dropped to zero. The failures look like the working version until they do not.

This is what the business owner did not understand. The fact that anyone can now generate a website does not mean every website now serves its business. It means the gap between a website and a website that actually does the job got wider, not narrower. The tools moved the floor up. The ceiling moved up too. The space between is the expert's job.

When everyone can produce, what gets paid for is what is good. What gets paid for is taste. What gets paid for is the judgment to know which of the ten things the model generated is the one that solves the problem. What gets paid for is the experience to recognize that the convincing-looking output is structurally wrong and to fix it before it ships. None of those moved. AI did not eliminate the need for expertise. It made expertise the only thing left that differentiates output.

The designers and developers I talk to are not really worried about AI replacing them. They know what their job actually involves. What they are worried about is the person who signs their paycheck, a manager, a founder, a client, looking at an AI demo, deciding it covers the job, and acting on that decision. The fear is not the tools. The fear is the person above them in the org chart who does not understand what is actually inside their work mistaking the demo for the job.

The reality is the opposite of what those decisions will produce. The people with expertise are going to have more leverage than they ever did before, because they can produce at a pace that used to require a team, and what they produce will be the output that does not look like everything else.

The work is not going to dry up. The work is going to get harder to do badly and easier to do well, if you know what good is.

A year from now, two years from now, ten years from now, businesses are going to be sorting through more AI-generated output than they can use. The ones that hire someone with taste are going to stand out by such a margin that the question of whether they needed to hire that person at all will not come up. It will be obvious from looking at the result.

NDC is not obsolete. The version of design and development that was a commodity in 2023 is obsolete. The version that is taste, judgment, and the experience to use the new tools without producing slop is more valuable than I have ever seen it.

That is the conversation I wish I had been able to have with the business owner who told me my company was obsolete. I have it with myself most weeks instead.