A company founder becomes tech support for their own IT solutions

An interesting pattern has appeared, and it caught me too. When I realised that AI could build practically any digital solution, I immediately remembered my failures with developers: websites, bots, funnels, even simple landing pages. Several projects, money buried in them, and constant misunderstanding.

And then came the green light: finally, I could build for myself what I had spent so much time fighting with IT teams over.

You dictate to an agent what kind of website you want, and you get a result. Quite a good one, too. It creates the impression that explaining it once is enough, and a wonderful service is ready.

Something like the casino effect: you win the first time, get a huge dose of dopamine, and are shown how easy it is. Then comes a series of losses.

I went through roughly this picture over the past two years. A year ago, I also thought that I could do everything myself now and developers were no longer needed. But the past year brought a completely different side into view.

Between “AI created it” and “this is a finished solution that I like and that works reliably,” there is a stretch many people estimate at about twenty percent.

In my experience, it is exactly the reverse. Putting together a website is perhaps 10 to 20% of the work. On the outside, it already looks almost finished. The biggest part begins with the finishing touches, adjustments and revisions. Connecting functions, testing, design, stability. All of this takes time, patience and a deep understanding of the system.

And here the business owner quietly turns into one of their own employees or contractors.

Previously, the obstacle was having to write code by hand. Now you do not have to, and it feels like you have magic in your hands. But if you can drive a car, why not get behind the wheel of a truck yourself and deliver all the goods? You can do that too.

Building a service is simply more interesting.

Instead of organising the business, coordinating departments, researching the market and competitors, the founder connects bots and keeps refining spreadsheets. As if a grown man has replaced World of Tanks with building digital products. Then he becomes their support team himself. There is no more time, and the business has not become better.

I am speaking from my own experience. When I had a business with many employees, I commissioned a website that was supposed to connect people, services, clients and intermediaries. Over six months, I paid about $30,000. I received some crooked, barely working code that needed just as much money again, with no clear finish in sight. There was no finished solution.

Now I understand that I could put something like that together in a few hours with an agent, practically for free. But bringing it to a finished product is a separate, substantial job.

And this, I think, is where developers' prices and terms of work need to be reconsidered. How much code was written, how rare the language is, and how attractive the prototype looks: what difference does that make now?

You should pay for a result that works the way it was ordered. You still have to reach that result. And decide who will do it: a team of professionals or a founder who has created another job for themselves.

An entrepreneur friend described his experience like this: “I launched OpenClaw, it was sooo interesting, built a website and a couple of bots, then everything started breaking, I had to repair it, it pulls you in like crazy, so I left it for now.”