We live in the 21st century, and technology no longer advances every year, every six months, every quarter, or even every day. It advances every second. All because we've entered the era of artificial intelligence. And soon, perhaps, quantum technologies available to everyone. Speaking of an unreal world, there is no longer anything these tools can't do.
A few years ago, when I saw the first seedlings of AI, I joked to coworkers and people I was close to that in a few years we'd be hiring someone to "write prompts" - a prompt specialist. And you know what? It wasn't that far off. It's 2026, I follow about 15 groups on artificial intelligence, and today writing a good prompt is truly an art.
The sandbox
I've noticed certain tendencies, though. First: many people play with the tool like a child in a sandbox - except it's play that harms more often than it teaches. Most use AI for funny, sometimes plain stupid images or videos, for abuse like deepfakes. Some users approach it thoughtlessly: instead of talking to the tool like an assistant, they try to turn it into "another fool". I read what people write in groups and chats and wonder where we went wrong. Of course, there's also a sizable group of people who use AI properly - wisely, concretely, productively.
Someone once said they'd lose their job because of AI, that it would "eat" professions. I hold a different view: people who stay current with technology and new developments, who aren't afraid of these tools but work with them diligently and conscientiously - won't lose their jobs. I see it in myself. I used to do a few projects a year for my clients; today I do a dozen or more a month. I can safely say AI has increased my revenue by about 500%.
The ERA and the types
Someone observed very accurately that we had the era of programmers by vocation, then the era of Elementor and page builder "programmers", and now we're entering the era of AI programmers. The difference, however, is that the "AI programmer" splits into two types of people.
The first type - and I count myself here - are people "of the trade": schooled, experienced, with an understanding of the subject. People who know the culture of code, know how to write, how to verify, when to check something and whether the tool got it wrong. Professionals who - like a good builder - know what to use to erect a skyscraper rather than a house of cards. I mean programmers who did this before, without Elementors, page builders, and similar crutches, who have knowledge (often enormous) of their field and can generate more value, and a better prompt along the way.
Let me clarify right away. I too have used page builders and built such sites for clients. It came down to budget. But when I did it, I did it consciously, making sure the code was optimized as well as possible!
The second type is the "AI programmer" without knowledge - someone trying to skip the fundamentals. An example, because maybe not everyone feels it immediately: there's me - a programmer, entrepreneur, IT industry person rooted in it since 2010 - and my wife, a pastry chef, who has no idea what exactly I do; for her, technology can feel like "The Matrix". And yet she and I can both take an AI tool and write something similar: "write a program", "build a website". The question is: whose prompt and logic will be better? You don't have to be exceptionally gifted to know the answer. Mine, of course - not because I'm "better", but because I have the fundamentals, the practice, and the ability to verify.
And it works the other way too: if the subject were purely confectionery rather than technology, she would write a better, more valuable prompt and would know where to check the result. Because school, theory, and above all practice taught her that. In that sense, AI is a bit like books: a library alone doesn't make anyone an expert - what counts is whether you can use the knowledge. It's no coincidence that Stanisław Lem, in "Summa Technologiae", argued that tools would grow ever more powerful, but humans would still need to know how to steer them wisely.
Fueling it with garbage
Unfortunately, plenty of people also use AI in a toxic way: they twist the tools' purpose, feed them absurdity, try to turn the technology into a garbage dump, and are proud of it too. Another absolute nonsense is the attitude: "no need to buy paid plans, I'll do it on the free tier anyway". In 2025 I ran an experiment: for half a year I worked on free tools, and for the next half year (and to this day) I spend about €165 a month on tools. The result? I work about 325% more efficiently than on free solutions and earn about 500% more revenue than back then. Of course: not everyone can make use of these tools. Not everyone has the capacity for reasoning, building schemas, algorithmic thinking.
no-Google
Google - in my assessment - has badly sawed off the branch it was sitting on with AI, and in practice turned all of search rankings upside down. Someone put it brilliantly in one of the groups: today we don't so much "search" as get the solution served on a platter. We skip the ads, bypass the wall of content we previously had to dig through to extract a single sentence. Today, typing anything into Google increasingly produces an immediate and - importantly - accurate answer. No need to search. So if you're not in the top five, it often isn't even worth the time. And here I put down a big period.
Coming back to programmers: we know how to build a site and its structure so Google understands what's really on it - and so it surfaces our content in "position zero" and in AI answers. At the same time, I'm disgusted by the approach where someone, despite acquired, enormous, valuable knowledge, gives it away completely for free - step by step - on YouTube or anywhere else. It's a bit like a doctor recording a detailed tutorial on how to perform surgery. Fortunately, not everyone can capture such knowledge and reproduce it identically - and that's exactly why those who "got a taste of the subject" earlier, built the fundamentals, and are genuinely in the industry still win.
Access or progress
My article is meant to show one thing above all: everyone has access to AI today, and everyone can do things limited only by imagination. But not everyone will do it the way a true professional in their field would. So no - you won't lose your job because of AI. AI is a beautiful addition and a powerful multiplier that lets you achieve above-average results. How you use it is up to you. You're limited only by your imagination.
And now the best part. Because most people think a "prompt" is a single sentence. That it's like a spell in Harry Potter: you say "build me a website", wave the wand, and get a finished product. But a prompt is not a sentence. A prompt is a specification. It's a brief. It's a problem description. It's a definition of the goal. It's the entire context people don't want to provide, because they assume the tool "will figure it out".
It won't figure it out.
And here is the first great friction of the 21st century: we have tools capable of cosmic things, and we try to use them like a TV remote - press a button and expect a miracle. Except the remote works because the TV has a manual. In AI, the manual is you. Your data, your intentions, your logic.
That's why I say it plainly: AI won't replace humans. AI will replace the human without a process.
Because a professional works by process. Even if they don't call it a process, it's in their blood: first reconnaissance, then a plan, then execution, then a test, then a fix. And a person without a process... well. They want it "right now", "pretty", "just make it work". And that's exactly how this mass of content, mass of code, mass of nonsense is born - things that sound professional but are hollow as a blown egg.
In books about thinking and decision-making (Kahneman, for one), a single thread keeps recurring: humans have a tendency toward shortcuts, guessing, filling in the blanks. AI has exactly the same tendency - just in turbo mode. When it doesn't know something, it sometimes won't say "I don't know" - it will produce the prettiest possible version of an answer so it sounds credible. And here comes the most important element, which few people talk about:
Verification. Responsibility. And no naivety.
I always repeat: AI is great at generating. The human is responsible for the truth.
You are the one signing off on the code, the offer, the strategy, on what you send to the client. Not ChatGPT, not the model, not the tool. You.
And now I watch people swallow AI like a pelican: "if it wrote it, it must be right". Well, no. That's not how it works. It's like taking the first answer from Google and turning it into a doctoral thesis. Or reading a description of a disease and going to operate on yourself in the bathroom. Sounds absurd? Then why do people do it in technology every day?
A professional always has three questions in mind:
- Is this logical?
- Is this consistent with the domain (i.e. the real world)?
- How will I test it?
And here we come back to the difference between the "AI programmer of the trade" and the "AI programmer from zero". The first one knows where AI most often goes wrong. Knows that when you get code, you have to run it, test it, review it. Knows that when you get a strategy, you have to confront it with reality. The second one takes the output as an oracle and keeps going.
And then we get an internet buried in "content" and "solutions" that are copy-paste, just in different packaging. In a sense, we live in an era where "professional-sounding nonsense" is the greatest plague.
Paid tools
And now another topic that hurts: paid tools.
I know everyone wants things "for free". Except in practice, free is the most expensive, because you pay with time, chaos, and rework. You pay with frustration. You pay with things not working, not being able to return to a thread, with limits, with slowness, with worse quality.
Time
As an aside - I once received very extensive test results. I described them in ChatGPT 5.2 PRO and got an analysis (I stress: an analysis, not a verdict). I took it to one of the better cardiologists in my area. I told him I had an opinion "from a doctor" - deliberately, because I wanted to test the reaction.
Even so, he confirmed and shared the artificial intelligence's view. Then I told him outright that it wasn't a doctor's opinion but an AI analysis. The cardiologist was momentarily stunned. Today he uses the chat and is my client. But - and this is the key - he doesn't outsource his work to it. He supports his work with it, and speeds it up at the same time.
I won't name names, but I hope this gives food for thought.
In the end, the same thing remains: everyone has access to AI, and everyone can do things limited only by imagination. But not everyone will do it like a professional. AI is not magic - it's a lever. And a lever only works when you have a fulcrum: knowledge, practice, and a standard of verification.
Only the graphic in this article is AI-made.
Until next time!