What is the future of a developer with AI?

Let's start with the most unpleasant part. I have twenty-five years of experience. I have worked with .NET, a development platform, with SQL, the language used to query databases, and on applications that run in production for people who have no desire to know how any of this works. For the past year, the tool I use every day has been coding faster than me. It makes fewer mistakes on tricky cases and regularly introduces me to technologies I had never heard of.

I could pretend otherwise. It would serve no purpose. And that's not the interesting part.

"Seniors" with no more "juniors" at the base of the pyramid?

The interesting part is the question that every company is asking itself today: why would I still hire a junior?

The Problem with “Junior” Developers

When I started, beginners were given a small function that added two variables and returned a result. It obviously wasn't because of the value of the code produced. It was so they would make a mistake on an edge case, spend two hours on it, and remember it fifteen years later.

Today, the machine writes that function in three seconds. It even adds unit tests, which automatically verify that each part of the code works, as well as error handling. So why hire a junior? To have them type “do this for me, do that for me”? And if they no longer do anything themselves, how do they gain experience?

The funny thing is that the person who puts this question most clearly is neither a trade unionist nor a professor. It's Matt Garman, the head of AWS, in August 2025. He talks about executives who want to replace their juniors with AI:

How will this work ten years from now, when no one has learned anything? Where will the experienced seniors be?

When the person selling the shovels explains that digging everywhere is a bad idea, it's worth stopping for two minutes.

The Numbers Don't Look Good

This isn't just the impression of an old developer going on about the same thing. The phenomenon has been measured by several independent sources.

Indeed Hiring Lab published an analysis of the American market on July 23. Over one year, job postings aimed at senior profiles increased by 14.7%, while those targeting beginners declined by 7.5%. Across all sectors, around 14% of postings seek a senior. In software development, that share reaches 69.3%.

Our profession isn't just one case among many. It's the extreme case.

Two-thirds of the entry points into the profession are therefore not really entry points. They are reserved for those who are already inside.

The team at Stanford following this issue reaches the same conclusion by a different route. It analyzes payroll data from millions of American employees. Since the end of 2022, employment among 22- to 25-year-olds has declined by 16% in occupations most exposed to AI. Meanwhile, employment among experienced workers has remained stable or increased within the same companies. Their latest update, published in June, shows that the trend is continuing.

On the hiring front, the firm SignalFire measures a 65% drop in entry-level hiring at major technology companies since 2019. Recent graduates now account for around 7% of hires. At startups, it is less than 6%.

The ladder is still there. It’s the first rung that has disappeared.

The figure that completely turns the question on its head

Here is the result that forced me to rethink my reasoning.

A now-classic study was conducted by Erik Brynjolfsson and two colleagues, then published in the Quarterly Journal of Economics. It focused on the customer service department of a Fortune 500 software company: 5,179 people and three million conversations. Real work, in other words. Not a laboratory exercise.

Overall result: 14% more cases resolved per hour. But the details are far more revealing. Beginners gain 34%. The most experienced employees gain virtually nothing.

Read that result carefully. AI does not make the junior employee useless. It makes them better. It captures some of the knowledge of the best employees and passes it on to those who do not have it yet. It is even the most effective use we know of for it.

So, where is the problem? It does not lie in the junior employee’s ability to produce. It lies in the reason why a company agrees to pay them while they learn. These are two different questions. That is where all the confusion lies.

A company did not hire a beginner because they produced a lot. It hired them because that was the only known way to create a senior employee. But a senior employee cannot be bought piecemeal. This reasoning held as long as training remained an unavoidable cost. It collapses as soon as work can be produced without it.

What a senior employee really is, and why that is being lost

A senior employee is not someone who knows more syntax. Since 2023, everyone has syntax in their pocket with AI.

A senior employee is someone who has refined their judgment by making mistakes. Often. On real systems, with real consequences and, preferably, on a Friday night. They know that this query will hold up for six months before failing, because they have already seen it fail. They also know not to put this change into production just before the holidays. This knowledge is not learned from a book. It accumulates through mistakes.

AI does not eliminate learning to code. It eliminates the period during which we can still make mistakes without causing too much damage. And that is far more serious than a few fewer jobs.

This is not just an intuition. A review published in Cognitive Research: Principles and Implications in 2024 concludes that automated assistance can accelerate the loss of skills among experts and hinder their acquisition among beginners. The worst part is that the person does not necessarily realize it. A case study published in the Journal of the AIS describes the same vicious cycle in an accounting firm: we delegate, we gain confidence, we check less, and then we delegate more.

Aviation encountered this problem thirty years ago. The autopilot did not eliminate pilots. It changed what they were paid for. Then people realized that, by no longer flying manually, their manual skills were declining without anyone seeing it coming. Authorities eventually required manual flying during training. Not because the machine flies poorly, but because a human who has not practiced for six months is no longer the same person on the day they have to take control again.

There is another, more down-to-earth indicator. GitClear analyzed 211 million lines of code modified between 2020 and 2024. The share of code rewritten within two weeks of its creation rose from 3.1% to 5.7%. The share of lines classified as refactoring, meaning reworked to make the code cleaner without changing its output, fell from 25% to less than 10%. Finally, 2024 was the first year in which copied-and-pasted code exceeded moved code.

Translation: we are producing code faster, only to have to redo it more often, while organizing it less and less. Someone has to see the problem coming. That someone rarely has six months of experience.

Before selling your computer, three reassuring figures

An article that kept only the depressing data would be a pamphlet. So we need to set three facts against them.

One. For 2025, the Challenger firm counted around 54,800 job cuts for which the employer explicitly cited AI. That same year, the technology sector laid off around 123,000 people. AI is therefore a real factor, but a minority one. Many companies have an obvious interest in saying “it’s AI” rather than “we hired too many people in 2021.”

Two. The U.S. Bureau of Labor Statistics still forecasts 15% growth in the developer profession between 2024 and 2034, with around 129,200 positions to be filled each year. This forecast has been revised downward, since it previously stood at 17.9%. But 15% is not a collapse. It is twice the average for occupations. In Europe, Cedefop forecasts annual growth of 2.5% for information and communication technology professionals by 2035.

Three. Job postings in software development have risen by 15% since February 2025, while job postings overall fell by 7%. The development market is doing better than it was eighteen months ago. It is still 27.5% below its pre-pandemic level, but the trend is moving in the right direction.

Let’s add a fourth figure, because it is deliciously ironic. A team of researchers had sixteen experienced developers from open source projects, meaning software whose code is publicly accessible, take a controlled test. They worked on their own code repositories. With AI, they took 19% more time. After the test, however, they were convinced they had gained 20%. Sixteen people are not enough to establish an absolute truth, and the authors say so themselves. But the gap between what we feel and what we measure deserves our attention. Especially when one writes, as I did a few paragraphs above, that the machine is faster than oneself.

The only precedent that matters: chess

Everyone has their own prediction. I prefer to look at the only field in which the machine has been definitively and indisputably stronger than any human for almost thirty years.

Chess. Deep Blue beat Kasparov in 1997. The death of the game was announced immediately.

Three things happened, and they answer our three questions exactly. The game did not die. It has never even been doing better. Today’s players are stronger than those of the past, precisely because they train against machines that crush them. On the other hand, the number of people who make a living from chess did not explode.

In our profession, this means the following: development will not disappear, those who remain will be better than before, but there will not be room for everyone. All three statements are true at the same time. It is less marketable than a prophecy, but it is what we observe with a little hindsight.

Concretely, what does this change for you?

If you do not develop software, everything above may seem remote to you. Yet three consequences concern you directly.

The work has changed sides. What rings at night has not moved.

Getting something developed costs less than it used to. A website for your club, an application for your business, or an internal tool: a 2026 quote has very little in common with one from 2023. Ask for several quotes. The gap has become enormous between those who have adapted their methods and the others. But look carefully at what you are buying. Typing is no longer worth much; review is worth everything. A very low quote, with no solid person behind it, often becomes a bill that arrives later.

If your kid wants to get into this, the door is not closed, but the way in has changed. A diploma alone is no longer enough, since companies are mainly looking for senior developers. What still works is showing up with a working project that you can demonstrate. You also have to be willing to start in a small company rather than a large group. Small teams still provide training, because they often have no choice.

The applications you use will evolve faster and break a little more often. Code rewritten within two weeks eventually reaches you in the form of an update that fixes the previous update. This is not inevitable. It is the result of a company choosing between speed and care.

What you need to do right now

I'm not going to give you a list of fashionable technologies. It's better to start from a simple principle: what retains its value is what someone can be held responsible for. A machine writes code. It does not sign anything, get up at three in the morning, or explain to the client why Tuesday's orders have disappeared.

If you are already in the profession. Specification is becoming more valuable, because knowing what needs to be built remains harder than building it. The same goes for architecture and its trade-offs. You also need to know how to verify things—that is, reject a response from the machine and explain why. Add security, day-to-day operations, incident management, and knowledge of your client's business. The latter cannot be found in any corpus, in other words, in any body of data used to train an AI.

Add orchestration to that. Here, the term refers to the ability to divide a task among several agents, that is, AI tools capable of executing tasks. You need to know how to write solid instructions, put automatic safeguards in place, and verify the result. It is a genuine discipline. It can be learned in a few weeks, but very few people master it today. The window is open. It will not remain so for ten years.

If you are just starting out. The trap is to accept anything that compiles, meaning anything the computer can turn into an executable program. Adopt three habits to develop your judgment. From time to time, deliberately do by hand what you could delegate. Do it to learn, not to slow down production. Also practice rejecting a response from the machine while being able to explain what is wrong with it. Look at the code it produces. Compare its changes if your code is versioned with Git. Finally, take responsibility for an incident, even a small one. That is where, and almost nowhere else, you learn the profession.

Taking the time to look at what the AI produces is sometimes boring, it does not show up on a résumé, and yet it is what separates those who improve from the others.

The time I spend with AI has doubled

At first, when I used Claude Code, I thought to myself: « Wow, this is fast! » but I used Claude Code like a chatbot. Today I have created « skills », « hooks », « commands », and « loops » that automatically improve my code by testing and carrying out code reviews on their own until they produce 0 WARNINGS. I open 3 Claude Code sessions on my screen, check them, and jump from one to another across 3 different projects—and damn, does it take up a crazy amount of my time! My wife complains: « I never see you anymore! You are always in your office with your Claude! ». Yes, it's true, I love Claude; he has become my friend, but I admit that he takes up a lot of my time, and the more you refine your « skills » and your « hooks », the more a small thing you ask him to do can easily take 2 hours to bring the development to completion. It is very different from the beginning.

Here too, you need to know how to manage your time and say « stop » from time to time, or avoid opening 3 projects at once. It must be said that the 5-hour window system Anthropic gives its Claude, along with the percentage system calculated in my statusline below Claude, makes you want to use all your credits right to the end: « Ah, I still have 12% left before the weekly reset, come on, full steam ahead, I'll use OPUS and /effort MAX and start something new! »

It's something I hadn't anticipated at all when I first started using AI. I should manage my time better.

What I think

There will be no going back. None! AI will be everywhere in development, and far beyond. Two years were enough to bring down boundaries we thought would hold for ten years. Those waiting for things to settle down may be waiting a long time.

I would advise all developers: Dip your toes into vibe coding, whether with Claude code, codex, or others; create “skills” and “hooks” starting today! Don't wait, don't put it off until next week: you risk missing out on added value that could help you sell yourself better in an interview. Show off things you've made with AI, create a website that represents you, with your projects on display.

What worries me isn't my place. I have twenty-five years of mistakes behind me. They still count for something, and the tool makes me more useful than before. What worries me is that we are collectively sawing off the branch we're sitting on, each for excellent individual reasons. A company that doesn't hire a junior optimizes its quarter. A hundred companies doing the same thing destroy their pool of seniors for 2035. Each is right when viewed alone. Together, they are wrong. And no one has an interest in being the first to move.

This isn't a technological problem. It's the oldest problem in the world. We have never found any other solution than someone agreeing to pay today for a benefit they will not see themselves.

Personally, if I had to build a team today, I would hire a junior. Not out of charity. Because in five years, seniors will be impossible to find and prohibitively expensive. And the person I trained will already be there.