I'm a professional developer, and AI has significantly increased my output—I'd say by maybe 30 or 40 percent. GitHub Copilot has significantly changed the way I work with code.
However, I take pride in producing high-quality code quickly, which is why my rates are high. Using AI helps me increase my output while maintaining that level of quality.
My take on AI is that it is not going to replace humans anytime soon.
It is, however, putting significant pressure on the economy. Previously, setting up a functional, decent-quality project without much complexity took time—at least weeks. Now, such tasks are incredibly fast and easy; anyone can set them up in a few minutes using AI, even without any coding knowledge.
Success in most fields, however, is not just a measure of how fast you can build; it's also about how well you can execute. Current AI can offer advice, but it still cannot execute for you. Market success requires sensitivity, context, and adaptability. AI can help significantly if you know how to ask the right questions. But the economy is made of people, not AI (yet). To earn money, someone must give you money because they value what you offer. The arrival of LLMs hasn't changed this.
I feel the pressure. The corporation I work for is pushing for AI adoption, and the initial drawbacks and realizations are already becoming apparent.
First point: Customers, at best, don't care about your AI. They don't want it.
Second point: AI succeeds at making developers more productive but fails with higher complexity—though not for the reason people usually think. With the right prompt, GPT-5.4 can create fairly complex solutions, even more complex than many corporate business processes.
The real reason is that, at a certain level, complexity lies not in the total amount of information in the system, but in how the human aspect of the business translates when you try to formalize higher-level context. This is something most developers don't see (or care about). For example, guiding an AI to implement a feature in a repository is easy. However, asking an AI how 50 apps in your company should better communicate is a question it cannot answer well, simply because this specific question may have hidden dependencies linked to why your products are successful in the first place. Do you see what I mean? At a certain level, technological questions can no longer be purely technological; they become (and must be) political and business-oriented.
Every developer dreams of having a perfect description of what needs to be built, complete with precise analysis and exact information. But you never get that in a higher-level context, because business success isn't modeled by perfect ideas alone—it relies heavily on execution, which often involves acting on incomplete information.
This is why AI is not going to replace developers anytime soon.
Top comments (7)
the framing might be wrong - not whether AI replaces devs but whether it changes how many devs projects need.
I respectfully disagree with both points.
First, whether customers care about AI depends entirely on the customer. Many don't care how a product is built, only that it solves their problem. But there are also organizations that explicitly care about AI because they expect it to reduce costs or increase delivery speed. We're already seeing companies reduce budgets on contractors under the assumption that a team half the previous size can deliver similar output with AI assistance.
Second, I don't think the conclusion follows from the premise. AI certainly struggles with overall design, architecture, and business context. But that doesn't imply developers are protected from disruption.
What's becoming increasingly clear isn't that senior engineers will suddenly become 10× more valuable. It's that the market itself is changing. Higher hiring bars, fewer junior opportunities, more selective recruiting, and widespread layoffs are no longer isolated events - they're visible across much of the industry, globally.
The engineers who remain will spend more time supervising AI, reviewing architecture, validating correctness, integrating systems, and translating business intent into technical decisions. Those are valuable skills today. But I wouldn't assume they'll remain uniquely human forever. AI capabilities continue to improve, and many of today's "AI oversight" tasks are likely to become increasingly automated.
To me, the important question isn't whether AI will replace developers overnight - it won't. The more useful question is: where will you still create unique value five years from now? Finding that answer early may matter more than debating whether AI can or can't replace today's software engineer.
I'm not sure I follow. The first point was more about end customers not caring about chat bots and similar. But sure, that's subjective, it's from my POV.
And second point:
If you only employ engineers who work only with AI only then in time no one understands what the AI is even doing. You need juniors and junior positions. You need to allow people to become experienced with code. Otherwise day will come and no one will be able to untangle the mess your engineers created with AI to a point where using AI no longer effectively works at sustainable cost.
You've forgotten to mention the backslash that is already happening with too much reliance on AI. Also happening globally. Always fun to use the world "globally" with strangers on the internet.
So no, the world where engineers only steer the AI according to you is not happening today and not in five years. But what do I know right? :)
I would note one point: the industry has practically starved itself of junior positions, which is a serious long-term problem. In a sense, it's optimizing away its own future - or perhaps it simply doesn't believe that future will require as many software engineers.
But that doesn't contradict my argument; it reinforces it. Companies are changing hiring patterns because AI is changing the economics of software development. Whether that's the right long-term decision is a separate question, we are talking of facts.
As for "globally", I use that word deliberately, because it precisely describes the situation. Large-scale layoffs, decline in junior openings, and significantly higher hiring bars absolutely happen in the marketd of North America, Europe, and Asia. The exact numbers differ, but the overall direction is consistent.
Regarding the AI backlash, we've seen waves of skepticism since ChatGPT was released. Yet AI capabilities and adoption have continued to advance despite them. Short-term corrections don't necessarily invalidate the long-term trend. There's also notable survivorship bias in how these stories are reported. A company announcing a mass rehire becomes news because it's unusual. Continuing layoffs rarely make headlines anymore because they've become routine.
And that's really my point: none of us can say with confidence whether the market will reach equilibrium in another five years. But if you asked most engineers five years ago whether AI would materially shrink the software engineering job market by today, many would have dismissed the idea as alarmist. The pace of change has already exceeded what most people ever expected.
Agree, and the reason is more specific than 'AI isn't good enough yet.' AI replaced the part of the job that was always the cheapest, typing. What it didn't touch is the expensive part: deciding what to build, holding the system in your head, and knowing when the plausible output is wrong. Those are the parts that make a developer, and they get more valuable as generation gets cheaper, not less. The role changes; it doesn't disappear.
The point about complexity living in the human/political layer is the one most "AI replaces devs" takes miss completely.
You can prompt your way to a working feature. You can't prompt your way to knowing why two teams haven't talked to each other in 6 months and what that means for the architecture.
I like this post where you highlight the positive of AI, I am working in IT industry more than 20 years, definitely using dev assist tools is is complementing developers. Developers still have to learn how to use these tools to show their engineering skill, not just coding skill.