Developers won't be out of work because of AI, but juniors...

If you're a software developer, I don't think AI is going to put you out of work, at least in the short term. In the long term, well... we'll see what happens. Except if you're a junior, I think things change a bit there.

Remember: this is purely my opinion. You can agree, disagree, partially agree, etc. But in this case, before telling you why, I want to show you a question I asked ChatGPT. I'll show it to you, and we'll analyze it in a few moments, to the point I want to get to.

What does a junior developer do?

Before that, I want to talk a little about what a junior is, or what the role of a junior is in a company, which is what I'm trying to get to. And, before that... the blessed blog! For God's sake:

desarrollolibre.net/blog

For me, a junior can fill several roles, but I think the main one is making small developments. Understandably, precisely because of the junior's lack of experience, the person in charge will tell the junior:

“Look, we need to create a modal that, when clicked, pops up, does some stupid animation, shows you some information, and, well, does that.”

Ah, perfect, we already have the modal.

“Then create a form out there using such and such a thing, or figure out how to use it or create it in such and such a technology, and finish the development.”

“Ah, perfect. Create me a paginated list of certain publications. For this, remember, we have such a thing...”

That is, it's literally a person—and I think you can already see the references—to whom you tell a development you want them to do, something you can control. Something that you see as not going to be very creative. Understandably, you're not going to tell the junior, "Make me the complete application the client requested, which has 20 different modules," because that's going to be a four-headed snake.

You try to guide them a bit, guiding them on how to do their developments, and obviously answering their questions and so on.

The key point: limited tasks

What I'm getting at, and the most important point: the junior is making small developments. In other words, you're never going to give them free rein—as you might with a semi-senior or senior—to do whatever they see fit. Obviously, they're a person with little experience. And that's the point.

So, what's my point with this? The key factor is this: a junior is someone you're going to ask for very limited developments, and that's practically it.

ChatGPT/AI enters the scene

And this is where ChatGPT or some other AI comes in. For me, that's precisely what this type of AI allows for. I say ChatGPT because it's the one I use and I think it's the most general, but there are probably other, more specific ones that I haven't tried yet, since this one is enough for me.

My experience with AI has mostly been with ChatGPT, because it's served me well. Literally, it serves me well for what I need it for. I haven't had the need to try anything else.

What ChatGPT can do (and what it can't)

For example, if I need a small development project:

“I need to make a modal. I need to make pagination. I need to make a graph. I need to make a filtered list. I need to make a form with validations. I need to convert HTML to PDF. I need to make a Vue component that does this…”

I ask ChatGPT, and in a matter of seconds or minutes, I have a working base. Sometimes they even give me the complete code, explain how to use it, walk me through each part of the code, and if I have questions, I ask them and they answer. It's literally like having a junior developer working with me, but one who doesn't sleep, doesn't get paid, doesn't complain, and is always available.

And what about the juniors?

Now, what's the problem? If I had to hire someone, and I see that an AI like ChatGPT gives me a working solution in seconds—be careful, sometimes I have to adapt it a bit, or it's not perfect, but 80% of the work is done—why would I hire a junior?

And this is where many might say:

“Well, but ChatGPT doesn’t understand the context of an entire application.”

It's true, it doesn't quite do that. But here's the important point: if you're leading the project, if you're the senior or the architect, then you already know where the application is going. And if you have the time to coordinate and ask the AI ​​for small parts, you then assemble those parts yourself, as if you were receiving the work from a junior.

And you do it faster, cheaper, without depending on anyone.

Replacing junior employees with AI is "one of the dumbest things I've ever heard"

Amazon's cloud chief says replacing junior employees with AI is one of the dumbest things he's ever heard. And honestly, I agree. Let's slow down a bit and review some other quotes I found very interesting.

The most valuable skills don't depend on a degree

He also pointed out that the most valuable skills in the AI-driven economy aren't tied to a college degree.
So, if you're watching me and want to learn a new profession, you have no excuse: a degree is no longer a limitation to practicing professionally.

The future is constantly changing

Another quote I found famous was this: if you spend all your time learning something specific and think you'll still be an expert in it 30 years from now, I assure you it won't do you any good.
The reason is clear: 30 years is too long, everything evolves, and it probably won't be the same in the future.

What you should learn

Instead, he said students should focus on developing critical thinking, creativity, and the ability to adapt as technology evolves.
These are skills that can't be taken away from us, at least not in the short term. These are precisely what characterizes humans: critical thinking, creativity, and adaptability.

There should always be juniors

My favorite quote is the one that gives the title to all of this: replacing juniors with AI is the dumbest thing I've ever heard.
And following that logic, he asks: how will this work when in 10 years there's no one who's built or learned anything?
The truth is, we all start as juniors. If you eliminate those positions, you cut off the very foundation of learning and professional growth.

I graduated in computer science and the only company that called me was a fast food chain.

"I graduated in computer science, and the only company that called me for an interview was a fast-food chain," a McDonald's, so to speak. This connects with an opinion I shared a few weeks ago, in an article that said developers won't lose their jobs... but juniors will.

The problem of being junior, entry-level positions

By "juniors," I mean those with little work experience. This includes recent graduates—whether from a technical degree, a bachelor's degree, or an engineering degree—and also those learning programming at bootcamps. There used to be a popular saying that after taking a basic programming course, you automatically got a job.
To a certain extent, this was true, because software development always required people with different levels of experience. However, the situation has changed, and I've experienced it firsthand.

How AI Changed the Way I Work

I work alone, and before, I would have appreciated someone to help me with coding or writing articles. But now, AI fills that need in seconds and often with better results. The biggest impact, then, falls on entry-level positions: jobs that were previously reserved for juniors.

This doesn't just happen in IT; any digital profession with simple, repetitive tasks is vulnerable.

The reality of the job market

Another article mentioned that many graduates send thousands of applications without success. For example: someone sent 6,000 applications, got only 13 interviews, and one was at McDonald's… where they were rejected for lack of experience. It's a vicious cycle: to be a senior, you must first be a junior, but the opportunities to start are increasingly fewer.

Optimization and reduction of staff

AI doesn't just "replace" tasks, it optimizes processes. If a company with 20 developers can do the same work in less time thanks to AI, it can lay off some staff. Most will opt for this convenient solution rather than looking for more projects.
The "uncomfortable" alternative would be to expand and leverage that efficiency to grow, but few do so.

Conclusion and personal reflection

I think the problem isn't that "someone who uses AI better than you is taking your job," but rather that optimization itself leaves people out, especially juniors.
That's why I recommend taking advantage of these tools to create your own solutions, continue learning, and increase your opportunities.

The "10x Programmer" Trend and the Impact of Artificial Intelligence on Employment

In the technology sector, a phenomenon conceptualized as the AI-driven "10x Programmer" has emerged. This trend poses a complex scenario: the possibility of optimizing development workflows to the point of delegating the workload of multiple professionals to a single developer highly trained in the use of generative tools.

When a developer properly integrates AI into their workflow, they experience a drastic increase in productivity, managing to solve complex tasks in a fraction of the time traditionally required.

1. The Operational Dilemma of Companies Facing High Efficiency

Faced with this increased speed in software delivery, development organizations find themselves at three strategic crossroads:

• Reduction of Downtime: Allowing the professional to finish their assignments early and remain idle for the rest of the workday. This scenario is unfeasible in practice, given that companies seek maximum operational efficiency and demand compliance with the contracted working hours.
• Scaling Work Volume: Increasing the flow of projects and client acquisition.
• Staff Reduction (Technological Layoffs): Replacing part of the team with profiles that master AI tools with greater skill, thereby reducing operating costs.

The second path —increasing the workflow— presents serious real-world limitations. A company's capacity for expansion is neither linear nor infinite, and it faces complex market barriers:

+-----------------------------------------------------------------+
|                       SCALABILITY LIMITS                        |
+-----------------------------------------------------------------+
| • Saturated Local Market: If a regional agency absorbs the     |
|   traffic in its area, the number of new clients is capped.     |
| • Fixed Licensing Models: In embedded software or specific      |
|   industries, optimizing code does not double the sale of       |
|   existing licenses.                                            |
| • Disconnect with the Real Environment: The speed of technical |
|   development usually outpaces the speed of commercial          |
|   absorption.                                                   |
+-----------------------------------------------------------------+

Due to these market nuances, the most direct and simple solution for many organizations is to maintain traditional growth flows and adjust their staff sizes through layoffs, delegating the burden to professionals with AI-optimized skills.

2. Real Productivity in Software Development: Practical Experience

The adoption of Artificial Intelligence in large-scale projects demonstrates its value as an accelerator. In architectural refactoring and the development of new functionalities for complex platforms (such as authentication systems, shopping carts, or payment gateways), AI can solve between 80% and 90% of the initial mechanical work.

Even in areas where the developer does not possess primary specialization, such as graphic interface design or advanced web styling, these tools make it possible to enhance existing technical knowledge to build more pleasant and functional interfaces.

However, industry reports (such as analyses from specialized portals like Xataka) and software engineers agree on a critical factor: although tasks that used to take days are now structured in minutes, the developer must invest a significant percentage of time in analyzing, interpreting, and adapting the code generated by the AI.

3. The Technical Filter: Why the Human Factor Remains Essential

Artificial Intelligence presents severe limitations when faced with custom business logic or complex software architectures. A representative case study occurs when implementing local content pre-rendering systems for the generation of dynamic documents (such as technical books or Curriculum Vitae modules).

         CORRECT FLOW FOR CONTENT PRE-RENDERING
      
+-------------------+      +-------------------+      +-------------------+
|   CONTENT Field   | ---> |   PREVIEW View    | ---> | CONTENT_REND Field|
| (Source Code or   |      | (Rendered with    |      | (Processed HTML   |
|   Plain Text)     |      |    JavaScript)    |      |   ready for PDF)  |
+-------------------+      +-------------------+      +-------------------+
                                                              |
                                                              v
                                                       [ PDF Generation ]

When attempting to program this flow, language models often confuse the roles of the content fields (the data source) and content_render (the final HTML processed in the preview view). AI tends to introduce erroneous conditionals or execute duplicate data persistences during the save method, mixing up the scopes of both fields.

Conclusion

Unless dealing with the writing of generic, monotonous, and repetitive code, Artificial Intelligence tends to lose the context of complex systems. If the developer does not possess solid programming foundations to interpret, debug, and correct the inconsistencies generated by the model, the proposed solution becomes useless.

Therefore, the most efficient development methodology today does not consist of delegating total autonomy to an autonomous agent, but rather applying an incremental development approach: dividing large modules into microtasks and actively guiding the AI in executing controlled changes under constant human supervision.