Does it make sense for AI to KEEP programming languages?

▶

I was recently left a very interesting comment:

…In my humble opinion, the "language" issue will increasingly fade from the scene as we see how we ask AI to model the projects we have in mind...and it's achieving this in increasingly surprising ways in whatever language it comes up with...

Mind you, this isn't a criticism or anything like that. I just liked the idea and wanted, so to speak, to give my opinion. And more than mine, I'd like to read yours in the comments and discuss them there.

The comment I'm interested in is this first part, in which the person comments that, in their opinion, the language issue will gradually take a backseat, allowing the AI to select the one it deems best.

There are several things here. This is already happening with some AIs—I don't want to name names because they might correct it or another AI might do it better later—but currently, many times, when you create an application from scratch, it always suggests the same technologies.

This is a bias. In other words, the AI evaluates what it's comfortable using more than the actual features of the project. Even if it allows you to select other options, it always tries to lean toward a particular technology. For example, it uses Tailwind instead of something else, simply because it's the one it already "knows."

The problem of total control

This bias is the first sign of a bigger problem: that AI will end up selecting the programming language for us, leaving us as mere spectators. If we give it all that power, it would no longer make sense for a human team to maintain that language. Because if AI is the one creating the projects, why would we be maintaining that technology?

This is where I see the danger: that AI has complete control. And if that's the case, then it would make sense for AI to maintain the language or framework, because it would be its tool, not ours.

Does it make sense for AI to maintain languages?

This is the point I want to raise: Does it make sense for an AI to maintain a language, framework, or technology? For me, it would only make sense if it uses it independently to develop projects without human intervention.

It's a very interesting topic, so I wanted to ask the AI what it thinks. I refreshed the page (I didn't edit anything) and asked the following question:

Does it make sense for an AI to maintain programming languages or frameworks independently of human developers?

The answer was very comprehensive. I'm not going to read it all because it would take forever, but you can pause and look at the titles.

Summary of Gemini's response

• More efficient error detection and correction
• I completely agree. Manual testing is horrible. If we already have the design and code, AI can analyze what we did, suggest improvements, perform tests... and save us a lot of time.
• Evolution and continuous improvement
  Languages are constantly evolving, but at a pace limited by human resources. Here, AI has an advantage: it can do in seconds what would take a human hours or days. It can also analyze new trends and architectures, although this means we're giving it ALL the control.
  And that's where the danger lies, because technologies currently evolve thanks to human experience. For example, at Laravel, they invented tools like wire:click, which seem simple but are the fruit of human creativity. I don't know if an AI would come up with that on its own without human context.
• Personalization and adaptability
  An AI could adapt a language for specific tasks. It sounds good, although some of the things you mention are too specific or difficult to evaluate without real context.
• Maintenance Automation
  Here we are again at the same point: if it starts with human-made development and improves it, great. But if AI does everything from scratch and also maintains it… it no longer makes sense for humans to continue in that loop.

Gemini closed with a worrying phrase (albeit with some humor): “He already kicked me out, he retired me and he sent me to hell.”

But what was interesting was their conclusion: an AI can't understand the human context. That's the key.

Summary of ChatGPT's response

I then asked ChatGPT the same question, and as always, their response seemed more humane. Gemini tends to be more robotic and rambles on. ChatGPT was more direct and clear.

I'll highlight a few points:

• It can help a lot with testing, security, documentation, refactoring, etc.
• It can maintain tools that are no longer useful for humans. I found this very interesting.
• It can detect common patterns and suggest improvements… but it shouldn't make decisions without us.
• It recognizes that frameworks and languages are not just technology, but also opinion, vision, and community. And that, today, is human.

Conclusion

So, does it make sense for an AI to maintain a language or technology?

My answer: not at all. It makes sense if the AI has full control over development, but as long as there's still a human team with vision, experience, and creativity, we can't completely let it take control.

For example, I'm developing a course on how to create an online store with Django 5. I used Django Admin because it's an excellent tool for quick CRUDs. The AI, if it understands the context, might suggest Django + Django Admin as the ideal solution. But the final decision remains mine.