This is quite interesting.

Today, I came across an article on InfoQ titled The Limits of Vibe Coding: 30 Million Developers Can’t Deliver Inspiration for 8 Billion People. Honestly, it stung a bit to read, but I have to admit—it hits on a question we’ve all been avoiding: After all the hype around AI programming tools, how many more people can actually write code now?

The answer might be harsh—it’s still the same tech crowd.

The article points out that while AI has lowered the barrier to programming—auto-completing code, explaining errors, even writing entire functions—it’s still fundamentally serving those 30 million developers worldwide. As for everyone else? They still don’t understand what an IDE is, what dependency management means, let alone debugging. No matter how smart the tools get, the cognitive gap remains.

This reminds me of a scenario. A while back, a friend excitedly told me, “AI can write code now—does that mean I can build an app?” They opened GitHub and gave up within three minutes—just setting up the environment was enough to scare them off. See? No matter how user-friendly the tool, the complexity of the tech ecosystem is still a wall.

It’s a paradox. On one hand, we talk about AI being “inclusive,” about empowering all 8 billion people to express their creativity. On the other hand, the tech world is mostly just patting itself on the back. “Vibe Coding” sounds cool, but its “vibe” is still a techie vibe. When outsiders step in, they’re met with jargon and quickly leave.

One point from the article really resonated with me: AI reduces operational costs, not understanding costs. You can use ChatGPT to generate code, but that doesn’t mean you know why it works—let alone how to fix bugs. Programming is, at its core, about logical expression, and tools alone can’t teach you how to build that logic.

Of course, some might say, “Give it time—not everyone knew how to use computers back in the day.” But the problem is, the speed of technological advancement and the difficulty of adoption are widening the gap. The more powerful AI tools become, the more complex the tech stack grows, and the more newcomers have to catch up.

So, what does “programming for all” really mean? Maybe it’s not about everyone writing Python, but about turning ideas into digital products in a more natural way—like describing what you want in plain language and having AI generate a functional app, skipping code, configuration, even the concept of “programming” itself.

But we’re far from that. Right now, AI programming tools feel more like giving techies a sharper knife, not building a bridge for everyone else.

(Original article link: https://www.infoq.cn/article/xyz123 — worth a read, the arguments are sharp.)

One last gripe: The tech world sometimes feels like a xianxia novel—the higher your “cultivation level,” the easier you think the basics are, forgetting that most people haven’t even tested their “spiritual roots.”