You’ve seen the demos. Someone types a sentence, an AI spits out a working app, and the internet loses its mind. “AI can code now!” they shout. But they’re missing the point entirely.
Recently, someone asked an AI assistant named Astra to build playable 4D chess. Yes, four-dimensional chess. The impossible, mathematically brain-bending game. And it worked. It’s fun, it has an interactive tesseract, and it makes you feel like you’re playing with a higher dimension.
The real breakthrough wasn’t that the AI understood four-dimensional space. The breakthrough was that a human’s idea became a playable reality in minutes, without writing a single line of math.
But look closely at the comments on that 4D chess post. People aren’t just praising the machine. They’re praising the design. One user wrote, “I liked the explanations, graphics, interactive tesseract, all nice attentions to detail.” Then they asked the real question: “How much guidance did you give Astra? Did you give a single prompt, an example site to mimic, or was there more to the design and make process?”
That comment hits on the exact tension nobody wants to talk about. We want to believe in a fully autonomous AI that just does the work. But the reality is far more interesting. The AI didn’t just wake up and decide to build a hypercube chess board. A human had the vision. A human curated the outputs. A human decided that two surfaces were better than a filled hypercube for gameplay.
We are obsessed with whether AI can build things. We should be obsessed with who is steering it.
The AI was the builder; the human was the architect. The skill of “prompting” is really just product judgment in disguise. It’s the ability to specify, critique, and iteratively steer the output until it becomes something delightful.
When the cost of building software drops to zero, the only remaining scarcity is taste.
You’ve probably felt this shift happening. The barrier to entry for complex, interactive experiences isn’t coding anymore—it’s having a clear vision and the ability to refine what emerges. If you can’t tell the difference between a generic AI output and a genuinely delightful user experience, you’re going to be replaced by someone who can.
The 4D chess game is a marvel. It proves that abstract, complex ideas are now just a conversation away from becoming real. But don’t be fooled into thinking the machine did all the magic. The magic was in the human steering it.
The future belongs to the curators. The AI will write the code, but humans will still have to decide what is worth building.
FAQ
Q: Doesn't the AI do all the heavy lifting if it writes the code?
A: The AI executes the syntax, but it has no concept of user delight. If you don't actively steer it, critique its outputs, and define the user experience, you just get a generic, soulless app. The human provides the heavy lifting of product judgment.
Q: Do I still need to learn how to code?
A: To build things, no. To understand complex systems and debug edge cases, yes. But the competitive edge has shifted entirely from syntax memorization to product vision and iterative design.
Q: Is prompting actually a real skill or just a buzzword?
A: It's a buzzword if you think it's about magic phrases. It's a hard skill if you realize it's actually rapid prototyping, quality assurance, and user experience design rolled into one.