My Code Works. An AI Wrote It. Neither of Us Knows Why.
here's a meme we've all posted some version of. "My code works and I have no idea why." Classic. Universal. Every developer since 1970 has felt it. But it changed this year, and nobody really announced it. Now it goes: "The AI wrote code that works. Neither of us knows why." Same joke. Completely different problem. The old version was funny because it was temporary Back then, "I don't know why it…
My code works. An AI wrote it. Neither of us understands why. The joke has shifted from developers struggling to figure out why their code works to developers deploying AI-generated code they don't fully comprehend.
By 2026, 92% of US-based developers are expected to rely on AI coding tools daily, and 41% of all code will be AI-generated globally. Surprisingly, 40% of junior developers deploy code without fully grasping it. This raises questions about whether developers are inadvertently shipping code they don't truly understand, and if AI is truly making them faster or if they're just under the impression it is.
Andrej Karpathy, who coined the term "vibe coding," now considers it outdated, as AI agents are increasingly capable of planning, testing, and correcting code across entire codebases. The entry-level software developer job is disappearing for 33 consecutive months, just as the primary skill demanded - judgement - is becoming increasingly important.
The issue isn't that AI generates bad code, but that working code can sometimes feel finished, when it rarely is. The rule now is simple: if you can't explain why the code works, it's not finished. This applies regardless of how well the tests perform. The real challenge is not the AI-generated code, but rather trusting it when we don't fully understand how it works.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
