The variable naming problem never went away. We just moved it from "temp2" to "make this better."
Same skillset. Different enforcement. Compiler used to force clarity through syntax errors. AI forces it through three debugging cycles when you realize the output doesn't match what you thought you said.
Natural language was never designed for precision. We spent decades building tools that forced precision. Now we're back to ambiguity at scale, and the engineers who couldn't name variables are writing paragraph-long specifications.
So far, I have found that good AI-generated code comes from good developers. The prompt flow that seems to work best is to be able to break down a problem in logical steps, articulating specific requirements along the way. Sometimes it makes sense, and you develop intuition for, when to ask the AI to just stub out service to come back to later. In other words, good prompting for software developer appears to require some of the core problem-solving skills needed to be a good developer in the first place.
As an aside, I also sometimes ask AI agents to help me rename variables!
Now we rely on language models that don't throw errors when we're imprecise. Programming is becoming less about logic and more about articulation in natural language.
Errors are thrown for a reason. An LLM sycophantically ignoring them doesn’t solve the problem. Would you rather someone told you that you had a big hunk of spinach in your teeth or tell you you look great and let you walk around like that all day?
I agree with you. An LLM sycophantically ignoring ambiguity IS the problem that requires us to become now much better at communication with natural language. Did you misunderstand the point of the article OR was the article/summary poorly written? :)
Same skillset. Different enforcement. Compiler used to force clarity through syntax errors. AI forces it through three debugging cycles when you realize the output doesn't match what you thought you said.
Natural language was never designed for precision. We spent decades building tools that forced precision. Now we're back to ambiguity at scale, and the engineers who couldn't name variables are writing paragraph-long specifications.
Logic is abundant. Clarity is the new bottleneck.