Ask HN: What is no longer worth learning in programming due to AI?
I'll start:
1. You don't need to learn frameworks anymore. New languages other than the ones you're already comfortable with. You no longer need to feel guilty for not knowing Rust or Go if you're a C++ or Python guru. No need to nag yourself about learning VHDL if you're a Verilog dev. No need to worry about C# if you're a Java person. No need to sweat makefiles if you're a CMake fan. We will see continuing progress in programming languages, certainly, but that progress will be aimed at making them more accessible to AI, not to humans. - You don’t need to learn debugging, just regenerate until it works. - You don’t need to learn types, the model knows what you meant. - You don’t need to learn security, just ask the LLM to make it secure. - You don’t need to learn performance, you can always scale later. You don’t need to learn performance, you can always scale later. My (half-joking) attitude was always, "Performance? That sounds like Intel's job, not mine. My code magically gets faster if I do nothing at all, so I'll focus on new stuff instead." That didn't scale very well at all, as it turned out.