I keep seeing developers say they barely write code anymore because AI handles much of it. That made me wonder: are there employers who would hire someone specifically to write prompts, or is programming knowledge still necessary to get and keep that kind of job?
4 Answers
Calling AI-assisted programming “just prompting” is a bit like calling traditional programming “just typing.” The important part is knowing what to build, recognizing bad output, and making sure the finished software actually works.
A senior developer will usually get much better results from AI than a beginner because they understand architecture, data flow, security, edge cases, and failure modes. In fact, those skills become even more important when AI writes most of the code. Entry-level roles may become more competitive, not disappear into prompt-only positions.
Some companies may hire people who mostly use AI tools, but you’ll still need enough technical knowledge to verify the answers and fix what the model gets wrong. Otherwise, you’re a liability rather than an engineer.
That’s a good path, but hiring processes may still test your knowledge before you’re allowed to rely on AI. Once you’re hired, the tool matters less than delivering correct work on time.
Probably not for prompting alone. Employers generally want someone who understands the task, can evaluate the AI’s output, debug problems, and take responsibility for the result. Prompting is becoming part of many jobs, but it usually isn’t the entire job.
That makes sense. I was also wondering whether someone could use AI as a teacher, learn the fundamentals quickly, and then become useful that way.