Carson Gross, a computer science professor at Montana State University, is often asked whether students should still pursue computer programming now that AI can generate code. His answer is "Yes, and..." Programming is fundamentally about solving problems with computers and controlling the complexity of those solutions, skills he expects to remain valuable even as AI tools spread.
Gross warns that AI is especially dangerous for junior programmers. If students simply generate code instead of writing it, they miss the hands-on understanding that comes from working in the trenches. He tells his students, "Yes, AI can generate the code for this assignment. Don't let it. You have to write the code." Without that experience, they cannot reliably read code, and the ability to read code may become even more valuable in an AI-based future. Otherwise, they risk falling into the Sorcerer's Apprentice Trap, building systems they neither understand nor control.
Gross also disagrees with comparing the shift to AI-generated code with the move from assembly to high-level languages. Compilers are largely deterministic, while LLM output is not. High-level languages removed much accidental complexity, but LLM-generated code often adds accidental complexity through poor approaches and shortcuts. Reading code critically requires having written it.
Gross encourages students to use AI as a teaching assistant rather than a code generator. Getting stuck on accidental complexity, such as unfamiliar tool chains, can waste time and drive people out of computer science. Used well, AI can help students get past these roadblocks and deepen their understanding.
Source: Hacker News · Summarized by HeadlinesBriefing