Peter Müller, professor at ETH Zurich, argues that studying computer science retains its value despite advances in generative AI for code. His central thesis is that the discipline goes beyond writing lines of code: it encompasses system design, security, and the very creation of AI. The figure of the professional with analytical thinking remains indispensable.
The Technical Role of the Computer Scientist in the AI Era ?Ӥ
Current AI performs well in bounded contexts, but cannot conceive, decompose, and guarantee complex systems on its own. The computer scientist is key to defining precise requirements, breaking down problems, and verifying the correctness, security, and efficiency of the generated code. Their work shifts to a higher level of abstraction, supervising and directing the capabilities of AI tools.
So AI will program itself? Sure, and my router makes coffee ??/h3>
It's understandable to be enthusiastic, but thinking that AI will relieve us is like expecting an electric screwdriver to build a house by itself. It can tighten screws quickly, but someone must read the blueprints, choose the materials, and ensure the walls don't fall down. AI is that fast tool that, without supervision, might decide the door looks good on the ceiling.