Should You Still Learn to Code? htmx Author Says Yes
Original: Yes, and
Why This Matters
Establishes a practical framework for how CS education should adapt to AI tooling
Carson Gross, htmx creator and Montana State CS professor, argues programming remains a strong career path in the AI era—but warns juniors against letting AI write their code, as doing so stunts development of essential code-reading skills.
Carson Gross, creator of htmx and a computer science professor at Montana State University, addresses a question he hears constantly from students and family: is programming still worth learning given AI? His answer is 'yes, and'—with significant caveats.
Gross argues programming is fundamentally about two things: problem-solving with computers, and managing complexity. Neither skill becomes less valuable with AI; if anything, both matter more. His sharpest warning targets junior developers who outsource code generation to AI tools before they understand what the code is doing. 'You have to write the code,' he tells his students—because writing it is what builds the ability to read it, and readable code is the skill AI-era developers will need most.
He also pushes back on a popular analogy: that AI-assisted coding is like the jump from assembly to high-level languages. Compilers are deterministic; LLMs are not. High-level languages stripped out accidental complexity. LLM-generated code often adds it—through poor design choices, shortcuts, or mismatched approaches. Without the ability to read and audit that output, developers fall into what he calls 'The Sorcerer's Apprentice Trap': building systems they can't understand or control.
His positive case for AI is as a tutoring partner, not a code generator. Used that way, it helps learners get unstuck faster—reducing the environment friction that drives many students out of CS entirely.