Yes, And

TL;DR

Before you orderOffer from Amazon

Get audio and creator gear delivered free with Prime

• Fast, free delivery on millions of items
• Prime Video, Amazon Music and more included
• Member-only deals all year

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

In a February 27, 2026 essay on htmx.org, Montana State University computer science instructor Carson Gross advises people interested in programming to pursue it despite AI, while learning to write and read code themselves. He says AI can help explain concepts and clear learning roadblocks, but warns that relying on generated code may leave junior programmers unable to understand or control the systems they build.

Montana State University computer science instructor Carson Gross says people should still consider becoming programmers as AI coding tools spread, but argues they need to learn to write and read code themselves. In an essay published on htmx.org on February 27, 2026, Gross recommends using AI to explain concepts and help students get unstuck, not simply to generate their assignments.

Gross frames his answer to the question of whether someone should pursue programming as “Yes, and…” He argues that programming involves solving problems with computers and managing the complexity of solutions, skills he expects to remain useful. The essay presents that as his view, not as a forecast backed by employment data or a study of AI’s effects on hiring.

His main warning concerns junior programmers who rely on generated code before developing practical coding experience. Gross says writing code helps learners build an intuitive understanding that they need to read and evaluate it. Without that ability, he warns, a programmer may create systems they do not understand or control, a risk he calls the “Sorcerer’s Apprentice Trap.”

Gross also disputes the comparison between AI-generated code and the shift from assembly language to higher-level programming languages. He says compilers generally produce more predictable results from specified code constructs, while the code produced by a language model in response to a prompt is less predictable and may add unnecessary complexity. He argues that users need to inspect generated code to judge whether it is appropriate.

At the same time, Gross recommends AI as a teaching assistant. He says students can ask it to explain programming concepts or help them work through problems with tools and development environments. He has shared an AGENTS.md file intended to configure coding agents to act more like tutors than code generators. The essay does not report testing results for that file or compare learning outcomes among students who use it.

Learning to Program With AI

The essay addresses a practical question for students and people considering a technical career: what should they learn if AI can produce code? Gross’s answer is not to reject the tools, but to use them in ways that preserve opportunities to practise programming and understand the output. That distinction matters because generated code can be useful only if someone can assess whether it fits the task and behaves as intended.

Gross also suggests that the mix of skills valued in programming may change. He expects clear communication and knowledge of the organization or business where software is used to become more important if writing code by hand takes up a smaller share of the work. Those are his expectations; the essay does not establish how quickly such changes will happen or how they will affect particular jobs.

Gross’s Advice to Students

Gross writes that he teaches computer science at Montana State University and has been asked by relatives, friends and students whether programming remains a sensible career choice given AI. His essay is a personal account of how he answers that question and what guidance he gives students. It is not a university policy statement or a report on the labor market.

He describes getting stuck as a common obstacle in learning programming. Sometimes, he says, the barrier is not a lack of ability but confusion about a tool or development environment. Gross recounts that difficulty learning Unix while studying at Berkeley contributed to his leaving that university’s computer science program. He says AI can help learners get past such obstacles and return to the work of understanding programming.

For future programmers, Gross identifies writing, thinking and communicating clearly as skills worth developing, alongside an understanding of the organizations whose problems software is meant to solve. He argues that programming is ultimately about applying computers to problems, whether those problems arise in a business, government or another setting.

“Yes, AI can generate the code for this assignment. Don’t let it. You have to write the code.”

— Carson Gross, addressing his students in the essay

What the Essay Does Not Establish

Gross’s article is an instructor’s argument and advice, not evidence that programming careers will remain unchanged or that a particular way of using AI improves learning. It provides no employment projections, student performance data or assessment of how widely AI-generated code is used in professional settings.

The essay also does not say how quickly the relative value of writing code will change, which programming roles might be most affected, or whether communication and business knowledge will gain value across the field. Gross presents these points as possibilities and expectations, and the effects may differ by workplace and type of software. The material provided does not include an independent response from other educators, employers or researchers.

How Students Can Apply the Advice

Gross encourages students to practise writing code while using AI for explanations and help with obstacles. His shared AGENTS.md configuration is one practical example of how he wants coding agents to support learning rather than complete assignments. The essay does not set out a formal course change, a rollout schedule or a later evaluation of the file.

Readers looking for evidence about AI’s effect on programming careers will need additional information, such as hiring trends, workplace studies and research comparing learning methods. Gross’s immediate recommendation is narrower: learn to write code, learn to read it, and use AI to support understanding rather than treating generated output as a substitute for that understanding.

Key Questions

What is Carson Gross’s advice to people considering programming?

Gross says yes, they should consider it, while also learning how to work with AI. He argues that problem-solving and managing software complexity remain useful, and that students should build the ability to write and read code.

Does Gross oppose using AI to learn programming?

No. He describes AI as a potentially effective teaching assistant when students use it to explain concepts or help them get past technical roadblocks, rather than asking it to write all their code.

Why does Gross tell students to write code themselves?

He says writing code develops a practical understanding that helps students read and evaluate code. Without that ability, he warns, they may struggle to spot unsuitable approaches or control systems built with generated output.

Does the essay prove that programming jobs are safe from AI?

No. Gross offers his judgment that programming will remain viable, but the essay contains no job-market data or employment forecast. It also does not establish how AI will affect specific roles or when changes might occur.

Source: hn