Why College Programming Professors Require AI Use (And What Parents Should Know)
Last month, my 20-year-old son Sam dragged his duffel into the kitchen after a three-hour drive home for spring break, dumped his backpack on the dining chair, and pulled out his laptop before I even handed him a cold soda. He turned the screen toward me, eyebrows raised, and said, “Mom, my data structures professor is requiring us to use AI to write half our code this semester. What is that about?”
I was folding his clean laundry from a semester away, and I paused mid-t-shirt fold. Back when I took an introductory comp sci night class 25 years ago, getting caught with someone else’s code got you an automatic fail. Now professors were making students use AI to do the work? My first thought, which I didn’t say out loud, was that this couldn’t be real learning.
I asked Sam to walk me through what the assignment actually looked like, not just the requirement. He spun the laptop around, pulled up his course page, and pointed to the syllabus. First, you prompt AI to generate a working base for the sorting algorithm assignment we’re doing, he said. Then you test it. Ninety percent of the time, it works for the simple test cases, but it breaks for anything tricky. My job is to find the breaks, fix the code, and write two pages explaining what the AI got wrong and why my fix works. Seventy percent of our grade is that explanation, not the working code.
That night, after Sam went out to get burgers with his high school friends, I texted my old college roommate who teaches introductory and upper-division programming at a public university an hour away. I asked her straight up: Is this just letting kids cheat now? She called me back 10 minutes later, and I could hear her making tea in the background while she talked.
She said 10 years ago, the most important skill a new programmer could have was writing clean, correct syntax from memory. If you messed up a semicolon or got the loop order wrong, your code didn’t work, and you had to spend hours hunting down the mistake. That’s why we made students write every line from scratch, she said. Now? AI writes perfect syntax in 10 seconds, every time. Every entry-level programming job I see requires using AI to speed up base code writing. If we still test students on writing every line from scratch, we’re testing a skill they’ll never use once they graduate.
That got me thinking. A week later, Sam mentioned he’d had his first internship interview for the summer. The interviewer didn’t ask him to write a sorting algorithm from scratch on a whiteboard, like I’d heard interviewers used to do. They asked him to walk through a time he’d fixed a flaw in AI-generated code, and how he checked it for security gaps. That’s when it clicked for me: this isn’t cheating. It’s teaching the skill that’s actually needed right now.
As a parent, it’s easy to panic when your kid comes home with a new school requirement that doesn’t match what you experienced. I’ve talked to a handful of other parents of CS majors in the last few months, and almost all of them had the same first reaction I did: this isn’t real learning, they’re letting kids get away with everything. But there are a few simple things you can do to understand what’s actually going on, instead of jumping to conclusions.
First, when your kid says a professor requires AI for an assignment, ask them “What do you have to do after the AI generates the work?” instead of immediately asking if it’s cheating. That’s a small shift, but it gets you talking about the actual work your kid is doing, not the vague fear we all have about new tools changing education. Second, ask them to explain the problem they fixed in the AI output to you. You don’t need to understand a single thing about programming to do this. Sam walked me through his first assignment, where the AI’s algorithm crashed when it had to sort a list of 10,000 entries because it used too much memory. I didn’t follow the part about recursion optimization, but I could tell he understood exactly what went wrong, and how he fixed it. If he’d just copied the AI’s code, he wouldn’t have been able to explain any of it.
Third, skip the urge to say “this is crazy, they should teach it the old way.” It’s okay to be confused, but saying that out loud just makes your kid second-guess their work, and it closes off conversation. Instead, use the phrase “That sounds really different from when I was in school, tell me more.” It’s simple, it doesn’t judge, and it gets you the information you need to understand what’s going on.
None of this is perfect, of course. My roommate told me that professors are still figuring out the right balance. A lot of programs now keep the “write every line from scratch” rule for first and second year intro classes, so students still build a base understanding of how code works before they start using AI to speed up work. Some classes have had growing pains, where requirements weren’t clear, and some students skate by by copying AI explanations. But professors are adjusting, just like they always have when new tools come along.
Last night, Sam sent me a screenshot of his grade on that first AI assignment: an A-, with a note from the professor that said “Good catch on the memory leak, that’s exactly what we wanted you to find.” I still don’t know what a memory leak actually is, and I don’t know how programming education will look 10 years from now, when Sam’s been working full time. But I stopped worrying a while back. The point of college isn’t to repeat the experience we had when we were 20. It’s to prepare kids for the world they’re actually going to work in. Right now, that world needs programmers who can spot AI’s mistakes, not just write code from scratch. For the moment, that seems like enough.
Please indicate: Thinking In Educating » Why College Programming Professors Require AI Use (And What Parents Should Know)