AI execs keep giving advice that young programmers should focus on AI instead of doing internships, etc. this is abominable advice, for many reasons. I’m going to focus on two:

1.) Jobs teach you more than your discipline. At a job — even a remote one — you learn the insider particulars of your industry and a lot more about real development practices. Most students don’t get much experience working on teams, on larger projects, and that experience is beneficial.

2.) Even if you have to use LLM agents to generate code, your value-add to that relationship isn’t prompting skill: It’s programming skill. The people squeezing the most value out of these systems are seasoned programmers. Leaning on an LLM may well diminish that hard skill acquisition.

How do you feel about schools pushing it? My current school‘s programming curriculum added 30+ AI courses in the last two years, while, as you said, forego fostering working in groups/collaboration.

AI has been part of CS curriculum for many years. I took an AI class back in the '90s. 30+ courses seems really excessive, though. In all seriousness, I'm kind of worried we're going to lose a generation of college grads.

One shouldn’t spend any of their precious college education learning to integrate these tools into their workflow. The whole point of them is that they’re supposed to be easy. They are also changing so fast that anything you learn will be obsolete in 5 minutes. Focus on lasting things in school.

Learning about various types of AI, what they’re useful for, and how to build them is still a useful subject to study. But not before CS fundamentals, obviously. AI is a subfield. It’s not THE field.