[hiAndrewQuinn]

My wife is currently enrolled in a computer science program, and AI was absolutely vital for getting her over the 0 to 1 known programming languages chasm (Python, in her case). It has been less helpful (still very!) for C++ after that, but C++ after Python is much easier than C++ with no background at all. So I don't think ignoring the "just programming fundamentals" phase is the right call. I think that's where the greatest difference is going to be by far.

I think a great many more experienced people forget just how hard it is to learn your first, just how many new unintuitive concepts appear no matter what language you choose. Many, many more people are now able to learn that without getting frustrated to the point of giving up, because they all have a digital Aristotle in their pocket to guide them.

[hiAndrewQuinn]

After getting past the fundamentals, of course, I think the runaway feedback loop of AI will continue to consolidate real world programming work into an even smaller pool of currently-popular languages. JavaScript, Python, Go, Rust, Java, C, and C++ will all be in this cannot. Bash, too, if you count that - current AI tools are much much better at Bash than e.g. Fish, much to my dismay. So we'll see educational resources follow suit.

We will probably also see renewed interest in the "out of the box" options available on common platforms like Debian. By my count those would be awk, sed, and perl. [1] There's just something very satisfying about being able to spin up a bare server and get useful work done with just the essentials!

[1]: https://hiandrewquinn.github.io/til-site/posts/what-programm...