Re:Same as always (Score 1)

I think this is essentially correct: the core subjects of CS education are still needed, like mathematics for the other engineering professions and other career-specific programs largely taught in the first two years of engineering or IT-related degree schools. Algorithms, databasing, digital devices/logic, still have to be taught for the students to have a foundation. But that's not the issue. That portion of CS learning that most call "coding training"--just enough learning to code by hand: these really aren't what I'd call entry-level engineering or related coding-related jobs in business and IT, etc. But these jobs are the most affected by current prompt-based code development. (As my graduate software engineering profs used to say: there always seems to exist an insatiable demand for low quality but extremely low cost software.) So the issue is "what do you teach in these lower level classes to replace coding classes?" You teach the students to use LLM/GPT tools to do a small survey of different languages in their first entry-level class (as is done now)--because the students can produce a lot more working code quickly and with less effort, so more examples can be assigned and completed in the same class as before. Teach them the process of using LLMs/GPTs to do this. You need to get that out of the way so it then winds up on their resumes--even if they leave school early before graduation to take jobs. The way I view it is that it actually makes learning more relevant since in-depth classes in specific (high level) languages can largely be bypassed by the bulk of CS students. Only specific sub-specialties in CS might need specific language training/inspecting for code quality specific to the languages-not every student in CS.