Original link: Harvard Is Training Us for a World That No Longer Exists / The Hardvard Crimson.
I very much disagree with this:
Students concentrating in Computer Science face a similar dilemma. Computer Science concentrators are required to enroll in algorithm-based classes such as COMPSCI 1200: “Introduction to Algorithms and their Limitations,” a course focused on theoretical computer science. While theoretical concepts are important, I wonder whether they should be the focus at a moment when computer science concentrators face such difficult career prospects?
AI continues to dominate, with some companies even proclaiming that within months, AI will write the majority of code used by software engineers. Rather than spending semesters dedicated to learning the theoretical framework, it would be more valuable for students to learn skills directly applicable to the job market like prompt engineering and appropriate ways to integrate AI. If ChatGPT can generate code in thirty seconds, there seems to be little benefit for a student to onerously learn how to write it themselves.
Quickly generate some code whether fast typing, copy paste from Stack Overflow or asking a LLM to generate code is easy. Everybody can do that. Even the boss can ask a LLM to generate code. There is no well paid jobs in that.
On the other side I believe there are still well paid jobs for those that truly understand IT systems, which require some theoretical understanding as well. LLM's are actually pretty dumb.
AI is not a good reason not to study theoretical computer science. AI is a good reason not to "learn how to write code in 24 hours".