AI is destroying my CS degree

Purpose and Value of a CS Degree

• Many argue a CS degree was never meant to be a complete job toolkit, but a foundation in fundamentals and thinking about systems and data.
• Some see it mainly as a signaling credential and “foot in the door”; others value it as an end in itself for deep, structured learning.
• Several posters say the degree’s practical value has dropped because AI fills knowledge gaps that used to require humans, especially at the junior level.
• Others insist fundamentals (algorithms, systems, reasoning) still matter, especially to use AI effectively and design robust systems.

AI’s Impact on Jobs and Career Entry

• Strong concern that AI is “murdering” the junior market: easier to hire fewer seniors armed with AI than many juniors.
• Some predict fewer, more “elite” engineering roles, with title inflation among weak “seniors” being corrected.
• A minority frames this as a “market correction” against people in it just for the paycheck, which others call gatekeeping and insensitive to financial realities.
• Suggestions include specializing in a domain, cultivating sales/relationship skills, or pivoting away from CS entirely; others think that’s too pessimistic.

AI in University Teaching and Assessment

• Multiple commenters say the real problem in the linked story is poor teaching practice: AI-generated assignments not properly checked, errors fixed late, chaotic and shifting AI policies.
• Some defend that faculty are under heavy institutional pressure and facing unprecedented uncertainty about what and how to teach.
• Large-scale, cheap lecture models are seen as increasingly incompatible with AI-era cheating; oral exams and in-person, closed-notes assessments are proposed but are costly and hard to scale.
• There’s worry that this will widen inequality between expensive, small-seminar institutions and mass “big state” universities.

Using AI for Learning CS

• Enthusiasts describe AI as a 24/7 tutor that explains “what/why/how,” accelerates learning in CS and math, and helps weaker students more than traditional courses did.
• Others are skeptical that AI can teach computer science (as opposed to coding snippets), arguing it tends to give answers without building deep conceptual understanding unless used very carefully.
• Various strategies are suggested: doing work first then comparing with AI, having AI pose problems but hide solutions, using it for Socratic/maieutic dialogue, and focusing on depth over tool-churn.

Software Engineering Practice in the AI Era

• Practitioners report AI now writes much of their code, but human value remains in architecture, tradeoff reasoning, explaining designs, and debugging.
• AI is seen as excellent at boilerplate and patterns but still weak at system-level design and long-term complexity management.
• Some fear that if AI becomes the main productive asset, individual engineers become interchangeable; others think the ability to work well with and sometimes without AI will differentiate top engineers.
• Advice to students includes: master end-to-end software engineering, keep a high-level mental model of systems, build interdisciplinary/business skills, and treat AI as an accelerator, not a substitute, for understanding.

Broader Concerns and Emotional Reactions

• Many express frustration with churn, bureaucratic AI mandates, and the feeling that enjoyable “doing the work” is turning into “project managing AIs.”
• Others are energized by the rapid learning and experimentation AI enables.
• There is no consensus on whether AI is net-destroying or transforming CS careers; most agree the transition is chaotic and the long-term equilibrium remains unclear.