Skilling for a post-AI bubble world
retained text
I have a BS in CS from a good school and have 6 YoE in mobile platform development. My last job was resistant to AI for a while and I was laid off before my company did any AI assisted coding trials so I've never experienced the hell that is having to token max and pump out low quality code en masse or be fired. That being said, I've experimented with agentic software development and I am constantly disappointed with whatever the new hotness is. At this point, I use it at most for code review, writing documentation, and for bug fixing reference because when I give it any sort of agency it sticks it's fingers where it doesn't belong and writes code that would have made my old coworkers call a wellness check on me. i'm aware of the current reality of AI in software development and I want to acclimate to the current market expectations. My problem is that anyone aware of the economic reality of LLMs can see that having agents write everything forever in it's supremely token inefficient loops is just not feasible when we have to pay what the compute is actually worth. My question: for anyone working in an environment that is using but limiting LLM usage (possibly with a local LLM), what does the workflow look like? Do specific coding language skills matter as much as they use to? Does leetcode still determine if you deserve a job for memorizing trivia or is there a new baseline for technical interviews? Are managers too fried by AI to ever come back to reality? For the record, I avoid FAANG companies because I value my free time so I'm not concerned with whatever those goons are doing. I don't expect to make the boogaloo bucks of yester-year, I just don't see myself switching careers anytime soon and know Walmart still needs app guys.
document fetched 2026-11-06 00:23 · canonical https://reddit.com/r/cscareerquestions/comments/1xoprcs/skilling_for_a_postai_bubble_world