I work as a software engineer. A lot of my specialization is in being able to write good code, but AI has been getting better and better at that lately. I hope I maybe have a backup as working on AI because logically speaking it seems like AI research would be the "last job to get automated", but I also know a lot of non-AI coding, e.g. backend development, frontend development, programming language theory, data science, etc., and these skills might get obsoleted by AI. I'm not really good at ops work so I'm not counting that.
Resolution criterion: If I subjectively feel that my opportunities for coding jobs have become worse due to AI, for at least 6 months duration starting before 2032, this market resolves YES. Otherwise it resolves NO. If I have great opportunities within AI, this market may still resolve YES, as I will treat AI as a separate field from coding. (Unless we're speaking like, GOFAI, scaffolding, etc..) AI does not need to be as skilled as a programmer; if it is cheaper or makes bad coders as productive as me, then it might still be negatively affecting my career.
I will not be betting in this market.
Related questions
Wait are you guys betting on the basis of the Metaculus strong AGI market? You are aware that that market uses a too weak criterion, right?
https://manifold.markets/tailcalled/is-metaculus-strong-agi-question-to
@placebo_username I feel like this is going to depend a lot on the nature of the job. I don't think this is a likely outcome so I don't have good predictions for the nature of the job that you have in mind. My main question would be, why do they need a specialist to perform this job?
@tailcalled In the three months since I originally wrote this comment, at least two friends have switched from traditional coding jobs to prompt engineering roles. So I don't think this is a terribly unlikely outcome.
As for why a specialist is needed, I assume their employers expect someone with ML experience to be better at designing good prompts. I don't know if they've chosen the right specialists for the job; maybe in the long run it will turn out that prompt engineering skills correlate more with management experience than with coding ability.
@placebo_username If prompt engineering replaces programming, then the way I will treat it depends on how it works. If it turns out to be more like a manager role, then I will probably resolve this market YES. But if it is similar to coding, I will probably resolve this market NO.