- Exemplary
- Work has changed massively for me in the past year, but that's because of a combination of two factors: changed companies, and current company is using AI heavily. Irrespective of employers, professionally speaking, in the past 1.5 to 2.0 years, with respect to AI, I have gone from "near zero use, just testing the waters" to "prime seating on the bandwagon".
- I'm enjoying my work more, but a big part of that is more about me (change of mindset, attitude) than the industry changing.
- I'm working a little more than before, but that's my choice. Also, I'm finding a fair bit of my time is spent herding AI sessions (multiple in parallel).
- I am both hopeful and fearful for the mid-term future of SWE. Few can deny that recent developments and the current state of AI/LLMs with respect to SWE are an upheaval, an historic kind of change. I don't think humans will be completely yanked from the overall SWE process (yet?), but I do think that demand will shrink by anywhere from "a bit" to "moderately". At the same time, I think a certain subset of SWEs will stay in demand, or even be more sought out, namely, the senior ones who have years of experience doing things the traditional way (pre-LLM), and who can spot when an AI does something wrong, and guide it back on track. I fear for the quality level of software that will be created by those who don't have this knowledge, experience, and intuition. I also don't know if the fledglings in the industry (who don't have these things) can pick these things up, or at least, whether they can pick them up without serious, concerted effort and discipline ("no, I won't commit 5000 lines of unreviewed AI-generated code"). That all said, I think the possibilities are interesting and even exciting, if the code-writing (and planning, and system-architecting) capabilities of AI continues to improve at the same pace it has been in the last year or so.
- Advice? Don't dismiss or ignore AI in SWE. At least give it an honest try and make an assessment. Some people are drinking the kool aid, other people are abstaining on moral grounds, some people have had bad experiences, but it only makes sense to make a decision based on facts ("I tried it and..."), not assumptions. IMO, those who have not tried it are missing out. Furthermore, I think there will be a rising trend in the industry that hiring will be based on AI(-using) skills, and, in my experience, that actually takes a bit of time to develop and hone, similar to other skills in our field. I don't think it's something that you just get good at in under a week after you flip a switch and decide to start using AI.
- My career trajectory... I think the main thing that has and will change is how much I use AI/LLMs in my work. I don't see this going away, except maybe if they raise consumer prices to exorbitant levels. I don't see my company cutting AI even if they raise corporate pricing, though.
- Background: big company (>1k employees); "many" years of experience (don't want to date myself); I'm a senior SWE; working remotely.