I'm curious about how artificial intelligence and language models have changed the landscape of our development jobs. Have you noticed any significant improvements in productivity or ease of work since using these tools? Is developing software noticeably simpler now, or has it brought about new challenges?
5 Answers
Overall, I feel it's about zero change. Some repetitive tasks are done faster, but for complex ones, I end up getting mixed results from the AI and have to do a lot myself anyway. Plus, I've tried multiple LLMs, and honestly, they don't seem to offer any significant advantages over one another.
It's not that it's easier, but everything feels faster now. I save time not having to look up syntax for different libraries, and I like using AI to summarize my code in plain terms for stakeholders.
But doesn't being faster mean it's easier?
Honestly, I find my job has become trickier. I often have to fix code generated by others using LLMs who don't fully understand what they're doing. I foresee a nightmare down the line when significant changes are needed to these poorly generated codebases.
For me, it's about managing expectations. I get more work done, but my employer seems to think that LLMs mean I can churn out everything instantly. This isn't the case; they need to understand it doesn't actually reduce the effort required.
I don’t think it’s easier at all. Sure, I get through more tasks, but the workload has just increased. It reminds me of the Red Queen Problem; I'm running just to stay in the same place!
I feel that way too. My job keeps saying we should be faster now, but I keep reminding them it still takes work to get LLMs to do what you need.
Yes! My workplace keeps telling us using LLMs should speed things up, but it still takes time to get them to do what I need, and often I just end up doing it myself.