I have tried using chatGPT a few times. Early on I had no luck but lately it has been a better experience. Maybe it has gotten better or maybe I have started asking better questions. I used it the other day to help me with some regex strings and I even enjoyed the experience.
cloudking · 12h ago
Have you tried any of the agentic coding editors like Cursor, Windsurf etc?
skwee357 · 5h ago
I tried, but I don’t find them useful.
I tried supermaven, and llama.cpp, as autocomplete in steroids, but they got annoying pretty fast.
I now use ChatGPT and Claude to ask questions about thing I don’t know how to do, or ask them to rewrite/optimize a specific piece of code.
scrapheap · 5h ago
I'm actively not using AI coding assistance as I think it would weaken my ability in the long run.
diatone · 7h ago
Yes definitely, it tends to autocomplete patterns that extend beyond basic intellisense.
Also there are scenarios where wiring up exhaustive cases would be tedious or require clever use of a vim macro, that a call to AI tends to do in slightly less time but much less mental overhead - allowing me to move on with less fatigue. Overall I cover more ground this way.
It tends to excel when I’ve structured code in a way that’s easily copyable or extensible, such that I can ask the AI to replicate pattern A but for API B. Again this saves me dropping into a low level understanding of an API integration unless I absolutely need to. I tend to check if I need to based on test cases, observability, and performance monitoring —- all the same shit I’d usually use to determine whether or not I should edit the code manually anyway.
Overall, obvious net benefit when used prudently. I’ve seen juniors use AI to write messy code that’s challenging to debug, understand, and maintain. It really is just another tool. You can use it to make your life easier or you can use it to make your life harder.
sherdil2022 · 11h ago
I use - but cautiouslyc, that is I don’t do ‘vide coding’ (I hate the term). Why? I started building a project in Rust - which I was very new. Using LLMs helped me speed up the process of development and testing, and I verify what was spit out by the LLM (unlike vibe coding).
Granted, it is frustrating at times since I end up spending lot of time verifying and iterating.
But overall it is definitely efficient to use LLMs for developing software.
cloudking · 10h ago
Do you think you are more productive overall?
sherdil2022 · 1h ago
Hard to say but using LLMs properly would make developing things faster In general - yes.
In my case, my baseline mood and motivation fluctuates - so I only. see a small increase in productivity overall - if at all.
I have also noticed excitement to build something new and exciting wore off a bit - because it is ‘just an LLM prompt or two away’.
owebmaster · 11h ago
Yes. I can make it generate code that is on par with my own - not sure if this is good or bad, but it help me do my work faster and with better quality.
I tried supermaven, and llama.cpp, as autocomplete in steroids, but they got annoying pretty fast.
I now use ChatGPT and Claude to ask questions about thing I don’t know how to do, or ask them to rewrite/optimize a specific piece of code.
Also there are scenarios where wiring up exhaustive cases would be tedious or require clever use of a vim macro, that a call to AI tends to do in slightly less time but much less mental overhead - allowing me to move on with less fatigue. Overall I cover more ground this way.
It tends to excel when I’ve structured code in a way that’s easily copyable or extensible, such that I can ask the AI to replicate pattern A but for API B. Again this saves me dropping into a low level understanding of an API integration unless I absolutely need to. I tend to check if I need to based on test cases, observability, and performance monitoring —- all the same shit I’d usually use to determine whether or not I should edit the code manually anyway.
Overall, obvious net benefit when used prudently. I’ve seen juniors use AI to write messy code that’s challenging to debug, understand, and maintain. It really is just another tool. You can use it to make your life easier or you can use it to make your life harder.
Granted, it is frustrating at times since I end up spending lot of time verifying and iterating.
But overall it is definitely efficient to use LLMs for developing software.
In my case, my baseline mood and motivation fluctuates - so I only. see a small increase in productivity overall - if at all.
I have also noticed excitement to build something new and exciting wore off a bit - because it is ‘just an LLM prompt or two away’.