I've been using Claude Code for a couple of days

This is very far from my experience. It's been impressive on a few things but as of now I'm on day 3 of trying to get it to fix an issue in an open source library I maintain that I haven't had time to deal with. I'm in an endless loop where it keeps having the same epiphany about the root cause, then it implements a "fix", then tries to verify via a test script, then when it runs into difficulty it keeps adding workarounds to the test script. I can't get it focused on the fact that it's acting on behalf of the library author and that hacking the test script has no value.

I am not disposed to AI skepticism. I'd love it if a tool existed that worked as this guy claims. Claude Code is the best tool of its type that I've worked with but I'd put it at "on balance a time-saver in a lot of cases, way more trouble than it's worth in many others".

LLMs seem to work a lot better with statically typed languages where the compiler can give feedback.

I do not know how to code, nothing beyond the simplest things but I am obsessed with AI development. Several month’s ago I decided to see if Claude could do the work for me.

My test was to create a tower defense game that would run in a web browser. It took me about eight prompts, each time refining it, including agreeing to the suggesting that Claude recommended and seeing Claude agree with me on the bugs that I pointed out.

It was mind blowing. It’s not pretty at all but is recognizable as a TD game. I thanked Claude and said that was enough and Claude signed off as well saying, well if you’re going to continue to refine it, do these four things. I was really stunned.

Yeah, so I'm a 'hacker' (MIT definition here). I've only taken a single class in javascript with Sun based workstations about 20 years ago now (god, I'm old). I hated it.

All my work now is in python and SQL now and though I've watched a lot of youtube videos and plunked at StackOverFlow for ~15 years, I've never had formal education in either language. Like, it takes me about as long to set up the libraries and dependancies in python as it does to write my code. My formal job titles have never had 'programmer' in them.

My code, as such, is just to get something done. Mostly this is hardware interfacing stuff, but a little software too. I've had my code get passed up the chain and incorporated into codebases, but that's only happened a handful of times. The vast majority of the code I write is never seen by others. My code hasn't had to be maintainable as I've never really had to maintain it for more than 5 years max. I've used git on projects before, but I don't really see the need these days. The largest program I've written is ~1M lines of code. Most code is about 100 lines long now. I almost always know what 'working' means, in that I know the output that I want to see (again, mostly working in hardware). I almost never run tests of the code.

I've had the same issues you have had with LLMs, where they get stuck and I have to try to restart the process. Usually this happens to me in about 20 back and forths. I'm mostly just pasting relevant code snippets and the errors back into the LLM for a while until things work for me. Again, I know what 'working' looks like from the start.

Typically, I'll start the session with an LLM by telling it the problem I have, what I want the code skeleton to look like, and then what I want the output to look like. Then it will give the psuedo code, then I walk it through each portion of the psuedo code. Then I get to errors and debugging. Usually about half of this is just libraries and versions in python. Then, I get to the errors of the code itself. I can typically find what line of code is causing the error just from the terminal outputs. I'll talk with the LLM about that line of code, trying to understand it from the error. Then, on to the next error. Repeat that process until I get the working output I desire. I'm never expecting the right code out of the first LLM interaction, but I am expecting (and seeing) that the time it takes to get to working code is faster.

The time it would usually take me to get through all this before LLMs was about 2 weeks of work (~80 hours) per project. Now it takes me about half a day (~4 hours), and it's getting faster.

I'm not in the camp of thinking that AI is going to take my job. I am in the camp of thinking that AI is going to finally let me burn down the list of things that we really need to do around here.

Thank you for the reply!

This surrendering to the LLM has been going around a lot lately. I can only guess it is from people that haven't tried it very much themselves but love to repeat experiences from other people.