Probably about 3 minutes? That's my main usage of these types of coding tools, honestly. I already know generally what I want to happen and validating that the LLM is on the right track / reached the right solution is easy.
I'm not saying that "Claude Code makes me 300% more effective", but I guess it did for this (simple) task.
Are you comparing only the generation time with your coding time? How are the figures if you include the necessary step of checking the generated code? And how do these times change if you are coding in a very complex environment?
To be clear, I explicitly picked this task because my judgement call was that it was going to be faster to use an AI than coding it myself. Checking the generated code was easy for me to do (whether I or the AI wrote it).
I don't know what you mean by "how do these times change if you are coding in a very complex environment", but to restate my original post: I'm fearful that Claude's additional autonomy will allow it to waste money (and time) pursuing useless ideas.
This shouldn't be factored in unless you never debug your own code. At any rate, I don't know about OP but usually when I get an LLM to write out some code, I also have it write out the tests for it as well.
I recently made some changes to a website generated by a non-technical user using Dreamweaver. The initial state was quite poor, with inline CSS and what appeared to be haphazard copy-pasting within a WYSIWYG editor.
Although I’m not proficient in HTML, CSS, or JavaScript, I have some understanding of what good code looks like. Through several iterations, I managed to complete the task in a single evening, which would have required me a week or two to relearn and apply the necessary skills. Not only the code is better organised, it’s half the size and the website looks better.
time spent is not the only question. how much thought it takes, however impossible that may be to measure, is another one. If an LLM assisted programmer is able to solve the problem without deep focus, while responding to emails and attending meetings, vs the programmer who can't, is time really the only metric we can have here?
My scrum/agile coach says, by parallelizing prompts, a single developer can babysit multiple changes in the same time slice. By having a sequence of prompts ready before hand, a single developer can pipeline those one after the other. With an IDE that helps schedule such work, a single developer can effectively hyper-thread their developmental workflow. If the developer is epoll'ing at 10x the hertz... that's another force multiplier. Of course context switches & side-channels are of concern, but a voice over my shoulder tells me that as long as memory safety is guaranteed, everything should turn up alrigd3adb33f.
