> The biggest problem with AI PRs is the sheer amount of churn, extra code and lack of intent with the PRs it generates.
But this isn't an LLM problem; this is a problem of undisciplined engineers who feel they need to cram extra stuff in a PR. If an engineer doesn't look at the output of the LLM and generate extra work then it's still on them, right?
> The only way I can describe the latter is that an AI-only PR feels to me like a painting where everything is high detail - and you have to comb over each part before you understand why it's there because so much is superfluous
This just indicates that the engineer doesn't know how to use the tool. Hell they can ask the LLM to split the work into focused PRs and Claude will be happy to do it and the results might no even be half bad.
> Also when I'm _using_ the agent; at least 50 percent of my time is spent telling it to stop with it's approach so it doesn't go down a useless rabbit hole and waste tokens.
If this is happening often then the tool is probably not fit for the job.
I don't believe so - it's not as though the original prompts asked for extra code churn (note that as soon as you look and edit the LLM code output extensively it ceases to be vibe coding, which I was talking about in my OP).
I'm not talking about extra feature s; I'm talking about for the same single feature the code is either convoulted because the algorithm is overly complicated or the abstractions are just wrong for the domain.
The PRs typically are already focused in that they address a single feature; or at least a single "usable" feature in a complex system which necessarily has a lot of connected parts and behaviors.
> then the tool is probably not fit for the job.
Perhaps; but with an LLM I haven't found which jobs it _does_ work for and which it doesn't. I already use planning mode extensively; and capture the major points, but then it makes a stupid decision mid implementation and just starts churning.
> It requires discipline, which varies a lot between developers, between projects, current mood, and so on.
In the beginning you might be careful doing small changes, but after a while you might get more tempted to accept the output for what it is, because ultimately that's much easier.
Counterpoint: how is this any different from how things were pre-LLMs? I have seen, in the same codebase, some throughly well-written and tested PRs that read like Shakespeare and some of the laziest slop that even no LLM would ever write because humans have an unlimited capacity for laziness.
You catch the bad stuff through oversight, process, automated and manual checks, and the ultimate threat that your job depends on your ability to deliver so you better allocate at least enough energy into this so that you can ship moderately working code.
Helium? Rare earth minerals? Having to mine ever deeper because there are essentially no easily accessible mineral deposits? The fact that mining has enormous costs and the potential to permanently destroy sources of fresh water?
This is likely a very significant factor as urbanization has been extremely rapid, and historically cities kept their populations afloat by a constant influx of people from rural areas.
> You don't need to have life all figured out before you have kids.
Sure, you just need to do away with international trips, going out, losing your group of friends, losing your chances for higher education and career progression and all of the associated prestige.
Yes, if you value spending time with parents of same-aged children. My social life is still fine, but the people I spend my time with are competely different. Not better, nor worse, just entirely different.
The results are in existing products being built out faster, in more internal tooling that would have previously not been cost-effective. Almost every software engineering shop is just doing the same thing as before but faster, as AI doesn't really give you an edge in the product side of things.
But this isn't an LLM problem; this is a problem of undisciplined engineers who feel they need to cram extra stuff in a PR. If an engineer doesn't look at the output of the LLM and generate extra work then it's still on them, right?
> The only way I can describe the latter is that an AI-only PR feels to me like a painting where everything is high detail - and you have to comb over each part before you understand why it's there because so much is superfluous
This just indicates that the engineer doesn't know how to use the tool. Hell they can ask the LLM to split the work into focused PRs and Claude will be happy to do it and the results might no even be half bad.
> Also when I'm _using_ the agent; at least 50 percent of my time is spent telling it to stop with it's approach so it doesn't go down a useless rabbit hole and waste tokens.
If this is happening often then the tool is probably not fit for the job.
reply