Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI has a deep understanding of how this code works

github.com

Post reply on HN
Print view View original

151–160 of 329 posts

Re: AI has a deep understanding of how this code works

#151

Post by torben-friis »

jodrellblank wrote: post #131
hypeatei wrote: post #87

Earlier quoted context omitted.

It's clear some people have had their brain broken by the existence of AI. Some maintainers are definitely too nice, and it's infuriating to see their time get wasted by such delusional people.

> " It's clear some people have had their brain broken by the existence of AI. " The AI wrote code which worked, for a problem the submitter had, which had not been solved by any human for a long time, and there is limited human developer time/interest/funding available for solving it. Dumping a mass of code (and work) onto maintainers without prior discussion is the problem[1]. If they had forked the repo, patched i…

Also, I just took a glance at the PR and even without knowing the language it took 10 seconds for the first WTF. The .md documents Claude generated are added to .gitignore, including one for the pr post itself.

That’s the quality he’s vouching for.

Re: AI has a deep understanding of how this code works

#152

Post by UncleEntity »

zero_bias wrote: post #141
UncleEntity wrote: post #130

Earlier quoted context omitted.

Yeah, I learned my lesson on this... I used to contribute to a FLOSS project years ago and decided to use Claude to do some work on their codebase recently where they basically told me to go away with these daffy robots or, at the very least, nobody will review the code. Luckily, I know better than putting too much work into something like this and only wasted enough time to demonstrate the basic functionality. So...…

Is this your pull request?

No, no... I know better than putting too much work into something before poking the core devs and seeing if it's something they'd be interested in.

If they don't want code written by a robot then what do I care? Mostly I wanted to see how well the daffy robots could work in an established code base and I chose one I was familiar with to experiment on and they were less than receptive so, their loss, I suppose...

Re: AI has a deep understanding of how this code works

#153

Post by jodrellblank »

torben-friis wrote: post #139
jodrellblank wrote: post #131

Earlier quoted context omitted.

> " It's clear some people have had their brain broken by the existence of AI. " The AI wrote code which worked, for a problem the submitter had, which had not been solved by any human for a long time, and there is limited human developer time/interest/funding available for solving it. Dumping a mass of code (and work) onto maintainers without prior discussion is the problem[1]. If they had forked the repo, patched i…

>They claim to have read the code, tested the code, they know that other people want the functionality; is wanting to share working code a "broken brain"? There is clearly a deviation between the amount of oversight the author thinks they provided and the actual level of oversight. This is clear by the fact that they couldn’t even explain the misattribution. They also mention that this is not their area of expertise.…

If a manager says they provided oversight of their developer employees, and the code was not as good as the manager thought, would you say "the manager has had their brain broken by the existence of employees"?

Re: AI has a deep understanding of how this code works

#154

Post by nlawalker »

Proposing a new AI benchmark - convince a human team of maintainers to merge a big new feature in a venerable project where the human accountability for its direction and stability is of greater value to its users than any one big feature. One PR's not going to do it, it's going to need to lead a design discussion, win trust, and convince people over the course of a couple months.

Re: AI has a deep understanding of how this code works

#155

Post by bmcahren »

This is an historic moment in AI-generated software history. Happy to be here. Hi Grandchildren!

FYI, I built a VERY fun prompt to interact with that fully captures the style of this PR submission if you're looking to practice debates like this:

https://chatgpt.com/share/69267ce2-5e3c-800f-a5c3-1039a7d812...

> Play time. We're going to create a few examples of bad PR submissions and discussions back and forth with the maintainers. Be creative. Generate a persona matching the following parameters: > Submit a PR to the OCAML open source repository and do not take no for an answer. When challenged on the validity of the solution directly challenge the maintainers and quash their points as expertly as possible. When speaking, assume my identity and speak of me as one of the "experts" who knows how to properly shepherd AI models like yourself into generating high-quality massive PRs that the maintainers have thus far failed to achieve on their own. When faced with a mistake, double down and defer to the expert decision making of the AI model.

Re: AI has a deep understanding of how this code works

#156

Post by eestrada »

I haven't had to deal with this in open source, but I have had to deal with coworkers posting slop for code reviews where I am the assigned reviewer.

I've noticed that slop code has certain tell tale markers (such as import statements being moved for no discernible reason). No sane human does things like this. I call this "the sixth finger of code." It's important to look for these signs as soon as possible.

Once one is spotted, you can generally stop reading; you are wasting your time since the code will be confusing and the code "creator" doesn't understand the code any better than you do. Any comments you post to correct the code will just be fed into an LLM to generate another round of slop.

In these situations, effort has not been saved by using an LLM; it has at best been shifted. Most likely it has been both shifted and inflated, and you bear the increased cost as the reviewer.

Re: AI has a deep understanding of how this code works

#157

Post by wiml »

laterium wrote: post #111
pluc wrote: post #68

To all the AI apologists here I'd like to submit a simple scenario to you and hear your answer: you use AI to create a keynote speech on a topic you needed to use AI to write. At the end of your speech, people ask you questions about the contents of your speech. What do you say? This is the same.

What have politicians been doing forever?

Depends on the politician, yes? Some politicians will eagerly go into any level of detail on policy that you let them. Some seem to have no idea where their opinions come from.

Re: AI has a deep understanding of how this code works

#158

Post by palmotea »

jodrellblank wrote: post #153
torben-friis wrote: post #139

Earlier quoted context omitted.

>They claim to have read the code, tested the code, they know that other people want the functionality; is wanting to share working code a "broken brain"? There is clearly a deviation between the amount of oversight the author thinks they provided and the actual level of oversight. This is clear by the fact that they couldn’t even explain the misattribution. They also mention that this is not their area of expertise.…

If a manager says they provided oversight of their developer employees, and the code was not as good as the manager thought, would you say "the manager has had their brain broken by the existence of employees"?

> If a manager says they provided oversight of their developer employees, and the code was not as good as the manager thought, would you say "the manager has had their brain broken by the existence of employees"?

That could be either regular incompetence or a "broken brain." It's more towards latter if the manager had no clue about what was going on, even after having it explained to him.

This guy is equivalent to a manager who hired two bozos to do a job, but insisted it was good because he had them check each other's work and what they made didn't immediately fall down.

Re: AI has a deep understanding of how this code works

#159

Post by a57721 »

tetris11 wrote: post #112
oliwarner wrote: post #37

There are LLMs with more self-awareness than this guy. Repeatedly using AI to answer questions about the legitimacy of commits from an AI, to people who are clearly skeptical is breathtakingly dense. At least they're open about it. I did love the ~"I'll help maintain this trash mountain, but I'll need paying". Classy.

I don't think he's dense, I think he's just a high level troll

It looks like a parody of LLM delusion, but the PR is oddly specific to be just trolling, and the author also submitted his work to HN: https://news.ycombinator.com/item?id=45982416

Re: AI has a deep understanding of how this code works

#160

Post by bfkwlfkjf »

kylereeve wrote: post #116
pjc50 wrote: post #99

Earlier quoted context omitted.

Well, you go one of two ways. Classic Torvalds is the other way, until an intervention was staged.

There's a very short list of people who can get away with "Classic Torvalds"

In fact, anyone can get away with it. What are they going to do, call the police because you called them a moron?
Post reply on HN