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

AI tooling must be disclosed for contributions

github.com

Post reply on HN
Print view View original

331–340 of 482 posts

Re: AI tooling must be disclosed for contributions

#331

Post by ekianjo »

ineedasername wrote: post #291
slg wrote: post #285

Earlier quoted context omitted.

>societal desire that generating IP continue to be a viable economic work product It is strange that you think the law is settled when I don't think even this "societal desire" is completely settled just yet.

Maybe I should clarify: Society, in general, supports the idea that writers, artists, film makers, coders, etc— everyone who creates IP- should have a place in the economy. Basically just that it should be possible to make a living and have a career at it. It can be spun different ways, and those differences are important, but this is the basic thing. This doesn’t seem like a disputable statement to me. For anyone wh…

> Society, in general, supports the idea that writers, artists, film makers, coders, etc

Coders don't get paid every single time their code runs. Why bundle different rights together?

Re: AI tooling must be disclosed for contributions

#332

Post by aspenmayer »

popalchemist wrote: post #113
alfalfasprout wrote: post #97

Earlier quoted context omitted.

> Courts (at least in the US) have already ruled that use of ingested data for training is transformative This is far from settled law. Let's not mischaracterize it. Even so, an AI regurgitating proprietary code that's licensed in some other way is a very real risk.

No more so than regurgitating an entire book. While it could technically be possible in the case of certain repos that are ubiquitous on the internet (and therefore overrepresented in training data to the point that they are "regurgitated" verbatim, in whole), it is extremely unlikely and would only occur after deliberate prompting. The NYT suit against Open AI shows (in discovery) that the NYT was only able to get p…

> No more so than regurgitating an entire book.

Like this?

Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book - https://news.ycombinator.com/context?id=44972296 - 67 days ago (313 comments)

Re: AI tooling must be disclosed for contributions

#333

Post by wahnfrieden »

nullc wrote: post #306
wahnfrieden wrote: post #288

Earlier quoted context omitted.

This whole original HN post is about ghostty btw Having the original prompts (in sequence and across potentially multiple models) can be valuable but is not necessarily useful in replicating the results because of the slot machine nature of it

> This whole original HN post is about ghostty btw Sure though I believe few commenters care much about ghostty specifically and are primarily discussing the policy abstractly! > because of the slot machine nature of it One could use deterministically sampled LLMs with exact integer arithmetic... There is nothing fundamental preventing it from being completely reproducible.

Can't do that with state of the art LLMs and no sign of that changing (as they like to retain control over model behaviors). I would not want to use or contribute to a project that embraces LLMs yet disallows leading models.

Besides, the output of an LLM is not really any more trustworthy (even if reproducible) than the contribution of an anonymous actor. Both require review of outputs. Reproducibility of output from prompt doesn't mean that the output followed a traceable logic such that you can skip a full manual code review as with your mass renaming example. LLMs produce antagonistic output from innocuous prompting from time to time, too.

Re: AI tooling must be disclosed for contributions

#334

Post by wahnfrieden »

kentm wrote: post #325
nullc wrote: post #151

Earlier quoted context omitted.

> is that they don’t know either It would be nice if they did, in fact, say they didn't know. But more often they just waste your time making their chatbot argue with you. And the chatbots are outrageous gaslighters. All big OSS projects have had the occasional bullshitter/gaslighter show up. But LLMs have increased the incidence level of these sorts of contributors by many orders of magnitude-- I consider it an open…

There was some post that comes to mind of an example of this. Some project had a security issue reported that was not a security issue, and when asking questions it became extremely obvious that someone was just feeding the conversation into an LLM. There was no security issue. I can imagine this is happening more and more as people are trying to slam in LLM generated code everywhere. Everyone promoting LLMs, especia…

The ghostty creator disagrees re: the productivity of un-reviewed generated PRs: https://x.com/mitchellh/status/1957930725996654718

Re: AI tooling must be disclosed for contributions

#335

Post by fluidcruft »

eschaton wrote: post #177
koolba wrote: post #56

Earlier quoted context omitted.

> You must ask "Do I trust where this came from?" And to answer that, you need to know where it come from. No you don’t. You can’t outsource trust determinations. Especially to the people you claim not to trust! You make the judgement call by looking at the code and your known history of the contributor. Nobody cares if contributors use an LLM or a magnetic needle to generate code. They care if bad code gets introduc…

You’re completely incorrect. People care a lot about where code came from. They need to be able to trust that code you’re contributing was not copied from a project under AGPLv3, if the project you’re contributing to is under a different license. Stop trying to equate LLM-generated code with indexing-based autocomplete. They’re not the same thing at all: LLM-generated code is equivalent to code copied off Stack Overf…

How does an "I didn't use AI" pledge provide any assurance/provenance that submitted code wasn't copied from an AGPLv3 reference?

Re: AI tooling must be disclosed for contributions

#336

Post by aspenmayer »

ekianjo wrote: post #331
ineedasername wrote: post #291

Earlier quoted context omitted.

Maybe I should clarify: Society, in general, supports the idea that writers, artists, film makers, coders, etc— everyone who creates IP- should have a place in the economy. Basically just that it should be possible to make a living and have a career at it. It can be spun different ways, and those differences are important, but this is the basic thing. This doesn’t seem like a disputable statement to me. For anyone wh…

> Society, in general, supports the idea that writers, artists, film makers, coders, etc Coders don't get paid every single time their code runs. Why bundle different rights together?

> Coders don't get paid every single time their code runs.

They do if they code the API correctly.

> Why bundle different rights together?

Why are mineral rights sold separately to most land deeds?

Re: AI tooling must be disclosed for contributions

#337

Post by eschaton »

bigyabai wrote: post #244
eschaton wrote: post #186

Earlier quoted context omitted.

Some courts at some levels. It’s by no means settled law.

Most licensed software is unsettled law, if we're being that pedantic.

Not really, no. If you’re specifically referring to, say, GPL or BSD or other Open Source licenses, it’s a bit more unsettled, but software licensing as a whole has several decades of case law at this point.

Re: AI tooling must be disclosed for contributions

#338

Post by eschaton »

koolba wrote: post #255
eschaton wrote: post #177

Earlier quoted context omitted.

You’re completely incorrect. People care a lot about where code came from. They need to be able to trust that code you’re contributing was not copied from a project under AGPLv3, if the project you’re contributing to is under a different license. Stop trying to equate LLM-generated code with indexing-based autocomplete. They’re not the same thing at all: LLM-generated code is equivalent to code copied off Stack Overf…

I’m not equating any type of code generation. I’m saying that as a maintainer you have to evaluate any submission on the merits, not on a series of yes/no questions provided by the submitter. And your own judgement is influenced by what you know about the submitter.

And I’m saying, as a maintainer, you have to and are doing both, even if you don’t think you are.

For example, you either make your contributors attest that their changes are original or that they have the right to contribute their changes—or you assume this of them and consider it implicit in their submission.

What you (probably) don’t do is welcome contributions that the contributors do not have the right to make.

Re: AI tooling must be disclosed for contributions

#339

Post by nojs »

freedomben wrote: post #18
rattlesnakedave wrote: post #12

In my personal projects I also require all contributors to disclose rather they’ve used an editor with any autocomplete features enabled.

Heh, that's a great way to make a point, but right now AI is nowhere near what a traditional editor autocomplete is. Yes you can use it that way, but it's by no means limited to that. If you think of AI as a fancy autocomplete, that's a good personal philosophy, but there are plenty of people that aren't using it that way

The line isn’t as clear as you might think, eg jetbrains has a mini on-device neural net powered autocomplete:

https://www.jetbrains.com/help/idea/full-line-code-completio...

Re: AI tooling must be disclosed for contributions

#340

Post by eschaton »

fluidcruft wrote: post #335
eschaton wrote: post #177

Earlier quoted context omitted.

You’re completely incorrect. People care a lot about where code came from. They need to be able to trust that code you’re contributing was not copied from a project under AGPLv3, if the project you’re contributing to is under a different license. Stop trying to equate LLM-generated code with indexing-based autocomplete. They’re not the same thing at all: LLM-generated code is equivalent to code copied off Stack Overf…

How does an "I didn't use AI" pledge provide any assurance/provenance that submitted code wasn't copied from an AGPLv3 reference?

It doesn’t, it provides an assurance (but not provenance) you didn’t use AI.

Assuring you didn’t include any AGPLv3 code in your contribution is exactly the same kind of assurance. It also doesn’t provide any provenance.

Conflating assurance with provenance is bogus because the former is about making a representation that, if false, exposes the person making it to liability. For most situations that’s sufficient that provenance isn’t needed.

Post reply on HN