A lot of people might also have coherent reasons to think that analogy applies equally to the US House of Representatives and the US Senate.

Peoples’ feelings about the nations they are born into and told to love from birth are complex and multifaceted. The people I know who grew up in the USSR have both good and bad things to say about it, just like the people I know who grew up in the USA (like me) at the same time (the 1970s-1990s) have both good and bad things to say about it. And that isn’t just about our own experiences growing up in these respective nations, but about learning our birth nations’ true histories, and how closely (or not) the ideals espoused by their founders and politicians and important figures in their histories were reflected in their actions.

Thus I really, truly do believe it’s ambiguous for someone to say, without any further context, that they find an image of a legislature with some screen shots of an IDE placed into it offensive. Is it offensive because it’s referencing a body they consider evil or is it offensive because it’s trivializing a body they consider good? Without context it’s impossible to know, and acting like everyone shares the same context about this is just refusal to engage with the world as it is rather than the world as you’d like it to be.

Given the existence of both groups I think just the claim that it’s offensive, without explaining why, is ambiguous and just reacting defensively doesn’t address that.

You’re the one who made the statement. It’s on you to support it.

How so?

With respect to the former, “fraud” is a shorthand for “fraudulent misrepresentation,” which is what you’re doing when you take someone else’s IP and try to contribute it to a project without securing the right to do so. It can be read as implicit in the attempt to contribute to the project that you have secured this permission (or do not need to, because the work is original to you). Whether the code came out of an LLM or was copied from another project or Stack Overflow doesn’t matter, it’s that you’re misrepresenting the rights you have that’s the fraudulent part.

For the latter, I specifically pointed out that the gain from fraudulent misrepresentation need not be monetary. The gain can be reputational or any other sort of benefit. For example, someone pretending to a fictional person to gain access to a space they otherwise wouldn’t is still committing fraud.

Finally, you’re wrong about whether the output of an LLM infringes copyright of material in its training set. Just running a copyrighted work through an LLM does not remove the copyright on that work if reproduced by the LLM.

When you send a patch or pull request to a project, you’re saying (implicitly) that you have the necessary rights to contribute the intellectual property it contains. If you used an LLM to “generate” some of it, that is not necessarily the case.

A similar situation would occur if you agreed to pay someone else to create a patch, and then submitted it under your own name without paying them. Because it’s a work for hire, it’s not yours until they’re paid for it, so you’re fraudulently misrepresenting your rights to that patch to the project. If you did pay the creator, you don’t have to attribute them unless it’s in the contract between you and the creator, or unless the project requires such attribution.

The LLM the person used was trained on a very large corpus of Open Source code, and reproduced that code exactly. Just like LLMs have reproduced chapters of books and articles from the New York Times exactly.

It’s clear on its face that LLMs can and do store and reproduce copyrighted works; using a form of (somewhat) lossy data compression. And using a lossy stochastic or perceptual form of compression to reproduce a copyrighted work doesn’t somehow make it not storage or reproduction, otherwise sharing MP3 files wouldn’t be copyright infringement.

Anyone engaging in responsible risk management should assume that anything LLM-generated is infringing until determined otherwise by the courts, not the other way around.

They are in fact committing fraud if they do not attribute the code in their commit properly, because by committing it they’re claiming to have rights by virtue of authorship that they do not have. (Namely, the right to contribute that code to the project,.) They may also be committing copyright infringement, depending on the copyright and license status of some code they found via Google or Stack Overflow.

It’s always fascinating to me to see how many people on Hacker News have such extremely poor understanding of how intellectual property actually works, and how misrepresenting themselves or their work can actually have consequences.

Often this is also spelled out in a project’s contribution guidelines, and some projects have even had more explicit copyright assignment policies they required contributors to agree to, but the lack of such guidelines or assignment policies does not mean the custom as normally observed in the field is irrelevant.

You might consider that there is a very large incentive by the large and public players in this market to promote the idea that this is not true, that they consider themselves large and powerful enough to actually flout the law, and that they plan to use the argument that enforcement will be too damaging to the economy to make their view the “new normal.”

This playbook has been run before, by Uber and Lyft, by AirBnB, by Tesla with “FSD,” and so on. It’s very clearly the approach being taken.

Unfortunately, a large number of people are being told—and here, you can see many who believe it—that the output of an LLM either carries no copyright or is copyright by the one prompting it. In other words, even right here on Hacker News it’s widely believed that LLMs “launder” copyright.

It also is copyright infringement, because what the LLM “generates” are actually portions of its training set, which were covered by copyright. Just passing through an LLM does not remove that copyright from that work.