From the creators of easy money laundering (crypto bros), we now bring you easy money laundering 2: intellectual property laundering, coming to a theatre near you soon!

>From the creators of easy money laundering (crypto bros),

Is there even any evidence that "crypto bros" and "AI bros" are even the same set of people other than being vaguely "tech" and hated by HN? At best you have someone like Altman who founded openai and had a crypto project (worldcoin), but the latter was approximately used by nobody. What about everyone else? Did Ilya Sutskever have a shitcoin a few years ago? Maybe Changpeng Zhao has an AI lab?

Elmo Musk was pumping and dumping dogecoin for years before pivoting to starting an LLM Ponzi.

Okay, but "fad-use-of-GPU bros" doesn't roll off the tongue as well.

> and had a crypto project (worldcoin)

That was a biometric surveillance project disguised as a crypto project.

> Is there even any evidence that "crypto bros" and "AI bros" are even the same set of people

No, the "AI" people are far worse. I always had a choice to /not/ use crypto. The "AI" people want to hamfistedly shove their flawed investment into every product under the sun.

I think the worst of the crypto people are probably the worst of the AI people too. Power/money-hungry grifters naturally move on to the most profitable grift when the old one peters out.

It's never/rarely the leaders. It's always the followers. This site has a high bar for engagement given the depth of subject matter, but if you check another social medium I promise you'll see this pattern. All over.

> AI allows companies to resell open source code as if they wrote it themselves doing an end run around all license terms. This is a major problem.

Has it been adjudicated that AI use actually allows that? That's definitely what the AI bros want (and will loudly assert), but that doesn't mean it's true.

We're just seeing that the copyright emperor has no clothes: companies want to infringe upon others' rights through a copyright laundering machine, but of course will insist that the laundered code they use is their property, protected by the holy copyright cudgel.

I don't think so. Because LLMs aren't legal persons (yet?!), they can neither have copyright to anything nor violate someone else's copyright. IANAL but the most reasonable legal interpretation is likely that any IP violations are actually committed by whoever it was who asked an LLM to "rewrite" something in a way that obviously counts as a derived work rather than a cleanroom implementation.

Correct.

You are misinterpreting my use of the word 'allow'. Think of it as 'enables' or 'makes it possible'. It is fairly obvious that AI does not grant permission as well as that there was no reference to any legal proceedings.

I worked on an extremely niche project revolving around an old DOS game. Code I worked on is often pretty much the only reference for some things.

It's trivially easy to get claude to scrape that and regurgitate it under any requested licence (some variable names changes, but exactly the same structure - though it got one of the lookup tables wrong, which is one of the few things you could argue aren't copyrighted there).

It'll even cheerfully tell you it's fetching the repository while "thinking". And it's clearly already in the training data - you can get it to detail specifics even disallowing that.

If I referenced copywritten code we didn't have the license for (as is the case for copyleft licenses if you don't follow the restrictions) while employed as a software engineer I'd be fired pretty quick from any corporation. And rightfully so.

People seem to have a strange idea with AI that "copyleft" code is free game to unilaterally re-license. Try doing that with leaked Microsoft code - you're breaking copyright just as much there, but a lot of people seem to perceive it very differently - and not just because of risk of enforcement but in moralizing about it too.

The overwhelming majority of devs do not concern themselves with nor are even familiar with the concept of software licenses, let alone how to abide by them. I argue that it's not that they think it's "free game to [...] re-license" so much so as they think it's just code and they can use it without the idea of a licence ever even crossing their mind.

Source: find literally anything on GitHub using dependencies that are MIT licensed and being distributed without following the terms that state you must also redistribute the licence for each

It goes much further than that. They use LLMs to create code for them and then they claim ownership to the code.

I think that is one of the main reasons there is so much pushback against this, a lot of people are now addicted to their stream of washed code and want to claim ownership over what is essentially a derived work. The key then becomes 'if a work could not have been written by the author that claims it does that claim survive'. I think it should not but there is plenty of disagreement on this.

Isn’t similar to looking up/copying code from stack overflow, Google or books? Use it as reference to write the code and claim ownership. My little understanding is that the whole copyright free ride for LLMS is because it is similar to the process of humans using content under copyright as reference to create something new and claim ownership.

No, it is not similar at all. There is the pesky little thing called 'copyright' which allows you to set terms under which you license your code. Stackoverflow, google and books all come with strings attached.

Your idea of how humans use content under copyright is mistaken.

Why it’s not similar at all?

I want to make an ajax request using jQuery. I look up an example in StackOverflow. I use a very similar code to the example given in the post and by not giving any attribution I just claim ownership.

Same with Spring in action books or looking up Java class references. Many times I look something up and use it as reference just tweaking the examples given.

Millions of programmers have done this.

LLMS in principle use the training data to generate an answer to the prompt, similar to the process I described.