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

Ask HN: Is open sourcing your code still an option with the rise of AI?

news.ycombinator.com

Post reply on HN
Print view View original

21–30 of 38 posts

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#21

Post by josephcsible »

muzani wrote: post #20
josephcsible wrote: post #9

Earlier quoted context omitted.

No such licensing concept could work. The AI companies' argument is that what they're doing counts as fair use, so they don't need to abide by any license. If that argument is true, then such a thing wouldn't do anything, and if it's false, then such a thing is unnecessary, because they're already violating even the most lax licenses (e.g., MIT) by failing to provide the required attribution.

Can't one just literally say that AI is not allowed? It's not fair use when you're going directly against the license.

> It's not fair use when you're going directly against the license.

Isn't the whole point of fair use that it's stuff you can do even if the copyright holder isn't okay with it?

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#22

Post by shrimp_emoji »

b20000 wrote: post #10
ericxanderson wrote: post #6

Earlier quoted context omitted.

Also, depending how the legalese shakes out, you could still be compensated if they stole your code.

and how are you going to be compensated when an AI model has been trained on it, and your code is spit out? and who will pay for the litigation?

For the legally paranoid, this is already a "fruit from the poisoned tree" situation.

Some people are informally "banned" from working on whole categories of open source projects because they've had provable exposure to closed source code in the same domain.

That's a partial motivator in maintaining pseudonymity online. If no one knows who you are, and they don't know you've done kernel work at Microsoft, you or the open source kernel project you're contributing to can't be sued by Microsoft, hypothetically, for "inspiration".

There's legal questions here for which there's never been precedent, so nobody knows where the line is -- and this is all before LLMs ever entered the picture. For other countries, whose courts don't rely on precedent, it's even more of a minefield since the legal outcome of a case is always undefined behavior.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#24

Post by palata »

codingdave wrote: post #19
b20000 wrote: post #14

Earlier quoted context omitted.

I wonder what this means for SEO. If someone develops content and someone else then generates very similar content using AI, how do the search engines know who to rank higher?

Everything evolves. The internet existed before the web, and we navigated its content. The web existed before search engines did, and we navigated its content. Content and ways to navigate it will continue to exist once search engines are no longer relevant. Maybe SEO simply doesn't matter?

SEO will become LLMO, with the added bonus that if you do it right, you can get the LLM to lie in your favour.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#25

Post by palata »

8organicbits wrote: post #5

You'll need to clarify what your specific goals and concerns are. Most of my code is open source because I want people to build derivatives. There's lots of license variations to consider.

The question is specifically about the problem with licenses, if I understand correctly. AI is now a big license-laundering machine, so there are not "lots of license variations to consider": AI just removes the license, whatever it is.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#26

Post by palata »

b20000 wrote: post #11
tjr wrote: post #8

A lot of companies avoid using GPL code at all, because they don't want to even accidentally find their own proprietary code subject to being released under the GPL. I can imagine that some similar licensing concept could be applied to copyrightable works with respect to AI training. Use a legally restricted work for AI training, and your entire AI training set is subject to free public release?

i don't think anyone training AI models cares about any of that right now

If engineers cared about ethics, we would know it.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#27

Post by palata »

talldayo wrote: post #2

> with the rise of AI and inability to enforce licenses or control whether code is used for training. The inability to precisely enforce license abuse has always been a problem with Open-Sourcing your code. I don't understand how someone would see AI as a dealbreaker in the real-world of Open Source pragmatism.

Because of the scale? AI is orders of magnitude better at laundering open source licenses than anything that existed before.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#28

Post by palata »

jvalencia wrote: post #4

I think this is the wrong question. The question assumes that AI is a derivative to your code. The reality is that there will come a day, sooner than you like, when the AI can replicate the application you wrote without your code. In that world, what does it matter what code is open vs closed?

> The reality is that there will come a day, sooner than you like

Were you one of those saying that Tesla's FSD would come "soon", too?

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#29

Post by palata »

elpocko wrote: post #16

The license under which I have published my projects applies equally to both humans and machines. I see no reason to differentiate.

Except that those who use AI say that the license does not apply ("fair use" giving them the right to ignore the license).

It's not a question of humans vs machines: it's rather a question of whether those who train AI get to ignore copyright or not.

Re: Ask HN: Is open sourcing your code still an option with the rise of AI?

#30

Post by elpocko »

palata wrote: post #29
elpocko wrote: post #16

The license under which I have published my projects applies equally to both humans and machines. I see no reason to differentiate.

Except that those who use AI say that the license does not apply ("fair use" giving them the right to ignore the license). It's not a question of humans vs machines: it's rather a question of whether those who train AI get to ignore copyright or not.

OP asked a specific question though.

FWIW, learning from copyrighted data should not be a copyright violation. This applies equally to both humans and machines. I see no reason to differentiate.

Post reply on HN