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

MIT Non-AI License

news.ycombinator.com

Post reply on HN
Print view View original

31–40 of 45 posts

Re: MIT Non-AI License

#31

Post by dumindunuwan »

linkdd wrote: post #29

> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software Key words are: - permission is [...] granted - free of charge - without restriction - use…

I understand these points. As someone who truly love open source, we can see open source projects are becoming just a free training materials for AI. After training LLMs using open-source projects AI can build far superior software one day and that software may be not free, not able to replace by any non-AI software project. We all know that day is not far and that period of time all open-source software might consider legacy as no individual contributor able to implement stuff the speed of AI. What you are protecting is not only a legacy system we build decade old requirements and also the death of the purpose of why people build free software.

What we have to focus is why we created free software, not word by word terms that not fulfill the requirement at this and future time period.

Re: MIT Non-AI License

#32

Post by tylerchilds »

nomel wrote: post #30
tylerchilds wrote: post #28

Earlier quoted context omitted.

Hilarious to get downvoted for defending the concept of Literature Without references and citations—- never mind. Go read Animal Farm

you comment did no such thing, in a way that would be recognizable by anyone else. it probably would have been a good comment if you did expand on that though, in some meaningful way.

Valid point. I shouldn’t have met condescension with condescension.

I don’t think I would have said anything if the opening statement wasn’t a “you do you, but”

Thanks for taking the time to call me out politely.

To be objective, we can ask the creator of tailwind how much honor he feels with his business just being weights in a model in someone else’s business

https://github.com/tailwindlabs/tailwindcss.com/pull/2388#is...

https://www.businessinsider.com/tailwind-engineer-layoffs-ai...

https://dev.to/kniraj/tailwind-css-lays-off-75-of-engineerin...

https://news.ycombinator.com/item?id=46527950

Going full hypertext computer history, this is why Ted Nelson believed in micro transactions on transclusions.

In modern terms— when ai bills by the tokens, those tokens should also get paid out to the source materials.

The business model is primarily broken, which is why that’s not happening, and why the main business use case is militarization of it.

Re: MIT Non-AI License

#33

Post by tylerchilds »

tylerchilds wrote: post #32
nomel wrote: post #30

Earlier quoted context omitted.

you comment did no such thing, in a way that would be recognizable by anyone else. it probably would have been a good comment if you did expand on that though, in some meaningful way.

Valid point. I shouldn’t have met condescension with condescension. I don’t think I would have said anything if the opening statement wasn’t a “you do you, but” Thanks for taking the time to call me out politely. To be objective, we can ask the creator of tailwind how much honor he feels with his business just being weights in a model in someone else’s business https://github.com/tailwindlabs/tailwindcss.com/pull/238…

I use ai daily but reciting the talking points that killed what open source used to mean and using them to further separate original authors from the impact of their work

That’s ai pilled

I write code under the mit license

I know the risk

Helping humans still makes it worth it

And technically these AI companies should have a /licenses route that lists every MIT piece of code their model was trained on.

That’s literally the only expectation I have from anyone as an active author using the MIT license, getting cited.

I think the legal AI defense is that the models themselves are a bastardized form of dynamic linking. I say the models are statically linked though, so they need to spill their sources.

Re: MIT Non-AI License

#34

Post by tylerchilds »

tylerchilds wrote: post #33
tylerchilds wrote: post #32

Earlier quoted context omitted.

Valid point. I shouldn’t have met condescension with condescension. I don’t think I would have said anything if the opening statement wasn’t a “you do you, but” Thanks for taking the time to call me out politely. To be objective, we can ask the creator of tailwind how much honor he feels with his business just being weights in a model in someone else’s business https://github.com/tailwindlabs/tailwindcss.com/pull/238…

I use ai daily but reciting the talking points that killed what open source used to mean and using them to further separate original authors from the impact of their work That’s ai pilled I write code under the mit license I know the risk Helping humans still makes it worth it And technically these AI companies should have a /licenses route that lists every MIT piece of code their model was trained on. That’s literal…

That’d be my question to the person I disrespected:

Why should I, as someone that’s not hypothetically giving back in code, continue to do so, when the social contract has been broken, where the always minimal expectation has been: Say my name?

Re: MIT Non-AI License

#36

Post by linkdd »

dumindunuwan wrote: post #31
linkdd wrote: post #29

> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software Key words are: - permission is [...] granted - free of charge - without restriction - use…

I understand these points. As someone who truly love open source, we can see open source projects are becoming just a free training materials for AI. After training LLMs using open-source projects AI can build far superior software one day and that software may be not free, not able to replace by any non-AI software project. We all know that day is not far and that period of time all open-source software might consid…

You can't say you love opensource and be mad that users are using the freedom you granted.

OpenSource projects are not becoming free training material for AI, AI companies are using a freedom OpenSource projects granted.

The claim that AI can build far superior software is dubious and I don't believe it one second. And even if it were true, that does not change anything.

With or without AI, permissive licenses (MIT, BSD, ISC, ...) always allowed the code to be used and redistributed in non opensource software. If you don't want that, use the GPL or a derive. If you don't believe that the GPL would be enforceable on the derivative works produced by AI, don't release your code as opensource.

OpenSource is essentially an ideology, that software should be free of use, and transparent, and freely shareable, without restriction. If you don't buy into that ideology, it's fine, but don't claim to love OpenSource when you don't. Just like a person who eats fish should not claim to be vegan.

AI will not be the end of OpenSource, firstly because it's a dead-end technology, it has already peaked years ago and is becoming worse with each new model. It does not have the ability to build complex software beyond a CRUD app (would you use a kernel that was entirely vibecoded? would you trust it the way you trust the Linux kernel?). Secondly, because OpenSource does not discriminate who gets to enjoy the freedom you granted.

You decided to "work for free" when you decided to distribute as OpenSource. If you don't want to work for free, maybe OpenSource is not for you.

Re: MIT Non-AI License

#37

Post by Gathering6678 »

dumindunuwan wrote: post #21
Gathering6678 wrote: post #19

How do you propose someone who uses this license to enforce such clauses?

Article 4(3) of the Digital Single Market (DSM) Directive establishes a legal bridge between AI governance and copyright law. https://www.ddg.fr/actualite/the-european-unions-code-of-pra...

I meant the enforceability of such clause: to the extent of my limited understanding of law, you would need to at least appear to prove that: someone has breached the agreement by, for example, using your code to train AI. I am not sure how it is possible.

Re: MIT Non-AI License

#38

Post by xign »

dumindunuwan wrote: post #31
linkdd wrote: post #29

> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software Key words are: - permission is [...] granted - free of charge - without restriction - use…

I understand these points. As someone who truly love open source, we can see open source projects are becoming just a free training materials for AI. After training LLMs using open-source projects AI can build far superior software one day and that software may be not free, not able to replace by any non-AI software project. We all know that day is not far and that period of time all open-source software might consid…

The whole point of open source license is that they are a legal document that can be enforced and have legal meaning. It's not just a feel-good article. Your argument is like saying to a client who you are drafting a contract to and say "oh yeah don't worry about the word by word terms in the contract, wink".

Also, this "non-AI" license is plainly not open source nor is it permissive. You can't really say you are a fan of open source when you use a license like this. The whole pt of the MIT license is that you just take it with no strings attached. You can use the software for good or for evil. It's not the license's job to decide.

There is nothing wrong with not liking open source, btw. The largest tech companies in the world all have their most critical software behind closed doors. I just really dislike it when people engage in double-speak and go on this open source clout chasing. This is also why all these hipsters startups (MongoDB, Redis, etc) all ended up enshittifying their open source products IMO, because culturally we are all trying to chase this "we ♥ open source" meme without thinking whether it makes sense.

If people say they "truly love open source", they should mean it.

Re: MIT Non-AI License

#39

Post by nomel »

tylerchilds wrote: post #34
tylerchilds wrote: post #33

Earlier quoted context omitted.

I use ai daily but reciting the talking points that killed what open source used to mean and using them to further separate original authors from the impact of their work That’s ai pilled I write code under the mit license I know the risk Helping humans still makes it worth it And technically these AI companies should have a /licenses route that lists every MIT piece of code their model was trained on. That’s literal…

That’d be my question to the person I disrespected: Why should I, as someone that’s not hypothetically giving back in code, continue to do so, when the social contract has been broken, where the always minimal expectation has been: Say my name?

> minimal expectation has been: Say my name?

MIT license specifically does not require public attribution for derivative works. You should be using a different license if that is your goal.

Re: MIT Non-AI License

#40

Post by tylerchilds »

nomel wrote: post #39
tylerchilds wrote: post #34

Earlier quoted context omitted.

That’d be my question to the person I disrespected: Why should I, as someone that’s not hypothetically giving back in code, continue to do so, when the social contract has been broken, where the always minimal expectation has been: Say my name?

> minimal expectation has been: Say my name? MIT license specifically does not require public attribution for derivative works. You should be using a different license if that is your goal.

I don’t want to insult you by pasting the full text here but this is the required bit

“The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.”

Which is here in my operating system

https://github.com/tylerchilds/plan98/blob/d27d975bf931a7d80...

I literally came up with a unique sdk for all “my elves” such that I can in fact see people in court for mishandling the software supply chain.

There’s a lot of software licensing misinformation out there and including my name and email with the rest of the license text is such a simple thing to misunderstand.

I’m sorry for any and all infractions you’ve committed across all MIT authors to date.

I’m not really planning to take anyone to court, but if you really believe what you’ve said to me here and you’ve been writing code that follows those beliefs

I’m not a lawyer, but you should probably consult a lawyer.

Post reply on HN