[tylerchilds]

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.

[tylerchilds]

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?

[nomel]

> 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.

[tylerchilds]

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.

[nomel]

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

Oops.

I don't think this applies to AI though [1].

IANAL, I am also not smart.

[1] https://www.nbcnews.com/tech/tech-news/federal-judge-rules-c...

[tylerchilds]

That’s about the inputs going in and even in the first paragraph disclaims other courts may side differently

We’re still not in the legal territory of the outputs on the other side, which is what’s actually more interesting to me.

The technical political maneuvering of to two sides of the accelerationist movement is the fair use bits, which is not even the point I’m debating.

Even there, they are discussing complete bodies of work as source material, which is exactly my point. My stuff can go in, but they still have to cite that I’m in there, which does matter on the other side.

I don’t think anyone deserves to be just weights and algorithms in a dark and shuddered library.

They want this to be a legal laundering device and that’s the bit I’m hung up on.

[tylerchilds]

I literally re-wrote how I wrote software post-ai to ai pill the ai, such that, when these models produce ASTs that match my signature, I do have a legal defense.

[tylerchilds]

Full circle:

We all deserve to be recognized beyond the false religion of the singularity.

In practice, that’s a bibliography. That’s what the MIT license guarantees— sources: cited, aka, literature in action.

[tylerchilds]

Plagiarism is the crux of what MIT protects

I do legally wave my right by using it to know every instance of it in the wild.

But even in those private instances, legally, that cannot be presented to a superior without the license without the copyright and the license text.

Just because the product is private doesn’t erase the name of the author.

There’s a saying “can you separate the art from the artist?”

And the punchline to sillyz.computer is: that’s the only thing banned.