3 ms·

I wrote something original in a language learning grammar online. I coined a term to describe something about how a certain language with negative language func

I wrote something original in a language learning grammar online. I coined a term to describe something about how a certain language with negative language functions. I tried talking to an LLM about it and sure and enough, it had scraped it and learned to use that terminology and explanation that I made up. I asked the LLM, "Where did that concept/term come from, who made it up?" It did a bunch of searching and tried to site a whole bunch of other sources which did NOT contain the terms of the explanation I had written. It simply would not cite or mention my source. It kept parroting my material semi-correctly and hiding the source. By any other actor that would be an egregious act of sloppiness, dishonesty, and plagiarism. Generative AI is speed-running a widespread corruption of truth. I do not believe any of the gains are worth this.

skygazer 14d ago
Your text may well have been in the training corpus, but not searchable with whatever terms and search engine that LLM used after you prompted it. It doesn’t have recall of sources of documents comprising its training corpus, unless the source is widely cited in other of its training documents. Sourcing and provenance aren’t currently an intentional part of LLM training. That may already have been obvious, and central to the complaint, but I was just knee-jerking to the anthropomorphic language.
- layer8 14d ago
  I would assume the parent understands that. The criticism is about the LLM behavior this results in. Being able to explain a behavior doesn’t necessarily excuse it. By some moral standards, you wouldn’t have trained an LLM that way, or wouldn’t have made it available, given the predictable outcome.
  - 59percentmore 14d ago
    It also ostensibly replaced a search system that would likely have been capable of finding the original document. It's damning that the new one can't.
- moralestapia 14d ago
  >Sourcing and provenance aren’t currently an intentional part of LLM training. Congrats, brother. That's the issue TFA and GP are bringing up. And I do agree it is unfair and equivalent to (intellectual) theft.
mistrial9 14d ago
> I do not believe any of the gains are worth this. You are authoring original work.. the Baron in the Castle collected your work and uses it. Who becomes wealthy and how?
- MentalM 14d ago
  That is pretty weak logic. I mean obviously it is the Baron who becomes wealthy. But what follows from this? By the same logic if Baron bought her baroness masterful dress to enjoy her look in that dress, and you when walked past the Castle saw the baroness in that dress, should you compensate to the Baron or not?
- adamddev1 14d ago
  It's more than that though. I authored the original work. Then the Baron in the Castle collected the work and uses it without fully understanding it. The Baron gives it out but in a partly-true, distorted form. He misunderstand how it applies and teaches people false clearly facts with it, with an air of authority. But when people try to find the source argument to see if the Baron is really on track, they have no idea. The Baron is just confidently parroting a compelling sounding set of words. For me the issue is more about the loss of truth than the loss of my IP.
theendisney 14d ago
Well done Confucius, you've invented a new creditpocalypse. Its when llms hallucinate new authors for new words. It almost makes it worth starting a blogsurection!
- 6510 13d ago
  A day ago blogsurection had zero Google search results, today it has 4 like: "The Rise and Fall of the Blog" and the search llm sniplet explains what the word means.
ttoinou 14d ago
France, 1658 <-> World, 2026. AI can also do the reverse : for example very few in the world know that the term midwit was originally coined in french by Blaise Pascal (« demi habiles ») centuries ago. Yes, that bell curve meme, he wrote about it. It’s not explained anywhere because of the language and culture separation between french and english. But if you ask enough questions to AI, the AI will eventually find about Blaise Pascal and tell you about this unknown link. It’s magic.
- bitwize 14d ago
  There used to be a meme on Slashdot: "In Soviet Russia, X Ys you!" It came from a joke by Russo-American comedian Yakov Smirnoff: "What a country! In LA, everybody is looking for a party; in Russia, the Party comes looking for you!" But I long wondered if Smirnoff was influenced by his countryman Leon Trotsky who observed: "You may not be interested in the Dialectic, but the Dialectic is interested in you."
baranul 14d ago
Your comment illustrates that AI companies are playing a kind of shell game. On the one hand there is theft of IP and violations of copyright, in the middle there is the obscuring of where things come from, and on the other hand there is the selling of it via tokens. More countries should debatably follow what Japan is doing[1], where they are creating new laws to force AI companies to show their training data and how they collected it. [1]: https://www.japantimes.co.jp/news/2026/08/19/japan/ai-training-data-disclosure https://www.japantimes.co.jp/news/2026/08/19/japan/ai-traini...

skygazer 14d ago

Your text may well have been in the training corpus, but not searchable with whatever terms and search engine that LLM used after you prompted it. It doesn’t have recall of sources of documents comprising its training corpus, unless the source is widely cited in other of its training documents. Sourcing and provenance aren’t currently an intentional part of LLM training. That may already have been obvious, and central to the complaint, but I was just knee-jerking to the anthropomorphic language.

layer8 14d ago
I would assume the parent understands that. The criticism is about the LLM behavior this results in. Being able to explain a behavior doesn’t necessarily excuse it. By some moral standards, you wouldn’t have trained an LLM that way, or wouldn’t have made it available, given the predictable outcome.
- 59percentmore 14d ago
  It also ostensibly replaced a search system that would likely have been capable of finding the original document. It's damning that the new one can't.
moralestapia 14d ago
>Sourcing and provenance aren’t currently an intentional part of LLM training. Congrats, brother. That's the issue TFA and GP are bringing up. And I do agree it is unfair and equivalent to (intellectual) theft.

mistrial9 14d ago

> I do not believe any of the gains are worth this. You are authoring original work.. the Baron in the Castle collected your work and uses it. Who becomes wealthy and how?

MentalM 14d ago
That is pretty weak logic. I mean obviously it is the Baron who becomes wealthy. But what follows from this? By the same logic if Baron bought her baroness masterful dress to enjoy her look in that dress, and you when walked past the Castle saw the baroness in that dress, should you compensate to the Baron or not?
adamddev1 14d ago
It's more than that though. I authored the original work. Then the Baron in the Castle collected the work and uses it without fully understanding it. The Baron gives it out but in a partly-true, distorted form. He misunderstand how it applies and teaches people false clearly facts with it, with an air of authority. But when people try to find the source argument to see if the Baron is really on track, they have no idea. The Baron is just confidently parroting a compelling sounding set of words. For me the issue is more about the loss of truth than the loss of my IP.

theendisney 14d ago

Well done Confucius, you've invented a new creditpocalypse. Its when llms hallucinate new authors for new words. It almost makes it worth starting a blogsurection!

6510 13d ago
A day ago blogsurection had zero Google search results, today it has 4 like: "The Rise and Fall of the Blog" and the search llm sniplet explains what the word means.

ttoinou 14d ago

France, 1658 <-> World, 2026. AI can also do the reverse : for example very few in the world know that the term midwit was originally coined in french by Blaise Pascal (« demi habiles ») centuries ago. Yes, that bell curve meme, he wrote about it. It’s not explained anywhere because of the language and culture separation between french and english. But if you ask enough questions to AI, the AI will eventually find about Blaise Pascal and tell you about this unknown link. It’s magic.

bitwize 14d ago
There used to be a meme on Slashdot: "In Soviet Russia, X Ys you!" It came from a joke by Russo-American comedian Yakov Smirnoff: "What a country! In LA, everybody is looking for a party; in Russia, the Party comes looking for you!" But I long wondered if Smirnoff was influenced by his countryman Leon Trotsky who observed: "You may not be interested in the Dialectic, but the Dialectic is interested in you."

baranul 14d ago

Your comment illustrates that AI companies are playing a kind of shell game. On the one hand there is theft of IP and violations of copyright, in the middle there is the obscuring of where things come from, and on the other hand there is the selling of it via tokens. More countries should debatably follow what Japan is doing[1], where they are creating new laws to force AI companies to show their training data and how they collected it. [1]: https://www.japantimes.co.jp/news/2026/08/19/japan/ai-training-data-disclosure https://www.japantimes.co.jp/news/2026/08/19/japan/ai-traini...