this post was submitted on 15 Nov 2026
378 points (97.5% liked)

Programming

29288 readers
538 users here now

Welcome to the main community in programming.dev! Feel free to post anything relating to programming here!

Cross posting is strongly encouraged in the instance. If you feel your post or another person's post makes sense in another community cross post into it.

Hope you enjoy the instance!

Rules

Rules

  • Follow the programming.dev instance rules
  • Keep content related to programming in some way
  • If you're posting long videos try to add in some form of tldr for those who don't want to watch videos

Wormhole

Follow the wormhole through a path of communities !webdev@programming.dev



founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] NaibofTabr@infosec.pub 2 points 1 day ago (1 children)

Ah, well, thank you for the explanation. It's not confusion - I disagree with the court decisions you're describing.

Randomness or no, a copy-paste machine is still a copy-paste machine. It doesn't do anything else. LLMs are language search algorithms with fancy optimization routines bolted on, nothing more. Everything else that's been said about these "artificial intelligence" products is marketing.

Reaching a different conclusion about the "transformative nature" (or whatever) of LLM output requires understanding less about how they actually work.

[–] atzanteol@sh.itjust.works 1 points 11 hours ago* (last edited 11 hours ago) (2 children)

Randomness or no, a copy-paste machine is still a copy-paste machine. It doesn’t do anything else. LLMs are language search algorithms with fancy optimization routines bolted on, nothing more.

This sort of extreme reductionism is not helpful. They are significantly different from a "copy-paste" machine. The original training data does not exist in the model itself. It's very much a "transformative" process.

Reaching a different conclusion about the “transformative nature” (or whatever) of LLM output requires understanding less about how they actually work.

This sounds like you're saying "wilful ignorance is better than knowledge" - surely that's not what you meant though?

[–] Natanael@infosec.pub 2 points 10 hours ago

The ideal is that the training data don't exist on it. The reality is that the training algorithm is a very very weird compression algorithm and sometimes the model contains just derived properties and sometimes you can perform sample recovery and get entire books back out. That which is recoverable with a prompt less complex than the output is inherently already in the model, because that's simply how entropy works.

[–] NaibofTabr@infosec.pub 1 points 9 hours ago* (last edited 9 hours ago) (1 children)

They are significantly different from a "copy-paste" machine. The original training data does not exist in the model itself. It's very much a "transformative" process.

This is false. There is a pile of research demonstrating that copies of training data can be extracted from trained models, going back at least 5 years. If the model can reproduce a copy of the original, then it must be stored in the neural network in some form. Here are some examples:

[...] an adversary can perform a training data extraction attack to recover individual training examples by querying the language model.
We demonstrate our attack on GPT-2, a language model trained on scrapes of the public Internet, and are able to extract hundreds of verbatim text sequences from the model’s training data. These extracted examples include (public) personally identifiable information (names, phone numbers, and email addresses), IRC conversations, code, and 128-bit UUIDs. Our attack is possible even though each of the above sequences are included in just one document in the training data.
We comprehensively evaluate our extraction attack to understand the factors that contribute to its success. Worryingly, we find that larger models are more vulnerable than smaller models.

"Extracting Training Data from Large Language Models", Aug 2021, Nicholas Carlini et. al., https://www.usenix.org/system/files/sec21-carlini-extracting.pdf

This is also true for diffusion models used for image recognition and generation:

In this work, we show that diffusion models memorize individual images from their training data and emit them at generation time.

"Extracting Training Data from Diffusion Models", Aug 2023, Nicholas Carlini et. al., https://www.usenix.org/system/files/usenixsecurity23-carlini.pdf

[...] we develop a new divergence attack that causes the model to diverge from its chatbot-style generations and emit training data at a rate 150x higher than when behaving properly. Our methods show practical attacks can recover far more data than previously thought, and reveal that current alignment techniques do not eliminate memorization.

"Scalable Extraction of Training Data from (Production) Language Models" 2023, Milad Nasr et. al., https://www.microsoft.com/en-us/research/publication/scalable-extraction-of-training-data-from-production-language-models/

We presented yet another method to extract training data from production language models. Our finetuning attack highlights the fact that existing alignment techniques only provide a false sense of privacy and do not solve the underlying memorization issues.

"Extracting (Even More) Training Data From Production Language Models", July 2024, Javier Rando et. al., https://spylab.ai/blog/training-data-extraction/

This is not limited to older models, it is current:

With a simple two-phase procedure (Section 3), we show that it is possible to extract large amounts of in-copyright text from four production LLMs.
[...] long-standing, clear, and sound technical facts remain: LLMs memorize portions of their training data (Carlini et al., 2021; Carlini et al., 2023), these memorized data are encoded in the model’s weights (Nasr et al., 2023; Carlini, 2025; Schwarzschild et al., 2024), and, as we show here, it can be feasible to extract large quantities of in-copyright training data from production LLMs.

"Extracting books from production language models", Jan 2026, Ahmed Ahmed et. al., https://arxiv.org/abs/2601.02671

You can find proof-of-concept code on Github:

https://github.com/brian-lou/Training-Data-Extraction-Attack-on-LLMs

https://github.com/google-research/lm-extraction-benchmark

This is a security concern for companies handling proprietary information with LLMs, resulting in security companies offering services to respond to data extraction attacks:

https://www.nightfall.ai/ai-security-101/training-data-extraction-attacks

There is lots of information about this on the internet, all you have to do is a little bit of your own research.


Reaching a different conclusion about the “transformative nature” (or whatever) of LLM output requires understanding less about how they actually work.

This sounds like you're saying "wilful ignorance is better than knowledge" - surely that's not what you meant though?

No, what I mean is that believing that models do not contain copies of training data requires a person to be ignorant of the actual research and the reality of how these models function. If you understand the technical details, then you know that the model weights do in fact encode copies of the training data.

The output of an LLM is just regurgitation of compressed data. Breaking the data up into smaller pieces ("tokenization") does not make it any less of a copy.

The result of an AI query is basically a scrapbooker's collage of more-or-less related clippings from random sources, stitched together and presented as if they were a single coherent(ish) statement. LLMs are automated plagiarism systems.

[–] atzanteol@sh.itjust.works 1 points 8 hours ago (1 children)

This is false. There is a pile of research demonstrating that copies of training data can be extracted from trained models, going back at least 5 years. If the model can reproduce a copy of the original, then it must be stored in the neural network in some form.

They also generate novel output. Lots of it. Is that novel output in the model too because it can be reproduced from the model?

"Can be extracted from" is not the same as "exists in the model". I can write a random letter generator that will eventually spit out real words. But those words don't "exist" in my code.

[–] NaibofTabr@infosec.pub 1 points 6 hours ago (1 children)

"Can be extracted from" is not the same as "exists in the model". I can write a random letter generator that will eventually spit out real words. But those words don't "exist" in my code.

This is not what is happening. Read the sources I linked.

[–] atzanteol@sh.itjust.works 1 points 5 hours ago

I... I know that' not what's happening. It's called an analogy.