• 0 Posts
  • 36 Comments
Joined 1 year ago
cake
Cake day: June 21st, 2023

help-circle
  • If something is possible, and this simply indeed is, someone is going to develop it regardless of how we feel about it, so it’s important for non-malicious actors to make people aware of the potential negative impacts so we can start to develop ways to handle them before actively malicious actors start deploying it.

    Critical businesses and governments need to know that identity verification via video and voice is much less trustworthy than it used to be, and so if you’re currently doing that, you need to mitigate these risks. There are tools, namely public-private key cryptography, that can be used to verify identity in a much tighter way, and we’re probably going to need to start implementing them in more places.








  • The key element here is that an LLM does not actually have access to its training data, and at least as of now, I’m skeptical that it’s technologically feasible to search through the entire training corpus, which is an absolutely enormous amount of data, for every query, in order to determine potential copyright violations, especially when you don’t know exactly which portions of the response you need to use in your search. Even then, that only catches verbatim (or near verbatim) violations, and plenty of copyright questions are a lot fuzzier.

    For instance, say you tell GPT to generate a fan fiction story involving a romance between Draco Malfoy and Harry Potter. This would unquestionably violate JK Rowling’s copyright on the characters if you published the output for commercial gain, but you might be okay if you just plop it on a fan fic site for free. You’re unquestionably okay if you never publish it at all and just keep it to yourself (well, a lawyer might still argue that this harms JK Rowling by damaging her profit if she were to publish a Malfoy-Harry romance, since people can just generate their own instead of buying hers, but that’s a messier question). But, it’s also possible that, in the process of generating this story, GPT might unwittingly directly copy chunks of renowned fan fiction masterpiece My Immortal. Should GPT allow this, or would the copyright-management AI strike it? Legally, it’s something of a murky question.

    For yet another angle, there is of course a whole host of public domain text out there. GPT probably knows the text of the Lord’s Prayer, for instance, and so even though that output would perfectly match some training material, it’s legally perfectly okay. So, a copyright police AI would need to know the copyright status of all its training material, which is not something you can super easily determine by just ingesting the broad internet.


  • AI haters are not applying the same standards to humans that they do to generative AI

    I don’t think it should go unquestioned that the same standards should apply. No human is able to look at billions of creative works and then create a million new works in an hour. There’s a meaningfully different level of scale here, and so it’s not necessarily obvious that the same standards should apply.

    If it’s spitting out sentences that are direct quotes from an article someone wrote before and doesn’t disclose the source then yeah that is an issue.

    A fundamental issue is that LLMs simply cannot do this. They can query a webpage, find a relevant chunk, and spit that back at you with a citation, but it is simply impossible for them to actually generate a response to a query, realize that they’ve generated a meaningful amount of copyrighted material, and disclose its source, because it literally does not know its source. This is not a fixable issue unless the fundamental approach to these models changes.


  • There is literally no resemblance between the training works and the model.

    This is way too strong a statement when some LLMs can spit out copyrighted works verbatim.

    https://www.404media.co/google-researchers-attack-convinces-chatgpt-to-reveal-its-training-data/

    A team of researchers primarily from Google’s DeepMind systematically convinced ChatGPT to reveal snippets of the data it was trained on using a new type of attack prompt which asked a production model of the chatbot to repeat specific words forever.

    Often, that “random content” is long passages of text scraped directly from the internet. I was able to find verbatim passages the researchers published from ChatGPT on the open internet: Notably, even the number of times it repeats the word “book” shows up in a Google Books search for a children’s book of math problems. Some of the specific content published by these researchers is scraped directly from CNN, Goodreads, WordPress blogs, on fandom wikis, and which contain verbatim passages from Terms of Service agreements, Stack Overflow source code, copyrighted legal disclaimers, Wikipedia pages, a casino wholesaling website, news blogs, and random internet comments.

    Beyond that, copyright law was designed under the circumstances where creative works are only ever produced by humans, with all the inherent limitations of time, scale, and ability that come with that. Those circumstances have now fundamentally changed, and while I won’t be so bold as to pretend to know what the ideal legal framework is going forward, I think it’s also a much bolder statement than people think to say that fair use as currently applied to humans should apply equally to AI and that this should be accepted without question.











  • Sure. My point is that, as far as I believe anyone is currently aware, there is no evidence that any law enforcement agency has ever accessed the content of encrypted WhatsApp messages. That does not mean that it has never happened either, but anyone positively claiming so is doing it without actual evidence, which is something we should probably avoid doing.