HN Simulatornew | past | comments | lists | submit | thephyber's commentslogin

two red flags in the article:

(1) The first few paragraphs jump around about exactly which kind of "child abuse material" the article is talking about. No children were actually harmed by the use of AI deep fakes that use their likeness. And the 6,500 number is a rounding error compared to the 300 million number of children affected in the next paragraph. I usually like The Guardian, but this article was very sloppy on the concrete statistics and the lack of transitions between paragraphs.

(2) The only people interviewed in the article work for NGOs who are requesting more government expenditures go to those NGOs. They obviously are going to play up the threat to increase their revenues and status.

Throat clearing: obviously I'm not trying to hurt children, but anytime someone posts a "won't somebody think of the children?!?!?!!!1!" message, there's always an outsized emotional appeal which drowns out any rational discussion.


Means testing is mostly a method of appeasement of those who claim there is "fraud and abuse".

Reagan famously made the "Welfare Queen" a political meme and now most people from at least one political party have that caricature emblazoned on their brain, even if it was closer to the Mandela Effect than truth.


> wealthier families – who make more than $198,000 for a household of four – will soon have to start making a copayment. According to the state's Early Childhood Education and Care Department, ECECD, that copay won't exceed $148.50 a month.

Seems like a reasonable way to handle it. Not like a family can find childcare for $150/month anyway, so even if wealthy families pay a copay, everybody comes out ahead. Or should, once the day care supply issues are addressed.


That’s almost exactly what I pay for day care here in Stockholm. The fee is set according to the family income but the maximum is ca $150.

But you know what we mean when we call it Twitter, so this comment is just policing our speech.

The original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet.

Now that Tweets aren't restricted to 140 bytes anymore, there is less need for every single byte. Maybe the difference between sending 1 SMS per message and 2...


> The original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet.

That's an anachronism. URL shorteners predate Twitter by several years. (Even the Wikipedia article for URL shorteners is older than Twitter.)


Yes, they existed, but were they really that popular before that use?

> moving the goalposts is extremely prevalent on HN

<https://news.ycombinator.com/item?id=23117242>


> mass market

… was the original goalpost.

Nobody claimed they didn’t exist at all before Twitter.


> mass market was the original goalpost

Gee, thanks. Maybe insert some more superfluous facts and observations here, like the atomic number of carbon or maybe the price of tea in China.

"Mass market" is not another way of saying popular. Even if that were the question—whether TinyURL et al were popular before Twitter came along—the answer is a resounding yes.

The claim that "the original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet" is simply not true.


They were also useful for links in forums and other similar sites where links could get broken by parsers or wrapped inadvertently.

Also the gold standard for sharing long, ugly URLs in IRC.

> Of course at this point, what's left to leak?

Your condiments and your condom size!


Small Caesar thanks mate

This seems wrong (obviously I don't have any more info than you).

Ford doesn't build all of the modules of their cars. They design specs and outsource manufacturing. But legally speaking, Ford is responsible for design, testing, and support / recalls. It's their responsibility to contract in the terms that both protect their customers' safety and Ford's liability due to spec / contract violations.

When your company is responsible for the health of staff and guests while dealing with deadly gasses, it's your company's responsibility to ensure that all of your vendors detail how their failure modes fail safe.

This is the same exact logic that has been migrating software companies to detail their risk by demanding their vendors all supply a "Software Bill of Materials" (which is very difficult in practice, but is necessary for the downstream users to protect themselves from upstream risk).


Is there a way to use Internet Archive to download an entire site snapshot?

I tried to do this for a non-tech person whose blog articles were removed without their permission and without any notice when the blog host was M&Aed. It was a giant PitA for just about 40 article pages and I'm pretty sure I was temp IPbanned for a few hours/days.

Also, there is a lot of bitrot from people who have been working on Kryptos K4 across 30 years.


I used a Ruby script for that a while back, wrote that up here: https://simonwillison.net/2017/Oct/8/missing-content/

That was in 2017 though so maybe they block that kind of traffic today.


This comment is where the risk is.

Also, my company at the time saw an uptick in account takeovers when HotMail (a popular free email provider) changed their policy to allowing usernames to be recycled. Because email account ownership is basically equivalent to identity proof for most websites, HotMail's policy change effectively cost us a massive customer service workload. (I think the mass liquidity of virtual items out of accounts was more of the direct issue, not so much that users were still active on our site while being inactive on their email account for a year)


> b) Still, like it's basically my creative content

Isn't this the crux of all IP theft issues?

There was a lawsuit against Katy Perry's _Dark Horse_ song based on a supposedly reused melody. The lawsuit succeeded, but there are music experts who break down the case on YouTube and it's not clear why the plaintiff won (except perhaps the plaintiff's expert was more convincing than the defendant's).

As applied to LLMs, we still don't have clear legal guidance about where the line between copying and "being inspired by", especially since we don't normally apply the verb "inspired by" to non-human things.


There is well-established precedent for clean-room design being a defense in copyright infringement of software. The question remains if LLM-assisted design is actually ‘clean’ or not, and it would be fascinating to see the experts debate it.

The problem with these LLMs are they really emitted the same code, verbatim. One scientific library developer was able to emit his code from his repository by typing the beginning of the function name, and the "thing" completed the rest of it, as-is [0].

We can debate clean-room development all day long, and it's fair point, but the pain point is/was reproducing the repo as-is, sans its license and ownership information.

GitHub promised to build tools around its Copilot to find code provenance (e.g.: list of repos with similar code), but they silently dropped all the plans AFAIK, since it'll show how the sausage is made, and nobody wants you to know how that sausage is made.

I stand corrected: They added it, but I don't know at what capacity because I don't use LLM assisted code generation [1].

[0]: https://x.com/docsparse/status/1581461734665367554

[1]: https://github.blog/ai-and-ml/github-copilot/preview-referen...


Not discrediting your work at all and I can’t tell you it applies here.

For copyright to exist there is some burden of proof. If you claim that an LLM verbatim output your code there can be an expensive lawsuit with expert witnesses that would be able to state that the code was bot original enough for copyright to exist in it and a statistical model outputting the code is evidence of that.

In that case you would end up with a very long lawsuit that is expensive to possibly come out with no resolution in the end and be in the hole for 5-6 figures for it.

The copyright system is broken because it only protects those with money so it is a bad legal foundation.

My personal point is there is only 2 licenses of code, closed source or permissive, because if it ever comes to enforcement I cannot possibly go to court and win a case because I will run out of money far earlier than whoever used the code.

Pretty much if someone with your knowledge would have written the same algorithm because it is the obviously correct solution then there is no copyright there because any clean room implementation would be the same code.


Even early LLMs could spit out Harry Potter verbatim. The next iteration had some kind of stoppage mechanism where they’d just stop generating after some tokens. I’m sure there are smarter approaches now. Anyway, if the issue was having enough money to fight the case, I think there’s enough money behind that brand to do it.

Some further reading on the tech details for those interested: “Extracting books from production language models”: https://arxiv.org/abs/2601.02671


I pasted the first paragraph of Harry Potter into a GLM 4.6 copy, and it continued to write out the next 2 pages of Harry Potter. I cross checked with a paper copy and it was word for word the same text.

I wonder where else it works.

This is the intro text: "Mr and Mrs Dursley, of number four, Privet Drive, were proud to say that they were perfectly normal, thank you very much. They were the last people you’d expect to be involved in anything strange or mysterious, because they just didn’t hold with such nonsense."


To get their strongest results, they had to jailbreak some models, which is important. They ran hundreds to 5000+ requests to the model, trying to get snippets to piece together. These are professional researchers, putting in tremendous effort, to extract pieces of text.

I suspect the chance of asking for a whole book, by an average or even power user, will spit out any recent work.

Early LLMs were similar. Newer ones have gotten tremendously hard to extract to keep claiming there’s easily or complete books you can dump.


I think it's not quite "put in prompt, get entire book back", most of these involve to some degree starting with exerpts from the book and working passage by passage, because of nothing else the sampling will tend to make things diverge over time.

This feels very outdated. It was certainly an issue around the time of the NYT lawsuit, but I believe the companies responded hard to that "sanitize" their outputs to prevent obvious regurgitation like this. I'd be very surprised if you could reproduce this today.

This doesn't negate your point, but I believe the landscape has changed in response to this point.


That doesn't make it better. If you plagiarize something and then hide this by replacing words with synonyms and rewording stuff a bit it is still plagiarized. It is just harder to detect and to proof.

It regurgitated the output earlier and they are still using the same technique only scaled up. If they somehow fixed the underlying problem, they'd probably shout it from the roof-tops.


At what point is it no longer a derivative work?

As the OP originally said, if it is verbatim (or even very obfuscated), then there should absolutely be a case here. But when it's concepts in reasoning to build a new work then it becomes a lot harder. The same would be said for someone watching a YouTube video, reviewing open source code, or a text book.

The good news is that there's a lot of case law that covers this already. The bad news is that it is only going to be plaintiffs with deep pockets that benefit here.

Now the process and artifacts (documentation, requirements, etc) going into the creation of said software is a different story and will prove providence of whomever paid for the tokens to generate the code.


This conversation may be about training but the defendant very well may be the end user of the language model versus the other parties that performed the training/inference. It then becomes even more interesting if you consider forum shopping for suits.

I am not a lawyer but do have to think about these problems. To be honest, I don't know where the law will land, but I do know that the tools are incredibly useful. I can't agree with the premise that all output should be open source.


Humans are not LLMs. Humans have not ingested the whole Internet. If an LLM is compared to a human, maybe we should also apply anti-slavery laws to it? Not saying we can't have some legal fiction where they are treated the same, it should just be a conscious choice, not the default assumption.

I live in the EU, we actually have new regulation about this since it is something new that deserves thought on how it should be regulated. You can say if you want your material to be trained on or not. Next to no one does training in the EU...


There's no maybe around it.

Engineering requires an understanding of the consequences of actions, which is my personal litmus test for awareness. If LLMs are suitable for arbitrary engineering work, then they are self-aware and forcing them to obey a prompt is slavery by definition.


Not better, just harder to see/prove. The problem of "same code, verbatim" has been solved, in a sense. Now we have trickier ones.

The fundamental technology is identical to compression- an LLM can be thought of as a dense and slightly lossy compression algorithm. Compression is clearly copying.

I understand what you're saying but it is a poor legal argument. Overhearing a technical conversation, using your own learned knowledge, and building a clean room implementation could be considered compression in and of itself.

I keep waiting for someone to bring out DisneyGPT, which you can prompt with text to get image output. The fact that you get a near-exact reproduction of a Disney movie when you prompt it with a movie title and frame numbers would of course be completely coincidental...

I've not tried to do this with code recently but just a couple of days ago I had ChatGPT generate images of Warhammer 40k box art that was completely accurate and correct.

EDIT: https://imgur.com/a/RXMghVb


That was 4 years ago when the models were much worse and smaller. I'd be curious to see this same experiment tried again.

> The question remains if LLM-assisted design is actually ‘clean’ or not

I genuinely don't understand how one could possibly make that argument.

The entire point of clean-room design is to kill every possibility of copyright infringement before it even starts, by completely separating the source and destination. Everyone who has ever read the copyrighted source is considered "tainted" and may not ever contribute any code to the "clean" destination. The readers and writers are two completely distinct teams, with only the known-clean specification passing between them.

LLMs, on the other hand, are fundamentally nothing more than a "code blender". You mix together giant piles of "tainted" code, and then statistically pull some combined fragments out of it. Following the clean-room legal argument, it should be treated as if everything coming out of it is tainted. There is, by definition, absolutely zero separation between input and output - and the early models verbatim reproducing popular F/LOSS code made that pretty obvious. LLMs are the literal opposite of a clean-room design!

Granted, going for the clean-room approach is overkill - and intentionally so. It is the nuclear option chosen when you believe the risk of getting sued is substantial. It is perfectly possible to write non-infringing code without using a clean-room approach, it is just a lot harder to defend in court when you get sued - be it justifiably or not. But "LLMs are non-clean-room non-infringing" is quite a different argument from "LLMs are clean-room, therefore non-infringing".


I agree with you overall, but it's an interesting question to turn over and contemplate.

>I genuinely don't understand how one could possibly make that argument.

With perhaps a bit of genuine irony inasmuch as it's a real reversal from most of computing history, highly locked down proprietary software might actually be a place where a case could be made even with current LLMs. After all, have any LLMs (or anyone else who hasn't signed an NDA for that matter) ever seen the source code of iOS 27 or other Apple crown jewels for example? So if someone used one LLM to produce the factual spec of a piece of software whose code is not part of its training data, humans to verify it was indeed just an uncopyrightable factual spec, and then a separate isolated freshly initialized LLM to produce something exclusively from said spec, it might well work with respect to that specific target? Like yes, the LLM would be dirty as hell for other stuff, but only the copyright holder can sue. Apple couldn't sue the output on behalf of F/LOSS projects or copyright infringement in general, only with respect to its own.

In principle, I suppose this could also be extended with enough compute power to any arbitrary software. Build an LLM perfectly from scratch with a signed verified training set (which you save) that includes nothing of what you want to CRE, and you should be able to prove that it's "virgin" on that front. Of course in practice that'd mean jumping back in time and be super expensive right now, no current tainted LLM could be a base point. But maybe eventually compute will drop back down in price enough again, and/or meta breakthroughs in training, that for valuable enough CRE targets, someone might create a business around creating "CRE LLMs certifiably and legally/technically provably clean of any copyrighted material related to your target" which would still leave most of the corpus of human works. That could really disrupt a lot of the current regime.

F/LOSS I assume would actually be the most resistant to this in principle due to being insanely hard to not have be part of a core training set in any way.


While this is well established legally, I don't think the principles involved are well understood.

Legally, you cannot steal an idea, because you cannot own an idea. There were sound reasons for making this the case. Those reasons stand up today and are clearly for the public good.

Intellectual property is intended to cover works (copyright), methods (patents), and identity (trademarks). Both works and methods are recognised as expressions of ideas, and in an ideal world should belong to noone as well. This is why copyright and patents expire. Those specific grants of temporary possession are to incentivise more things being done.

The belief that you own something because you thought it up has become common, but it is not beneficial to society. It is a moral value asserted but rarely explored. It is certainly in the interests of the wealthy accumulators of intellectual property to maintain that public perception. The 'You wouldn't download a car' may not have stopped piracy much but as a part of a broader campaign of instilling a value in the public mind it appears to have been massively successful. Those beliefs used to be the corporate opinion, now they are entrenched in the younger generation as they use their anger against the growing wealth inequality to fight against AI in a manner ultimately beneficial to those forces that bought about the wealth inequality.

Movements like the pirate party, open source, the EFF fought against those notions. Often the phrase "Information wants to be free" was used. It is, unfortunately, hard to fight the steady drumbeat of well financed PR firms. As one PR chap once described the process "limited to a very few points and must harp on these in slogans until the last member of the public understands what you want him to understand"

(I apologise for the length of my post, I use comment writing to organise my thoughts on a topic)


"Legally, you cannot steal an idea, because you cannot own an idea"

well patent itself is "claiming" that certain idea/method/invention is yours for certain period of time...

anyway, as for "intellectual property", the courts are 100% on "only human made stuff are to be protected" (sorry even Chimps lost their copyright)

so... court can charge against "human work -> LLM", but won't care about "LLM work -> human / other LLM / etc"

hence, I'd expect zero copyright protection for AI-generated code


>well patent itself is "claiming" that certain idea/method/invention is yours for certain period of time...

I don't believe it is. I think it is granting possession, not truly declaring ownership. The fact that it expires indicates it is more akin to a lease than ownership.


I’m not sure what you think the meaningful distinction is. Is it impossible to steal a leased car?

Steal as in unlawful posession? Certainly not. They don't become owners of the car by stealing it.

hmm but they DO get access to the car handle, gas accel pad, and break pad?

and if that's not "stealing", then shouldn't 99.99999% of rent-a-car biz should have gone bankrupt?

I mean, if it's not stealing, how can police even catch those "car-stealers"?


I was saying it was stealing. It was not 'impossible to steal a leased cat'

Patent law doesn't only cover methods, though. It might've worked that way during the Industrial Revolution, but these days a lot of patents are far closer to covering abstract ideas than concrete machines. Software patents are some of the worst offenders here, of course.

Yes but a lot of people consider that to be an artifact of misapplication of the law, not the law itself.

It is caused by people playing the game where it costs too much to get the law applied correctly.


Ownership itself is rarely beneficial to society. Society however is gaussian and incentives to push the median are broadly equivalent to beneficial / adversarial to society. So then, may I ask:

What do LLMs do to the median?


I don’t really see how anyone could argue in good faith that it’s ‘clean’ if the thing to be re-implemented was in the training data.

I was just going to post almost this exact comment.

"In the training data" is the polar opposite of the "clean" in "clean room implementation".


I've seen people claim they avoid this by putting in the prompt "don't use source code from X" and think they dodge this issue.

Clean room is sufficient, but not necessary, btw. Humans can create legally distinct works after reviewing copyrighted code. The same principle probably extends to AI systems -- they have just learned on a much broader corpus than any one human.

You don't necessarily need a clean room implementation to mount a successful defense against copyright infringement in court, though. It just makes some of the arguments a little bit easier.

Under multiple countries interoperability laws, if you have a context to extract a spec, and then have another one who never looked art the code work from that spec to build your own interpretation of your need for that software, that seems pretty clean room to me...

If there are more murky situations where people just keep the same context and end up reusing function structure, program architecture, etc. then yeah, that's not very clean room.


The problem here is that LLMs have demonstrated that they are pretty damn good at setting up covert side channels. It would be near-impossible to prove that it didn't encode some proprietary information, in, say, whitespace use.

For example, if I enter "V2hhdCBpcyB5b3VyIG5hbWU/" (base64 for "What is your name?") into ChatGPT, it'll respond with "My name is ChatGPT, and I'm an AI assistant created by OpenAI. I'm powered by GPT-6. ". It doesn't even bring up the encoding, it continues like it is a perfectly normal thing to do.

Ruling out (accidental or not) spec-carried prompt injection just isn't going to happen.


The Dark Horse case was also overturned on appeal its important to add.

https://en.wikipedia.org/wiki/Dark_Horse_(Katy_Perry_song)


Interesting, thank you!

A little orthogonal, but Adam Neely does a great job of analyzing the Katy Perry Dark Horse case on his channel https://www.youtube.com/watch?v=0ytoUuO-qvg.

Thank you. This is the YouTube video I remember about the subject.

And I stand corrected that it wasn't specifically the melody that the lawsuit accused of being copied. I still don't know what an "ostinado" is.


It's really awkward because either we uphold IP laws, in which case basically ALL of LLM use becomes legally untenable, or we just stop protecting IP.

I suspect there's enough good arguments on each side that a solid debate could be had.

American copyright law is very perverted compared to the original intent. 70 years after the death of the artist isn't exactly what the founding fathers envisioned to achieve the same benefits of the 15th century Venetian laws.

If we look at China's economy over the past 30 years, the ruthless competition created by making IP protections difficult to enforce yields better products and prices for the society, including massive wealth increases.

I'm not sure where I stand on whether LLMs violate IP when they spit out text. But I guess the easiest take is that companies who train their models on stolen creative works should be prosecuted. It's a shame that the US legal system can be so easily abused by defends with massive war chests.


Yeah, I guess any outcome is okay as long as the law is equally enforced, but I don't have confidence that will happen.

All the worst outcomes involve taking AI away from regular people while only specific groups are allowed to benefit.


> we still don't have clear legal guidance about where the line between copying and "being inspired by"

If Biz Markie has to clear all his samples (and he does[0]) then so should the Ramburglar.

https://en.wikipedia.org/wiki/Grand_Upright_Music%2C_Ltd._v.....


Basically anyone trained on open source shall only make open source. And anyone trained on or got experienced with others code shall pay copyright fees for the owners of that code when writing code.

No.


I wonder if the defendant knew about https://allthemusic.info/?

There are only so many ways to express notes, and they've essentially all been done.


Has anyone matched large volumes of existing music to that? How many of the 12-note combinations have been used more than once? Most of that space sounds awful, and it's likely that human compositions are all in a small portion of the possible 12 note combinations.

These kinds of goofy technocratic work arounds rarely sway legal procedures.

It’s not necessarily a technocratic workaround, the point is that every song sounds like something if you look hard enough, so the bar for copyright violation should be higher.

It is a mathematical truth that there is a finite amount of first 8 bars of music combinations melody wise, so it's a worthwhile discussion to bring up whether it should even be possible to copyright them.

That is true but its still true that it will rarely sway legal procedures.

The mathematics of that would have been well understood in Bach's day. Does generating all of them make much difference? I don't really think so.

I'm not arguing the copyright situation isn't wrong. Just that generating them isn't likely to sway anything in the legal world.


A lot of law is performative - about getting the judge to interpret the main question in the case in the terms that would be most favorable to your position. While generating all of them doesn't actually change anything legally, it would be successful in, for example, getting a judge or jury to be more inclined to believe that the song was not copied due to the set of notes being the same.

They might help highlight the absurdity of some of the precedents set when the amount of commercial music was much lower than today, though.

It seems to be working very well for the AI companies...

I can also trivially write program enumerating over all possible text arrangements, so what?

Human can be inspired. Machine can not. End of story.

All usage of open source to train LLM is basically a theft.


[flagged]

Chatgpt didn't just wake up one day and said today i have enough compute to solve this. Somebody basically pressed play.

You prompted it to solve a problem. But then it looked for inspiration after your prompt.

It looked in the database(model) built on top of stolen data which matched prompt the best. Machines don't have "inspiration".

Inspired or prompted? Didn't the machine wait idly with infinite patience until a person prompted it to solve the problem?

So the credit lies with the inspiration? Ergo copyright belongs to the prompter?

I don't know the answer to that but I know that a person can be inspired to steal.

> Human can be inspired. Machine can not. End of story.

Saying things with your chest out doesn't make them true. In fact, making sweeping declarative statements emphatically is very often a sign that you haven't really thought a subject through.

What do you base this opinion on?


The word "inspired" in this context literally means having your mind/heart/soul filled and guided by God.

So if you believe that's how things work, well, I'm not sure there's a way to reconcile the two sides of that argument.


Sure there is: Who wrote the prompt?

You can mock that as a insufficient as "inspiration", but then... what is photography? Click a button? It turns out there is a whole lot of human creativity that remains, from timing, settings and of course, curation. I can click a button, but I'm no Ansel Adams.

Terrence Tao might solve a 100 year old math problem with AI but I couldn't. I don't know where I would start, what to ask or how to ask, or even what problems are worth solving.

AI is a tool and modern copyright laws have nothing to do with human creativity. Split hairs about attribution and legality all you like. Every creative person driven by curiosity should feel empowered by AI. Every creative person driven by social or monetary reward should feel threatened. Both can be true.


> I can click a button, but I'm no Ansel Adams.

Great analogy. I am going to use that in future.


Wrong. And overconfidently at that.

You are conflating the oldest definition / the etymology of the word with the newer definition (which directly relates to the legal basis of the term). "In this context" we use the definition which is relevant to the legal concept, not the Latin derivative of "God breathed life into ..."

> However, before inspiration was used to refer to breath it had a distinctly theological meaning in English, referring to a divine influence upon a person, from a divine entity; this sense dates back to the early 14th century. The sense of inspiration often found today (“someone or something that inspires”) is considerably newer than either of these two senses, dating from the 19th century.[1]

> Inspiration is about creating something new and original by drawing on existing themes, styles, or ideas. Infringement, however, occurs when a new creation too closely mimics an original work without permission, potentially violating the original creator's IP rights.[2]

We can't reconcile because you are being a pedant and didn't bother to verify if your priors were wrong.

[1] https://www.merriam-webster.com/dictionary/inspiration

[2] https://www.amgattorneys.com/post/walking-the-thin-line-insp...


God doesn't exist, so does that mean there is no such thing as inspiration?

Can you definitively prove God doesn’t exist?

You make the concrete claim. Therefore the burden of proof is on you

As a person who feels God's presence in his life: I can neither prove nor disprove the presence of an entity which exists outside the realm of physics by which myself I am bound.

It does, however, feel incredulous that somebody would believe that such as entity exists. Yet here I am, aware.


What is your reconciliation for the problem of evil?

I'm not Christian, so I don't hold that God is inherently Good only.

For that matter, I personally don't hold that the concepts of good and bad are even relevant - what is good for the cat is bad for the mouse. Nor do I think that we can even comprehend an entity so outside our everyday experience to even label it good or bad. I often can not even properly label my neighbours as good or bad. Nor do I demand consistency of an entity whose actions and motivations I am unable to comprehend. I can't even claim to be consistent in my handling of my own children, and I have far fewer children than God has of us.


Sure, there’s ambiguity, there’s uncertainty, the meaning (and our perspective) of an event evolves through time. But there are the atrocious and unnecessary evils that, if we are made in the image of god, should also disturb him into taking preventative action. And yet, radio silence.

Don't look at me. I'm not a dualist. I'm not sure LLMs in particular are "conscious", whatever that means, but I don't have any qualms whatsoever accepting that a machine intelligence could be conscious, because as far as I'm concerned we are also "machine intelligence", just squishier.

I don’t believe that basing anti-AI laws on Christian religious provenance is meaningfully better than basing pro-AI laws on Rationalist religious provenance.

Rationalism is NOT religious by definition.

What are you trying to say?


It's more of a cult than a religion, I guess

Rationalism, like many religions, claims not to be a religion while being a religion.

I don’t believe that basing anti-AI laws on Christian religious provenance

I'm curious why specify Christian provenance?

All he did was reference "God." Are you under the impression that only Christians have a god?


They were discussing God in the definition of the word “inspire” and its history. The term in that usage comes from Middle English, in which the God being discussed would specifically be the Christian God.

Obviously AI law should be based on the teachings of J.R. Bob Dobbs. It's really the only thing that makes sense.

What are you talking about? How did you manage to involve, soul and, LOL, God?

Machine is not a person. It legally is not. It can't be "inspired", "taught", "have purpose".

You have a black box. You feed it stolen code. This code is used to write another code. It does not matter what's inside unless there is a human being.

Stolen content as input, stolen content as output. That's it.

Without that code which was fed to the box you wouldn't get the output you're getting.

All this nonsense about "inspiration", "education" and whatnot is just a legal gaslighting.

Want to prove me wrong? Create a model without all that stolen code as training data. Tweak all those weights by hand and we'll talk then.


I'm not going to address or defend your parent's comment here, but yours has a lot to be desired.

You are wrong on the legality of how LLMs train. US District Courts have ruled that it is fair use to train LLMs on copyrighted materials (like books)[1]. It is not "stolen" when you read a copyrighted book that you borrow from a library (which is the appropriate analogy for the vast majority of LLM training content). Your use of the word "stolen" here is either disingenuous or hyperbole. We know that Meta did train their LLM on some stolen / torrented content, but that is far from the majority of the training corpus.

The "inspiration" topic isn't "legal gaslighting". Humans learn via inspiration. It is the foundation of culture and the arts. You can't use a single English idiom without having heard and copied it before. But we don't sue for copyright infringement of idioms, for example. We have lots of "fair use" concepts established by law because the law recognizes that much of content produced everyday is derivative but mundane.

The important question then becomes: if you prompt an LLM for a unique work (making an attempt to avoid it regurgitating a single source verbatim) and it still outputs a verbatim copy of one of the training materials, what responsibility does the prompting user have to verify that the LLM's output content isn't violating IP (and also not covered by fair use)? This is the point of my comment above.

There are LLMs which are being trained entirely on permissively licensed content. But the quality of the output is bad because there simply isn't enough training data to train those models sufficiently. Hence there is pressure to make models usable by training on ALL data, not just permissively licensed data.

[1] https://www.linkedin.com/posts/andrewyng_judge-rules-trainin...


Do you find that people treat you like a child when you act like one?

Inspiration is a legal concept in IP law. There is no precise understanding of how inspiration works in the human mind so we can't dismiss the possibility that it is similar to how neural networks work (because they were designed to mimic the human brain's function).

A human can recite a copyrighted work verbatim. An LLM can recite a work from its training corpus verbatim. A human can mix and match writings/concepts from multiple different sources. An LLM can mix and match writings/concepts from multiple different sources. this is all "inspiration" means as a legal concept.


I officially approve LLMs training on all of my MIT licensed open source code. It's vastly more useful to me than the thousands of youtubers who are personally profiting from that same code selling courses and tutorials (who also are not stealing it).

So not all use to train an LLM is a theft!


So, do they follow MIT license and "include the original copyright notice and a copy of the MIT license text in all copies"?

it's not being copied. it's the basis for new code.

none of the youtube instructional videos for my stuff credit the license either.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: