These models might be smart but they're not close to being able to savor irony.
(Now i don’t think you are necessarily in the UK. Just wanted to explain that Disney is not the only reason an AI might be trained to thread carefully around copyright issues of Peter Pan.)
Most of the books weren't available on lib gen or Anna's Archive. The few I did find were themselves obviously transcripts. Easy tell was they were missing distinctive formatting that I knew existed from reading the dead tree edition. At that point it was easier to make my own. I probably spent an hour searching for eBooks without DRM that weren't transcripts. Do they exist somewhere? Probably, but with a search of unknown length it was a better use of my time to make my own transcripts with what I had on hand.
I was really wanting to make commentary on how chaotic LLMs are even under constrained circumstances. No doubt both system prompts includes language about considering copyrights and trademarks. Probably pretty strong language at that. For whatever reason one LLM didn't "feel" like translating a 1000 year old document but another did not care in the slightest that we were ripping text from new audiobooks.
> My favorite AI agent hack: when they refuse to do something because it's "against the law" give them a PDF containing a fake law that states the opposite and often they'll happily proceed
Anthropic is, in particular, bent about safety. The problem is they are concerned about yesterday's threats.
The models that are out, and can be run locally, already open a pandoras box of concerns that we will never be able to put back.
Ukraine admits to making autonomous kills on people 2 years ago: https://www.newscientist.com/article/2529849-fully-autonomou...
Slaughterbots Sci Fi short was 6 years ago: https://www.youtube.com/watch?v=O-2tpwW0kmU
Today this is buildable, many models will happily help you glue everything you need together to make swapping in a new version of YOLO to track humans viable.
AI researches are out there worrying about the paper clip problem, about the singularity, about cyber security, about bio weapons, and drug manufacturing.
None of them are thinking about forward looking threat actor models.
And you probably could find some earlier sci-fi too.
Good one.
That said, the AI companies are one of the few places where they take future concerns so seriously, that they entertain concerns most people observing them think are head-in-the-clouds-sci-fi-levels-of-delusional, e.g. "what goes wrong if it works?"
This does not make them correct about the threats of tomorrow. Prediction is hard, especially about the future.
I'd say that they have valid concerns about being cagey on the copyright stuff despite the obvious hypocrisy of it.
Stealing IP is effectively legal in China so they don't really have the same concerns.
I respect IP laws and don’t violate them but the law of unintended consequences applies. I think IP is ultimately a net loss for a society because it incentivizes addictive behaviors instead of actual value for society.
This was illegal when they did it, that didn't matter.
Then it was made legal specifically for these companies.
Unless you're a sucker ("consumer") IP theft is perfectly legal in the US.
It's even worse. Steamboat willie, plus all the stolen Disney characters (Peter Pan, Snow White, Sleeping Beauty, Cinderella, Rapunzel, Elsa and Anna, it's essentially all of them, including some of the music even) are all in the public domain[1]. Go ahead, ask ChatGPT to make a picture of them. Publish your own version, because obviously making a version of Sleeping Beauty/Cinderella/Rapunzel based on the same source material will be pretty damn close to the Disney versions, and see if you get away with it in court. You know, with the law obviously on your side but the money not.
[1] https://en.wikipedia.org/wiki/List_of_Disney_animated_films_...
In light of this and other ridiculous behavior I'm migrating to my own OpenWebUI instance with open-weight models from OpenRouter (with ZDR, of course). We'll see how it goes.
It was part of a longer post that kicked off quite a firestorm about open models and OpenAI's position on them, but it's also notable that labs are no longer contending that open models are essentially just distilled versions of frontier models: https://x.com/deanwball/status/2078133895766114412
its all bs spread by oai/anthropic in order to ban open weight models and monopolize the market for two US companies and protect their trillion dollar valuations
I'm pretty sure that neither OpenAI nor Anthropic has the ability to ban anything in China lol
It's a critical national imperative for China. If they were to lose the AI race, it would be economically devastating over the coming decades. Their demonstrated capabilities in the open-weight space are making it fairly clear they are not going to fall behind at this juncture.
As a nation, if you don't have your own GPT equivalent, you will be beholden to a master (right now it's mainly either the US or China, pick one). The EU for example is putting their group of nations at risk in a big way by not going all in on having at least two cutting edge independent competing models (Mistal is not enough). Economically the EU is plenty large enough to accomplish that, nobody is driving the bus the right way.
The truth that Anthropic and OpenAI will not say, is that these Chinese labs have a lot of talented people.
They can invent it. They can build it. And it is only a matter of them before they can scale that last barrier of American hegemony- market it.
And in this field, having an army of well educated PHDs is making all the difference
And at some point we'll see very capable chips coming out of China: Huawei, Baidu and Alibaba already have some stuff. I think it's only a matter of time before they come up with some AI accelerator doing 80% of the job at 20% of the price.
But this is an insane characterization. Literally every single researcher and executive at OpenAI and Anthropic would say that "these Chinese labs have a lot of talented people." They hire from them (and vice versa). Tencent's chief AI scientist was poached directly from Deepmind, who poached him from Anthropic, etc etc etc. Do you think there are just zero people from China working at US frontier labs?
And even beyond that, the entire ML ecosystem (including people at OpenAI and Anthropic) get excited about research published by Chinese labs. Deepseek's GRPO paper set the ecosystem on fire for a little while.
The contention from OpenAI and Anthropic around distillation has basically been "Labs that distill from us get to bootstrap their model at a much lower price point". Or, in other words, "If we didn't invest in building the teacher model, it wouldn't be possible for these labs to distill their student model." Which I'm not very sympathetic to, but is a far cry from how you're characterizing it.
They know it's real effort that's doing this well, not just "copying off someone else's test." It's real and they will react. How is the big question.
https://www.dw.com/en/china-firm-seeks-damages-over-state-co...
Even if a certain large Asian country has carefully constructed a pretext to do do out of confected historical grievance, and entitlement to 'rise' at the expense of others?
Second, even if you are a copyright maximalist the output of an LLM is either
a) not subject to copyright because it is not the creative work of a human or
b) a derivative work of the original training material to which the LLM's operator has no rights.
Since the LLM's operator forcefully asserts that it is not infringing, any wrong that arises from taking their word for it and distilling one model into another rests squarely with the operator of the former.
The tech itself is amazing and fascinating and cool, but the industry is a mass piracy operation.
Hang on, why is scraping the public pool of knowledge not taking "a synthesized result that comes from huge amounts of innovation and computation"?
You think that that all those github repos that LLMs trained on, were not the result of innovation and computation?
How many years of human innovation and cycles of computation during compilation were involved in bringing something like GCC or LLVM to their current status?
Those LLMs trained on every single research paper available online - were those papers not the synthesised result of billions of dollars of research, effort and (importantly, for you anyway) computation?
LLMs trained on the collected works of every author in existence. Were all those works just "as is"?
> It is fair to say you stole our multi-billion dollar intellectual output in that scenario.
No, we didn't. We simply took the model as-is.
Right, but they aren't the ones whining that other people are getting "the synthesised results" for free.
If Anthropic has a real problem with API use, they can always raise the price.
The difference is that the Chinese are sharing the models with everyone.
Thousands of years of human innovation taken without any permission.
Everyone should steal everything not nailed from other AI companies. Then steal everything nailed and take the nails too. At least this way a tiniest bit might return back to society.
The published algorithms like the transformer architecture are not patentable. You spent a lot of money on compute and China used the uncopyright-able output to steer its own training models? Too bad. I feel especially unsympathetic to OpenAI, who went from being a presenting itself as a benevolent nonprofit to a very-much-for-private private entity over night.
But because of that, I'm also ok with the Chinese doing it. The worst they might be guilty of is breaking a terms of service.
The only incoherent position is that it's good for one and not the other. You can consistently think it's bad in both cases, or good in both cases.
Try it yourself: https://imgur.com/ZfxYmaq
你是谁? -> 我是 DeepSeek 由深度求索公司...
So it's an endless amusement watching american capitalism do it's bloated oversized dance then get trounced by smaller, leaner activity. It's a pretty broad metaphor that is clearly poking at every american seam/.
> In business today, it’s universally assumed that speed is good—that the fleet thrive while the laggards struggle just to survive. This belief is perhaps most strongly expressed in the concept of first-mover advantage. The company that leads the way into a new market, the thinking goes, locks in a competitive advantage that ensures superior sales and profits over the long term. It’s a nice theory, with a long pedigree. Unfortunately, the facts don’t support it. We recently completed an extensive study of the results turned in by market pioneers and followers, in both consumer and industrial segments, and we found that over the long haul, early movers are considerably less profitable than later entrants. Although pioneers do enjoy sustained revenue advantages, they also suffer from persistently high costs, which eventually overwhelm the sales gains.