The Private Capture of Public Genius

This was interesting right up until "The fund pays every eligible American the same amount each year. "

I'm in Australia. I've contributed my share of dirt to the delta. Why do I not get a share of this?

I get that the frontier companies are (for the moment) US companies. But that's just corporate ownership, it's not what we're talking about. We're talking about compensating the people who wrote the training data for their contribution. That contribution came from all over the world, so the Corpus Fund needs to be paid all over the world.

Set it up in the UN, get the UN to provide the training data sets as a common good, and have the UN collect the money from all AI companies using the training data sets. And the UN should distribute the money in the most equitable manner globally (so most of it going to alleviate poverty, probably).

I'd happily trade my collected years of shitposts to help folks get out of poverty.

> Frontier science looks different today. It's rooted in model weights and GPUs. It is flooded with token spend and agentic loops. It blooms in data centers.

This seems like handwaving to me. Even the actual frontier science that is using ML (e.g. AlphaFold) isn't based on "token spend" or "agentic loops". I personally cannot think of a single example of frontier science that is rooted in LLMs. I am sure there are a few examples, but the idea that frontier science has somehow completely shifted its trajectory and methods based on LLMs feels like quite a stretch to me.

Got any counterexamples to show me I'm wrong?

We cannot always want to capture only the (temporary) winners whenever we see a lucrative business and expect to share a free ride. I'd also assume that most of the revenue these AI labs are making is turned into depreciating fixed capital (hardware) and OPEX at this point.

Why don't we capture Meta and Google as they allegedly take advantage of more publicly available information for profit? Let alone the truly valuable knowledge, like mathematics, has nothing to do with the majority of garbage posts that an average person would "contribute" on social media.

If we really want to tax or nationalize some economic activity, then, in my opinion, the target should be what it takes from society, not what it produces for society. By this logic, we should tax all labs, including those lagging ones, that utilize the public knowledge.

However, if everyone can access the public knowledge without rendering it less useful or reducing its available quantity, there should be no reason to tax it.

"Fair use" was always fuzzy. To be honest, I care a lot less about slurping up the public internet and private books to make models than about every profession on the planet being forced by their employers to create skills that automate their knowledge work. The latter is much more directly an expropriation, legitimized only by the shortage of work, i.e., market power.

This is a well written essay. I had hoped it might address the role of distillation and open source in diffusing ownership of this technology back to the public that made it possible. And the AI labs’ rank hypocrisy in this area.

The troubles over copyright infringement in AI training data remind me a bit of Eli Whitney and the cotton gin.

There he suffered massive patent infringement, that basically stopped being enforced due to the sheer economic importance of the cotton gin.

In a similar manner, I think there is a reasonably strong argument that it was wrong to use copyrighted material for AI training without paying royalties nor even asking for permission. But equally, every country wants to have the most powerful models and enforcing such royalties would make it effectively impossible to train them as the amount of material required would cost an insane amount in royalty fees.

So I expect the law will continue to turn a blind eye (perhaps enforcing some token payments like that $1.5B mentioned in the article) because "if we don't make these models, the Chinese will" etc.

There’s a quote about how in some articles a switch is quietly flipped in the middle where the article was talking about what is and suddenly the author has everything to say about what should be.

I googled for the quote but all I got is useless web spam and meme style graphics about quotes from writers. But AI told me it was David Hume and provided the full quote.

The real question is when the day will come that AI become the fertile muck that a new thing grows from and clings to and the legal system needs to adjust to. I hope it’s a good thing.

It was a brilliant article, and it succinctly captured the offenses to ethics and humanism posed by LLMs.

I'm not sure it'll get a lot of reception in the technocracy here on HN, whether of the AI booster or AI nihilist sort. However, I think it's a very comprehensive digestion of the questions that will swirl around the idea of LLMs as a public good in the near to medium future.

Anybody know where that gordon moore quote comes from? A little searching didn't produce sources for me.

It's been said a thousand times but the ability to produce copyrighted material is not copyright infringement.

I'll say it again:

You can produce copyrighted material all day long and it's not copyright infringement.

If I draw The Simpsons on a piece of paper - whether or not I used AI to create it - it's not copyright infringement.

Copyright infringement is if I tried to sell that Simpsons work as my own, putting up for consumption and reaping the monetary benefits.

It has never been illegal to produce copyrighted work. You people are in for a surprise dystopia if you keep pushing to make that a reality.

A similar appropriative-use vs. public-trust evaluation is playing out this year as the California State Water Resources Control Board reevaluates Los Angeles’ right to divert water from the Mono Basin in the Eastern Sierra Nevada.

The foundational case for Mono Lake as a public trust resource is National Audubon Society v. Superior Court (1983) [1]. The California Supreme Court evaluated appropriative water rights against the public trust doctrine, took both arguments to their logical extremes, and decided that neither was acceptable in itself.

In a pretty jaw-dropping passage, the Court summarized the Los Angeles Department of Water and Power’s position in relation to appropriative use of water diverted from a unique ecosystem hundreds of thousands of years old:

> Defendant DWP, on the other hand, argues that the public trust doctrine as to stream waters has been "subsumed" into the appropriative water rights system and, absorbed by that body of law, quietly disappeared; according to DWP, the recipient of a board license enjoys a vested right in perpetuity to take water without concern for the consequences to the trust.

The decision in Audubon rejected LADWP’s argument, but it remains a stark example of the beneficiary of a public resource recasting a conditional public license as a permanent private entitlement, apparently free from consequence of accountability for harm inflicted on the public trust.

I think this appropriative-use vs. public-trust/public-benefit discussion is going to define the coming decades. The landscape remains unsettled as it applies to water (especially in a changing climate), much less to data in a period of rapidly evolving technology.

With respect to data, progress could be made by formally establishing a public corpus as an accessible commons, with clear expectations and rights around individual contributions made to third-party platforms. Publicly funded research is still often locked behind paywalls. The contents of the Library of Congress, special collections, municipal libraries, university archives, and museums are publicly owned or publicly supported, yet remain largely inaccessible to the general public.

I expect the “leader” in LLM performance to keep changing, but the accumulated genius of public knowledge to remain far more durable, with periodic and incremental additions. Fighting over small reparations for every scraped post seems less transformative than building a public knowledge commons that anyone can use, converse with, search, train on, and learn from.

reCAPTCHA began as a tool that simultaneously authenticated users while helping verify OCR for the backlog of The New York Times and Project Gutenberg. Maybe it is time for a similar public project to digitize and make accessible the body of public knowledge without surreptitious and ethically dubious appropriation of copyrighted works. Authors, writers, and shitposters could opt in as desired.

I would take a public resource like that well ahead of a few bucks of compensation for my decades of shitposting, just as I'd take a thriving Mono Lake well ahead of compensation for it being relegated into lifeless alkali flat via appropriative water rights.

I'm surprised that the Author has missed a very important corollary to the diffusion of "genius theft" that they are bringing up.

And that is the diffusion of the beneficiaries. Maybe they think that OpenAI and Anthropic are actually worth, like a trillion dollars each and can therefore have value extracted from them. I'm not so convinced.

What if there aren't frontier labs spending billions on training a model. What if, instead, open source is at least mostly competitive with the top models. And if the models are open source (or weights, whatever you want to call it) the people benefiting are actually just rando people or startup founders.

What are you going to do if you want to extract this value from this diffuse set of beneficiaries? Put an arbitrary tax on anyone living on San Francisco or something??

The reality is that the author is trying to put the genie back into the bottle. All technological progress has winners and losers. It has people who are even benefiting from the rest of society and making personal gain based on that.

But, at the end of the day, doing accounting math on how much an individual benefited from a specific common good as vague as societal knowledge is impractical. And yet technological progress benefits all of society in a wholistic sense.

Additionally, the author focuses so much on the extraction of a public good, I am surprised that they failed to address that these labs are creating a public good as well. Who's to say that this "theft" is larger than the production of public goods that these labs give to the public in the first place.

I mean, my life has been massively improved by the fact that I have access to these models. And I'm not convinced that I have produce enough myself to outweigh this benefit that they are giving to me, so I consider it to be a fair trade.

tl;dr

Author fears pending/future copyright litigation against AI labs, proposes "Corpus Royalty"