> A.I. is built on our collective intelligence: our books, songs, artwork, journalism, computer code, scientific research, videos, conversations, images and ideas spanning generations
I know many here would scoff at nationalizing a private company, but AI is a usurpation of human knowledge and quite literally at times. (Every AI company was embroiled in copyright lawsuits and lord knows what Qwen et al are up to.)
In turn, everyone knows labor displacement is coming. My bet is the next recession will end up being brutal for this reason. To me, labor displacement and the social consequences are a potentially *catastrophic* negative externality. Should not there be a tax to offset the "frictional" unemployment? What happens when people lose a high skill job and will no longer be able to afford their mortgage?
Also, why are people always talking about AI as if its an angel or satan? The degree to which we're doomed is an open question, much like a tornado... so why aren't we thinking about taxes on AI like a tornado insurance fund?
> In turn, everyone knows labor displacement is coming
What % of Americans would agree to that. 10%?
I agree for those of us in the field, we are a lot more confident. But I dont think my retired parents have any idea what AI is or have ever touched anything more complicated than Alexia.
You can tax something without owning it. It is the owning part that bothers me. It implies that the government is going to shovel ridiculous amounts of money into these things and when the bubble finally pops We The People get nothing out of it.
I think you've misunderstood the proposal. The government levies a tax in the form of shares, not cash. It doesn't pay for the shares.
FTA: "It would create a sovereign wealth fund through a one-time 50 percent tax — not on the profits of OpenAI, Anthropic, xAI and other companies, but paid with something far more valuable than that: the stock."
I don't know what their plan is, but that is one way to "own half". I'm really hoping that isn't their plan, but this administration seems to love shoveling tax money to people who don't deserve it.
Taxing X% ownership means you get X%. You don't pay for it. So if the bubble pops, you still get X% ownership. To your point, it may be smarter to tax the IPO valuation and buy more later.
I'm mostly commenting on the story a few days ago regarding having index funds change their rules to automatically purchase shares in the impending IPOs so that passive investors would end up buying shares without their direct knowledge. It struck me as a way for executives to cash out of what they know is a bubble. A lot of people in that thread didn't seem to have any issue with that.
Labor displacement is always coming. Every new technology eliminates some old occupations and creates new ones. LLMs aren't unique in that regard. We should have a safety net to support and retrain displaced workers regardless of the technology.
And let's please not have any lazy, low-effort replies claiming that AI will somehow magically eliminate all jobs for humans.
> And let's please not have any lazy, low-effort replies claiming that AI will somehow magically eliminate all jobs for humans.
I don't think AI actually capable of doing so, but AI companies need to stop bragging about wanting to do this and making it their goal if you don't want people to keep bringing it up.
I would rather we have an alternative where you can buy Anthropic, OpenAI, SpaceX, etc stock but it has a few strict requirements:
* Investors are committing to keeping their money over a 10 year period, and you pay them dividends
* At the end of the 10 years you can withdraw it all, or keep your stock but withdraw the dividends, or keep your stock and reinvest the dividends for the next 10 years.
* Companies cannot be on any other type of "stock market"
* Companies need to be identified as producing something of significant value, like SpaceX's rockets would qualify.
* It can be part of anyone's 401k since these companies would be scrutinized.
Ideally a social media only company would NEVER qualify. I think the stock market is too happy to buy and sell, when some of these companies are a long term investment that has slowly paid off in ways most people cannot fathom or comprehend including healthcare discoveries.
They do this in Alaska's fund with great success[0] and Alaska is a deep red state. It's primarily funded by oil and gas revenue and other mineral royalties. Also thats only (at least) 25% of total revenue, the other 75% the state spends directly.
Norway's oil fund is another famous example mentioned in the article.[1]
The oil is on Alaska/Norway's land, so they can choose what they want to do with money from selling that oil including distributing it or purchasing assets. Sanders seem to be proposing to arbitrarily seize half of the companies.
The oil companies are exploiting a public asset, so it's obvious the public would benefit. Who are the AI companies paying for the training data they're using?
Because AI companies basically took everything we ever wrote, drew, recorded, posted, or thought and turned it into a product with the power to lie, propagandize, and manipulate the public with zero oversight. Walmart is a parasite using welfare to subsidize their operations but they didn't tell a judge that they were immune to copyright because they stole just so much damn information.
Yes but all the AI companies took all the public data, so when you pay for an AI model you are paying for the marginal service of building a model off that data, not for the data itself. What we should do is ensure that the data is available to more people to train AI models... but sadly this doesn't seem to be happening. Instead AI companies that were first-movers got to train off public data, and as the companies and businesses that own this data get wise they're going to start charging people to train off the data. This will make it much more difficult for anyone to train a model in the future as it will become expensive, and the companies that did happen to already train off public data will get a bit of incumbent's advantage.
I don't really buy this argument. When you buy a physical product, you are paying the entire product lifecycle, not just the marginal aspect of retail distribution. This is the same thing. The marginal inference has to come FROM somewhere. It doesn't just appear out of nowhere.
Yes. Took. As in: without permission. Didn't ask before hand, didn't provide a way to opt-out (although that would also be problematic), didn't ask for volunteers. Took.
If I can go to jail over it then they should too. Let's not judge them by some imaginary ideal world while judging individuals by the present crushing reality.
I interpreted it to mean people feel as though they didn’t consent to having their information trained on, because for many folks, they published articles, open source projects, etc. assuming that they were only helping other people. It’s quite a shock to see megacorps use such data to create machines which threaten the livelihoods of the original authors themselves.
Also, much of the data used to train LLMs are not strictly public domain. For example, copyrighted books and source code with attribution-requiring licenses feature heavily in many corpuses. There are still pending lawsuits against the labs here, yet they continue to push forward. It’s no surprise that there is popular demand for redistribution.
> A sovereign wealth fund? That seems like a great tool for a certain corrupt politician to use as a carrot to make CEO's bend to his/her whims.
Use the sovereign wealth fund to reduce income taxes. Other countries manage to run wealth funds without corruption. If a country can't do it then, I hate to break it to you, it's definitely not the fund's fault.
The focus on AI is just to capture some of the current zeitgeist. Socialists generally think that most large businesses be run for the public benefit in some form or another.
This is basically making that point with AI companies as their true influence is rapidly increasing. The rhetorical strategy here is to hook socialist ideas onto something people are already thinking about, a land-and-expand rhetorical strategy.
Personally I'm against the proposal, but the details are not the point because it's not going happen in this form. It's about changing minds first, then changing policy down the road.
> Let us be clear. Artificial intelligence was not created out of thin air. The data and language used by generative A.I. tools didn’t just pop into Sam Altman’s head or Elon Musk’s imagination. A.I. is built on our collective intelligence: our books, songs, artwork, journalism, computer code, scientific research, videos, conversations, images and ideas spanning generations.
And the ones spanning generations were completely fair game. The ones that were still under copyright are a different matter. (It has not, so far as I know, been legally decided, but it's probably going to depend on whether AI is a transformative use.)
> And the ones spanning generations were completely fair game.
No they weren't. Copyright at this point covers things for at least half a dozen generations back, and is intentionally made annoying enough that it is difficult to find out what is covered and what isn't. LLM companies didn't bother with any of that (they just pirated like your average online 13 year old), meanwhile archive.org got sued for pulling the music off ancient wax cylinders.
Clean up and dramatically shorten, restrict, or even eliminate copyright, and we can start talking about what's fair game or not. People were afraid to sing "Happy Birthday" in movies for probably 80 years, and the corporations that own all IP made it very clear at the time that they preferred for the status of "Happy Birthday" to remain unclear, and would send you a scary letter if you used it.
> probably going to depend on whether AI is a transformative use.
It's probably going to be entirely political, and decided through corruption. It's obviously a mechanical transformation. If rap DJ's got sued for cramming songs full of 80 extremely manipulated samples that you'd need a forensics expert to trace, and all sampled music had to revert to a form where they'd license a single song and re-release it, LLMs are a violation. DJ's doing an absolutely creative translation, and LLMs are not creative, they are pinball machines.
Lately I've been thinking about digital co-ops. Imagine a member owned server farm for running private AI models and cloud services. It could be small, like a neighborhood or citywide project running on cobbled together used hardware like that post the other day where someone got an older server GPU running on his PC. Or it could be a small data center for small businesses and individual users to use. Members could have more privacy than giving all their data to companies like Google and they could have a say in what hardware is used and what type of energy generation is used to power it. There seem to have been more posts over the last couple years about running home media servers and getting away from subscription services; maybe something like that could be an alternative to what we have now.
Strictly speaking, the big A.I. companies _want_ the public to own half of them. Passively. In index ETFs in their 401(k)s and other retirement portfolios. That way the get all the money without any of the actual influence.
Between California and the federal government the public is already going to receive half the value of the equity in cash. The only question is will they buy back in ?
This is more interesting than it looks, because it's using normal corporate control rather than regulation.
Obviously the current stakeholders hate losing control and wealth, but that's not the biggest issue.
Senator Sander's goal is not for some vast public to share the wealth, but for the government to have a veto on what gets done, to limit the collateral damage/exported costs. That's a classical government function.
However, the record of regulatory capture is nearly perfect, so it's likely the reverse would be achieved: the government-sponsored providers being a required intermediary in all knowledge work, with a corresponding incentive to seize those reins.
The probabilistic range of possibilities looks bleak: Now that all regulatory or quasi-governmental agencies of any import (Fed, FDA, EPA, Congress, Courts, CPB) have demonstrated remarkable plasticity to political whim, one would anticipate the worst would come of creating a political franchise out of fighting for control over AI; it would corrupt other aspects of politics.
Taking 50% of a few valuable companies for no compensation is just expropriation. Not that this has political legs, but even if it did, expect an enormous fight. The more obviously legal (and stupendously expensive) route would be eminent domain.
I don’t have an opinion on this specifically, but I am glad to hear a politician talking about the effects of AI.
I feel like I’m going crazy sometimes. Over the next few years we will see the biggest change to employment our country has ever seen. Our entire financial structure is about to be upended, and not a single politician is talking about it. It’s so weird that all I think about is AI, yet not a single politician seems to notice. (Or maybe they do and that’s why they’re pillaging the country.)
Wouldn't taxes give more to the public than nationalization? I'd like the benefits of Communism without the downsides.
There are efficiency benefits to the government owning stock vs. using the IRS for collection, that part I like. But I don't trust the US gov to use stock voting rights wisely.
I strongly dislike the belief that people should be compensated when others find ingenious ways to profit off of publicly-available data.
We live in a world where "creating value" (doing things that others find helpful) and "capturing value" (getting those people to pay you money) are two different things. If I give my mom a hug, I'm creating value, but that's not necessarily something I'm going to charge her for it. Most value created by people won't ever be captured. And that's a good thing imo.
It keeps the world moving, removes friction, and allows for authenticity. There's nothing wrong with wanting to capture value, of course. But the second you do that, you're a business. And "capturing value" has a huge set of tasks and responsibilities you now have to handle.
But there's an intuition that has gradually built up over centuries, alongside the growth of "intellectual property" as a concept. It's best paraphrased as, "I want to be compensated for the value I create, without doing any of the work to capture it. And if someone else finds an ingenious way to capture some of the value that I've created, then they should pay me."
To some degree, I understand and agree with the sentiment.
Nothing is built in a vacuum. No person or company is an island. Everything is built on top of public infrastructure and works created by the country, laid by our forefathers. This is just one of many reasons why I believe in a progressive tax system. To the extent that you're able to capture large amounts of value in America, a lot of that is made possible by the infra you're building on top of, which is owned by the public, and a progressive tax system is a good way to to share that with the public.
(Of course, this has its own problems, bc the government collecting taxes is not enough, it has to spend those funds wisely, for the benefit of all. Which it obviously doesn't do, at the federal level, or at many state and city levels. So I've always found it a bit perplexing for people to clamor for more taxation while caring little about how tax revenues are spent. But that's a discussion for another time.)
But overall, I don't like this intuition, because it's essentially rent-seeking behavior.
Capturing value is hard. Simply creating value is not enough. If you write a song, or you build an app, or you cook a meal, you still have all your work ahead of you to find a customer/consumer, and understand what they find valuable enough to pay for, and ensure your offering matches that, and do the marketing/sales to get it in front of them, and convince them to pay, and scale to more people, and manage your books, and do all of this profitably.
Expecting to be paid for simply creating value but doing none of the work to capture the value to me feels a little bit entitled. Or, at the very least, naive.
What's interesting is that certain industries have more or less entitlement here, depending on the influence of "intellectual property" in that industry.
For example, there's almost no concept of intellectual property in the cooking. If you invent a new recipe, you can't really patent it and tell everybody else that they're not allowed to make it. So, pretty much every chef is okay with the fact that they need to actually capture value by opening a restaurant or going to work for one.
It's similar in the software industry, where rather than patenting all of our software and trying to enforce it, we generally do the opposite and release software in an open-source way. We're quite aware that if we want to profit, we'll need to start our own startup, and we have no qualms with that.
But with writing, music, etc., you see a lot more creators who want to just do the creation part, who don't want to do the business part, but who then want the profits that the business part enables.
I can empathize for sure, I get it. But I think a world with less rent-seeking behavior is better. A world where more people understand what it takes to capture value and are willing to do it (or happy to just not do it) is a better world. A world where more people feel entitled to the profits earned by those who are able to capture value, I think, is worse.
As long as they're non-voting shares, I don't see the harm.
I assume not enough politicians in this senator's camp were given their early cut so this is retribution/a lesson to the abstract "Big Tech" to show that DC is still the city that rules the world.
Why non-voting shares? That kind of defeats the point of partial ownership.
If the models were built using the output of all of human effort over time, then society at large should absolutely have a voice in the direction of the companies.
This tech should be made for public benefit, not for purely profit and private interests.
Quite frankly, most companies should be worker co-ops instead, and its long passed time we start moving to that model.
> Why non-voting shares? That kind of defeats the point of partial ownership.
Not really, no. Ownership gets you a share of the profits and profits can be used to reduce income taxes. I think non-voting is wise. It prevents political and partisan meddling.
This would have a massive chilling effect on the private sector as a whole. IMO it would completely destroy investment in America. American companies and markets get extraordinary investor interest due to strong property rights. Once those rights are gone there will be massive capital flight and greatly reduced investment.
Imo this proposal is even worse than a billionaire wealth tax (which has all sorts of implementation issues).
Why does the public have a right to expropriate the property if AI companies specifically, as opposed to other types of companies? Just make broad rules that apply to everyone based on abstract principles. I’m fine even with very liberal economic approaches. If we want to raise corporate tax rates to 30%, fine, do that. Want to get creative? Half the equity of every company goes into a public fund. But this case by case, “sure is a nice company you got there” stuff is third-world shit.
This sounds like the new "thats people's retirement," because if we convert 50% of every AI company into a sovereign wealth fund (which is already a questionable seizure anyway), suddenly it will become politically untenable to do anything that might put that fund in danger, like... regulating anything, or even not bailing out a company thats struggling.
The quickest way to lose the AI battle to China and reinstate censorship and model political bias is to hand the industry over to the government. This doesn’t just benefit democrats, whatever political party is in power will use them as tools of political means. Not to mention almost certainly pricing won’t act in accordance with free market forces but instead be regulated. See healthcare.
Most of us are. That's what taxes are, in fact. We have private property rights, and we have to pay taxes. The hard-core anti-tax people try to make that a contradiction; most of the rest of us don't think of it that way, so their "taxation is theft" rhetoric falls flat.
But I think it's important that we keep it at some. Major erosions of private property rights in order "to tax the rich" make me nervous, because I don't want to lose those rights myself. And taking the rights from them, but I get to keep them, seems likely to not be a stable equilibrium.
I don't think LLMs will be the path to AGI, but anything nature can already build (a human mind) is proven to be possible. After that, it is just a science and engineering problem to replicate.
There are those who believe that a human mind is somehow magic and a special exception to the laws of physics, but I am not one of those people.
"Envy was once considered to be one of the seven deadly sins before it became one of the most admired virtues under its new name, social justice."
- Thomas Sowell
I'm sorry to say, but this is a losing position, as in one you only adopt when you've already lost and are trying to bargain. It presupposes that the AI companies are unregulateable, and that the only possible avenue of influence/dividend is through ownership. Contrast with the traditional idea that when companies create harm, the government works to stop that harm by default. Or that if these companies actually do succeed at rolling up up the distributed economy into a handful of centralized companies, the government steps in to tax their outsized gains to preserve some semblance of a distributed economy. Furthermore, what does said "ownership" actually do ? If the government is unable to regulate these companies, then it is also unable to reliably exercise a voting interest or insist on receiving dividends - if the companies are this powerful, then whoever actually controls them can always alter the terms and reject the "owners'" demands.
I am repulsed by this because it will obviously be the vehicle through which tax money will be directed into Altman & Co's pockets, but I also understand that they will get bailed out whether the government gets a share or not.
As long as they are voting shares, I don't see an increase in the harm. I'd like to see a legislative framework about how that ownership is handled that allows Congress and regulatory agencies to make decisions restricting how these companies will operate, but without any regard to the constitutional rights of the corporate persons or their owners.
I'm sick of the government arguing with monopolies, then taking dives. I want it to be abundantly clear that government has the ability to restrict these AI utility companies freely (such as their ability to feed on their customers), while still limiting the rights that the state has over the personal use of AI by private individuals. Partial state ownership will make that possible. Hell, let half their boards be publicly elected.
People on HN generally love municipal broadband. For good reasons. It is almost without exception better than any of the national ISPs. Cheaper too.
Municipal broadband is just 100% publicly-owned. That's what that means. When you have a national ISP, you might not get a service at all despite the ISP guaranteeing service in exchange for money from the state they've taken. You get a service that starts at $60 but somehow gets to $140 in a few years unless you do the annual cancel dance and if you do cancel you have no other options anyway. And what are you really paying for? Lobbying to make municipal broadband illegal.
And these same people will defend the status quo because of "property rights". Nobody here is Jeff Bezos. Does it seem like things are going well? Is this a legitimate belief in unfettered property rights? Or is it just that you believe you'll be Jeff Bezos one day so you'll benefit from the status quo?
This is the origin of the quote possibly misattributed to Steinbeck that Americans view themselves as "temporarily embarrassed millionaires".
You don’t have to be Jeff Bezos to believe you’re better off in the long run with strong property rights. To your example: I’ve got 10 gig fiber from a private company. My alternatives are 2 gig fiber from a private company, or 2 gig cable from a private company. I don’t have to be Jeff Bezos to wager that, if the people who ran Baltimore City Public Schools ran my broadband service, I wouldn’t have even 1 gig service.
The fact is, we have tried exactly what you’re suggesting. We don’t even need to get to the brutal communism of China or the Soviet Union. The “social democracy” of India and Bangladesh (where I’m from) left those countries’ economies in the toilet.
The best case scenario for social democracy was the DDR. My wife lived for a year in former eastern germany—in 2001, when the socialist government was as far back in time as Obama’s first term is to today—and folks recall being reasonably happy under socialism. But western germany was still much more prosperous, and it took huge solidarity payments to help the former east germany catch up.
Bernie Sanders. I mean, he's not always wrong, but he's, um, kind of enthusiastic about just taking stuff from those who have it, considerably more than the current understanding of private property (or even taxation) considers acceptable.
>kind of enthusiastic about just taking stuff from those who have it, considerably more than the current understanding of private property (or even taxation) considers acceptable.
Funny how you can use that description for how AI companies used everything for training data.
The obvious counterargument is that the AI labs took virtually all human writing, imagery, music, etc., without regard for licensing or copyright. It's fair to ask what they owe back to the commons they built their models on (and which they are in some sense helping to destroy).
China and Russia are simple examples of the fallout of "cutting king's heads".
So awesome in China that they went straight back to a lifetime emperor now.
The French Revolution and its fallout (communism, fascism) killed more humans than any other historic movement, triggering 2 world wars, including numerous genocides.
If these companies intend to destroy the fabric of society and jobs and livelihood of everyone, then that leaves us very few choices as a society. This is one of the tamest and most peaceful ones, even if it's just a start. Hopefully Sammy and friends choose wisely.
> A.I. is built on our collective intelligence: our books, songs, artwork, journalism, computer code, scientific research, videos, conversations, images and ideas spanning generations
I know many here would scoff at nationalizing a private company, but AI is a usurpation of human knowledge and quite literally at times. (Every AI company was embroiled in copyright lawsuits and lord knows what Qwen et al are up to.)
In turn, everyone knows labor displacement is coming. My bet is the next recession will end up being brutal for this reason. To me, labor displacement and the social consequences are a potentially *catastrophic* negative externality. Should not there be a tax to offset the "frictional" unemployment? What happens when people lose a high skill job and will no longer be able to afford their mortgage?
Also, why are people always talking about AI as if its an angel or satan? The degree to which we're doomed is an open question, much like a tornado... so why aren't we thinking about taxes on AI like a tornado insurance fund?
> In turn, everyone knows labor displacement is coming
What % of Americans would agree to that. 10%?
I agree for those of us in the field, we are a lot more confident. But I dont think my retired parents have any idea what AI is or have ever touched anything more complicated than Alexia.
You can tax something without owning it. It is the owning part that bothers me. It implies that the government is going to shovel ridiculous amounts of money into these things and when the bubble finally pops We The People get nothing out of it.
I think you've misunderstood the proposal. The government levies a tax in the form of shares, not cash. It doesn't pay for the shares.
FTA: "It would create a sovereign wealth fund through a one-time 50 percent tax — not on the profits of OpenAI, Anthropic, xAI and other companies, but paid with something far more valuable than that: the stock."
That sounds like a brilliant idea if you're a board member of one of these companies.
Who said anything about the government shoveling in money?
Usually to "own" something requires to spend money to get it.
The government gets half my paycheck without buying it from me.
Wait, is the idea to buy that 50%?
I don't know what their plan is, but that is one way to "own half". I'm really hoping that isn't their plan, but this administration seems to love shoveling tax money to people who don't deserve it.
Taxing X% ownership means you get X%. You don't pay for it. So if the bubble pops, you still get X% ownership. To your point, it may be smarter to tax the IPO valuation and buy more later.
I'm mostly commenting on the story a few days ago regarding having index funds change their rules to automatically purchase shares in the impending IPOs so that passive investors would end up buying shares without their direct knowledge. It struck me as a way for executives to cash out of what they know is a bubble. A lot of people in that thread didn't seem to have any issue with that.
Labor displacement is always coming. Every new technology eliminates some old occupations and creates new ones. LLMs aren't unique in that regard. We should have a safety net to support and retrain displaced workers regardless of the technology.
And let's please not have any lazy, low-effort replies claiming that AI will somehow magically eliminate all jobs for humans.
What are the jobs AI is creating, specifically?
> And let's please not have any lazy, low-effort replies claiming that AI will somehow magically eliminate all jobs for humans.
I don't think AI actually capable of doing so, but AI companies need to stop bragging about wanting to do this and making it their goal if you don't want people to keep bringing it up.
> And let's please not have any lazy, low-effort replies claiming that AI will somehow magically eliminate all jobs for humans.
This is essentially what a handful of c-suite execs have been telling the world for the past 2-3 years is it not?
I would rather we have an alternative where you can buy Anthropic, OpenAI, SpaceX, etc stock but it has a few strict requirements:
* Investors are committing to keeping their money over a 10 year period, and you pay them dividends
* At the end of the 10 years you can withdraw it all, or keep your stock but withdraw the dividends, or keep your stock and reinvest the dividends for the next 10 years.
* Companies cannot be on any other type of "stock market"
* Companies need to be identified as producing something of significant value, like SpaceX's rockets would qualify.
* It can be part of anyone's 401k since these companies would be scrutinized.
Ideally a social media only company would NEVER qualify. I think the stock market is too happy to buy and sell, when some of these companies are a long term investment that has slowly paid off in ways most people cannot fathom or comprehend including healthcare discoveries.
Why half of AI and not half of Walmart & Exxon & Apple?
Government spending is already ~40% of GDP.
And what do we get with this half?
A sovereign wealth fund? That seems like a great tool for a certain corrupt politician to use as a carrot to make CEO's bend to his/her whims.
What benefit does that have for anyone else?
You can't build a sovereign wealth fund if you're transferring all the money out. That's just more government spending, not a wealth fund...
They do this in Alaska's fund with great success[0] and Alaska is a deep red state. It's primarily funded by oil and gas revenue and other mineral royalties. Also thats only (at least) 25% of total revenue, the other 75% the state spends directly.
Norway's oil fund is another famous example mentioned in the article.[1]
[0]: https://apfc.org/
[1]: https://en.wikipedia.org/wiki/Government_Pension_Fund_of_Nor...
The oil is on Alaska/Norway's land, so they can choose what they want to do with money from selling that oil including distributing it or purchasing assets. Sanders seem to be proposing to arbitrarily seize half of the companies.
The oil companies are exploiting a public asset, so it's obvious the public would benefit. Who are the AI companies paying for the training data they're using?
They seized all of humanity's intellectual output first.
> The oil is on Alaska/Norway's land
These AI companies are on American land.
> Sanders seem to be proposing to arbitrarily seize half of the companies.
I guess it's only ok when the government arbitrarily seizes half my paycheck.
Because AI companies basically took everything we ever wrote, drew, recorded, posted, or thought and turned it into a product with the power to lie, propagandize, and manipulate the public with zero oversight. Walmart is a parasite using welfare to subsidize their operations but they didn't tell a judge that they were immune to copyright because they stole just so much damn information.
Yes but all the AI companies took all the public data, so when you pay for an AI model you are paying for the marginal service of building a model off that data, not for the data itself. What we should do is ensure that the data is available to more people to train AI models... but sadly this doesn't seem to be happening. Instead AI companies that were first-movers got to train off public data, and as the companies and businesses that own this data get wise they're going to start charging people to train off the data. This will make it much more difficult for anyone to train a model in the future as it will become expensive, and the companies that did happen to already train off public data will get a bit of incumbent's advantage.
I don't really buy this argument. When you buy a physical product, you are paying the entire product lifecycle, not just the marginal aspect of retail distribution. This is the same thing. The marginal inference has to come FROM somewhere. It doesn't just appear out of nowhere.
“Took?” AI companies aren’t removing the information from the public domain. What happened to “information wants to be free?”
Yes. Took. As in: without permission. Didn't ask before hand, didn't provide a way to opt-out (although that would also be problematic), didn't ask for volunteers. Took.
The word you're looking for is "copied".
Don't fall for the great lie of intellectual "property".
If I can go to jail over it then they should too. Let's not judge them by some imaginary ideal world while judging individuals by the present crushing reality.
I interpreted it to mean people feel as though they didn’t consent to having their information trained on, because for many folks, they published articles, open source projects, etc. assuming that they were only helping other people. It’s quite a shock to see megacorps use such data to create machines which threaten the livelihoods of the original authors themselves.
Also, much of the data used to train LLMs are not strictly public domain. For example, copyrighted books and source code with attribution-requiring licenses feature heavily in many corpuses. There are still pending lawsuits against the labs here, yet they continue to push forward. It’s no surprise that there is popular demand for redistribution.
Are you proposing the Aaron Schwartz treatment by the government for Zuckerberg, Altman, and Amodei?
Individuals have faced federal charges and served prison time for reselling copyrighted content. I don't see the same happening to AI execs.
> Government spending is already ~40% of GDP.
> A sovereign wealth fund? That seems like a great tool for a certain corrupt politician to use as a carrot to make CEO's bend to his/her whims.
Use the sovereign wealth fund to reduce income taxes. Other countries manage to run wealth funds without corruption. If a country can't do it then, I hate to break it to you, it's definitely not the fund's fault.
Because unlike AI companies Apple, Walmart, Exxon are not taking away all human generated knowledge and dumping in their data centers?
The focus on AI is just to capture some of the current zeitgeist. Socialists generally think that most large businesses be run for the public benefit in some form or another.
This is basically making that point with AI companies as their true influence is rapidly increasing. The rhetorical strategy here is to hook socialist ideas onto something people are already thinking about, a land-and-expand rhetorical strategy.
Personally I'm against the proposal, but the details are not the point because it's not going happen in this form. It's about changing minds first, then changing policy down the road.
From the article:
> Let us be clear. Artificial intelligence was not created out of thin air. The data and language used by generative A.I. tools didn’t just pop into Sam Altman’s head or Elon Musk’s imagination. A.I. is built on our collective intelligence: our books, songs, artwork, journalism, computer code, scientific research, videos, conversations, images and ideas spanning generations.
And the ones spanning generations were completely fair game. The ones that were still under copyright are a different matter. (It has not, so far as I know, been legally decided, but it's probably going to depend on whether AI is a transformative use.)
> The ones that were still under copyright are a different matter.
Given the sheer volume of information posted to the Internet in the last 40-50 years, I'd wager that covers 80% or more of the relevant input data.
Old text is relatively scarce in the grand scheme of things.
But I have no real clue, just spitballing.
> And the ones spanning generations were completely fair game.
No they weren't. Copyright at this point covers things for at least half a dozen generations back, and is intentionally made annoying enough that it is difficult to find out what is covered and what isn't. LLM companies didn't bother with any of that (they just pirated like your average online 13 year old), meanwhile archive.org got sued for pulling the music off ancient wax cylinders.
Clean up and dramatically shorten, restrict, or even eliminate copyright, and we can start talking about what's fair game or not. People were afraid to sing "Happy Birthday" in movies for probably 80 years, and the corporations that own all IP made it very clear at the time that they preferred for the status of "Happy Birthday" to remain unclear, and would send you a scary letter if you used it.
> probably going to depend on whether AI is a transformative use.
It's probably going to be entirely political, and decided through corruption. It's obviously a mechanical transformation. If rap DJ's got sued for cramming songs full of 80 extremely manipulated samples that you'd need a forensics expert to trace, and all sampled music had to revert to a form where they'd license a single song and re-release it, LLMs are a violation. DJ's doing an absolutely creative translation, and LLMs are not creative, they are pinball machines.
You get Social Security and Medicare. Which are cheaper now than they will ever be, as the population is only getting older.
They know it will tank and want you money to save their friends.
> sovereign wealth fund? That seems like a great tool for a certain corrupt politician to use as a carrot to make CEO's bend to his/her whims.
Taking ownership of these companies is a Bernie thing not a Trump thing, but a sovereign wealth fund is used to pay off national debt.
Lately I've been thinking about digital co-ops. Imagine a member owned server farm for running private AI models and cloud services. It could be small, like a neighborhood or citywide project running on cobbled together used hardware like that post the other day where someone got an older server GPU running on his PC. Or it could be a small data center for small businesses and individual users to use. Members could have more privacy than giving all their data to companies like Google and they could have a say in what hardware is used and what type of energy generation is used to power it. There seem to have been more posts over the last couple years about running home media servers and getting away from subscription services; maybe something like that could be an alternative to what we have now.
Strictly speaking, the big A.I. companies _want_ the public to own half of them. Passively. In index ETFs in their 401(k)s and other retirement portfolios. That way the get all the money without any of the actual influence.
Sure but they want the public to pay for the privilege of owning half.
Between California and the federal government the public is already going to receive half the value of the equity in cash. The only question is will they buy back in ?
This is more interesting than it looks, because it's using normal corporate control rather than regulation.
Obviously the current stakeholders hate losing control and wealth, but that's not the biggest issue.
Senator Sander's goal is not for some vast public to share the wealth, but for the government to have a veto on what gets done, to limit the collateral damage/exported costs. That's a classical government function.
However, the record of regulatory capture is nearly perfect, so it's likely the reverse would be achieved: the government-sponsored providers being a required intermediary in all knowledge work, with a corresponding incentive to seize those reins.
The probabilistic range of possibilities looks bleak: Now that all regulatory or quasi-governmental agencies of any import (Fed, FDA, EPA, Congress, Courts, CPB) have demonstrated remarkable plasticity to political whim, one would anticipate the worst would come of creating a political franchise out of fighting for control over AI; it would corrupt other aspects of politics.
Taking 50% of a few valuable companies for no compensation is just expropriation. Not that this has political legs, but even if it did, expect an enormous fight. The more obviously legal (and stupendously expensive) route would be eminent domain.
To paraphrase Margin Call, the pubic will soon be left holding the biggest bag of odorous Xcrement assembled in the history of IPOs.
We will all own half, just not the good half.
The public includes the rest of the world, not just a small minority living in the United States.
I don’t have an opinion on this specifically, but I am glad to hear a politician talking about the effects of AI.
I feel like I’m going crazy sometimes. Over the next few years we will see the biggest change to employment our country has ever seen. Our entire financial structure is about to be upended, and not a single politician is talking about it. It’s so weird that all I think about is AI, yet not a single politician seems to notice. (Or maybe they do and that’s why they’re pillaging the country.)
> It’s so weird that all I think about is AI,
Maybe this is a place to start?
Wouldn't taxes give more to the public than nationalization? I'd like the benefits of Communism without the downsides.
There are efficiency benefits to the government owning stock vs. using the IRS for collection, that part I like. But I don't trust the US gov to use stock voting rights wisely.
Once they IPO, the public will in fact own much of them.
I strongly dislike the belief that people should be compensated when others find ingenious ways to profit off of publicly-available data.
We live in a world where "creating value" (doing things that others find helpful) and "capturing value" (getting those people to pay you money) are two different things. If I give my mom a hug, I'm creating value, but that's not necessarily something I'm going to charge her for it. Most value created by people won't ever be captured. And that's a good thing imo.
It keeps the world moving, removes friction, and allows for authenticity. There's nothing wrong with wanting to capture value, of course. But the second you do that, you're a business. And "capturing value" has a huge set of tasks and responsibilities you now have to handle.
But there's an intuition that has gradually built up over centuries, alongside the growth of "intellectual property" as a concept. It's best paraphrased as, "I want to be compensated for the value I create, without doing any of the work to capture it. And if someone else finds an ingenious way to capture some of the value that I've created, then they should pay me."
To some degree, I understand and agree with the sentiment.
Nothing is built in a vacuum. No person or company is an island. Everything is built on top of public infrastructure and works created by the country, laid by our forefathers. This is just one of many reasons why I believe in a progressive tax system. To the extent that you're able to capture large amounts of value in America, a lot of that is made possible by the infra you're building on top of, which is owned by the public, and a progressive tax system is a good way to to share that with the public.
(Of course, this has its own problems, bc the government collecting taxes is not enough, it has to spend those funds wisely, for the benefit of all. Which it obviously doesn't do, at the federal level, or at many state and city levels. So I've always found it a bit perplexing for people to clamor for more taxation while caring little about how tax revenues are spent. But that's a discussion for another time.)
But overall, I don't like this intuition, because it's essentially rent-seeking behavior.
Capturing value is hard. Simply creating value is not enough. If you write a song, or you build an app, or you cook a meal, you still have all your work ahead of you to find a customer/consumer, and understand what they find valuable enough to pay for, and ensure your offering matches that, and do the marketing/sales to get it in front of them, and convince them to pay, and scale to more people, and manage your books, and do all of this profitably.
Expecting to be paid for simply creating value but doing none of the work to capture the value to me feels a little bit entitled. Or, at the very least, naive.
What's interesting is that certain industries have more or less entitlement here, depending on the influence of "intellectual property" in that industry.
For example, there's almost no concept of intellectual property in the cooking. If you invent a new recipe, you can't really patent it and tell everybody else that they're not allowed to make it. So, pretty much every chef is okay with the fact that they need to actually capture value by opening a restaurant or going to work for one.
It's similar in the software industry, where rather than patenting all of our software and trying to enforce it, we generally do the opposite and release software in an open-source way. We're quite aware that if we want to profit, we'll need to start our own startup, and we have no qualms with that.
But with writing, music, etc., you see a lot more creators who want to just do the creation part, who don't want to do the business part, but who then want the profits that the business part enables.
I can empathize for sure, I get it. But I think a world with less rent-seeking behavior is better. A world where more people understand what it takes to capture value and are willing to do it (or happy to just not do it) is a better world. A world where more people feel entitled to the profits earned by those who are able to capture value, I think, is worse.
The public should own 100% of AI companies. Why stop at 50%?
As long as they're non-voting shares, I don't see the harm.
I assume not enough politicians in this senator's camp were given their early cut so this is retribution/a lesson to the abstract "Big Tech" to show that DC is still the city that rules the world.
Why non-voting shares? That kind of defeats the point of partial ownership.
If the models were built using the output of all of human effort over time, then society at large should absolutely have a voice in the direction of the companies.
This tech should be made for public benefit, not for purely profit and private interests.
Quite frankly, most companies should be worker co-ops instead, and its long passed time we start moving to that model.
> Why non-voting shares? That kind of defeats the point of partial ownership.
Not really, no. Ownership gets you a share of the profits and profits can be used to reduce income taxes. I think non-voting is wise. It prevents political and partisan meddling.
This would have a massive chilling effect on the private sector as a whole. IMO it would completely destroy investment in America. American companies and markets get extraordinary investor interest due to strong property rights. Once those rights are gone there will be massive capital flight and greatly reduced investment.
Imo this proposal is even worse than a billionaire wealth tax (which has all sorts of implementation issues).
This would be for the benefit of the political elite, not “the public”.
Can we wait till after the bubble pops and see what pieces are left on the floor?
Why does the public have a right to expropriate the property if AI companies specifically, as opposed to other types of companies? Just make broad rules that apply to everyone based on abstract principles. I’m fine even with very liberal economic approaches. If we want to raise corporate tax rates to 30%, fine, do that. Want to get creative? Half the equity of every company goes into a public fund. But this case by case, “sure is a nice company you got there” stuff is third-world shit.
The idea is that AI companies expropriated our labor to train their sloppotrons first.
But given that this is Bernie, it's probably a stalking horse for future expropriations from other industries later.
This sounds like the new "thats people's retirement," because if we convert 50% of every AI company into a sovereign wealth fund (which is already a questionable seizure anyway), suddenly it will become politically untenable to do anything that might put that fund in danger, like... regulating anything, or even not bailing out a company thats struggling.
If a 1T company struggles do you really think we (the US) wouldn't bail them out just like the banks and the auto industry?
The quickest way to lose the AI battle to China and reinstate censorship and model political bias is to hand the industry over to the government. This doesn’t just benefit democrats, whatever political party is in power will use them as tools of political means. Not to mention almost certainly pricing won’t act in accordance with free market forces but instead be regulated. See healthcare.
Giving government property rights means taking away your own property rights. It’s zero sum.
I am willing to surrender some property rights for a functional distributive society, pragmatically.
Some.
Most of us are. That's what taxes are, in fact. We have private property rights, and we have to pay taxes. The hard-core anti-tax people try to make that a contradiction; most of the rest of us don't think of it that way, so their "taxation is theft" rhetoric falls flat.
But I think it's important that we keep it at some. Major erosions of private property rights in order "to tax the rich" make me nervous, because I don't want to lose those rights myself. And taking the rights from them, but I get to keep them, seems likely to not be a stable equilibrium.
Everybody thinks that they can get the government to take from them, but not from us.
They're already taking from us, if you work for a paycheck.
Only if it's AGI, which is astronomically impossible.
I don't think LLMs will be the path to AGI, but anything nature can already build (a human mind) is proven to be possible. After that, it is just a science and engineering problem to replicate.
There are those who believe that a human mind is somehow magic and a special exception to the laws of physics, but I am not one of those people.
The public should own half of everything that Senator Sanders owns as well.
Pretty sure he'd be down for that
"Envy was once considered to be one of the seven deadly sins before it became one of the most admired virtues under its new name, social justice." - Thomas Sowell
If we are pulling quotes out of a hat, why not some OG one?
“He who oppresses the poor taunts and insults his Maker, But he who is kind and merciful and gracious to the needy honors Him.” Proverbs 14:31
"Render therefore unto Caesar the things which are Caesar's..."`
Bernard, this may not be the best time. This market is gonna go nuclear...
Can the market implode already, so we can all say told you so.
I'm sorry to say, but this is a losing position, as in one you only adopt when you've already lost and are trying to bargain. It presupposes that the AI companies are unregulateable, and that the only possible avenue of influence/dividend is through ownership. Contrast with the traditional idea that when companies create harm, the government works to stop that harm by default. Or that if these companies actually do succeed at rolling up up the distributed economy into a handful of centralized companies, the government steps in to tax their outsized gains to preserve some semblance of a distributed economy. Furthermore, what does said "ownership" actually do ? If the government is unable to regulate these companies, then it is also unable to reliably exercise a voting interest or insist on receiving dividends - if the companies are this powerful, then whoever actually controls them can always alter the terms and reject the "owners'" demands.
The Public Should Own ALL of SpaceX.
I am repulsed by this because it will obviously be the vehicle through which tax money will be directed into Altman & Co's pockets, but I also understand that they will get bailed out whether the government gets a share or not.
As long as they are voting shares, I don't see an increase in the harm. I'd like to see a legislative framework about how that ownership is handled that allows Congress and regulatory agencies to make decisions restricting how these companies will operate, but without any regard to the constitutional rights of the corporate persons or their owners.
I'm sick of the government arguing with monopolies, then taking dives. I want it to be abundantly clear that government has the ability to restrict these AI utility companies freely (such as their ability to feed on their customers), while still limiting the rights that the state has over the personal use of AI by private individuals. Partial state ownership will make that possible. Hell, let half their boards be publicly elected.
People on HN generally love municipal broadband. For good reasons. It is almost without exception better than any of the national ISPs. Cheaper too.
Municipal broadband is just 100% publicly-owned. That's what that means. When you have a national ISP, you might not get a service at all despite the ISP guaranteeing service in exchange for money from the state they've taken. You get a service that starts at $60 but somehow gets to $140 in a few years unless you do the annual cancel dance and if you do cancel you have no other options anyway. And what are you really paying for? Lobbying to make municipal broadband illegal.
And these same people will defend the status quo because of "property rights". Nobody here is Jeff Bezos. Does it seem like things are going well? Is this a legitimate belief in unfettered property rights? Or is it just that you believe you'll be Jeff Bezos one day so you'll benefit from the status quo?
This is the origin of the quote possibly misattributed to Steinbeck that Americans view themselves as "temporarily embarrassed millionaires".
You don’t have to be Jeff Bezos to believe you’re better off in the long run with strong property rights. To your example: I’ve got 10 gig fiber from a private company. My alternatives are 2 gig fiber from a private company, or 2 gig cable from a private company. I don’t have to be Jeff Bezos to wager that, if the people who ran Baltimore City Public Schools ran my broadband service, I wouldn’t have even 1 gig service.
The fact is, we have tried exactly what you’re suggesting. We don’t even need to get to the brutal communism of China or the Soviet Union. The “social democracy” of India and Bangladesh (where I’m from) left those countries’ economies in the toilet.
The best case scenario for social democracy was the DDR. My wife lived for a year in former eastern germany—in 2001, when the socialist government was as far back in time as Obama’s first term is to today—and folks recall being reasonably happy under socialism. But western germany was still much more prosperous, and it took huge solidarity payments to help the former east germany catch up.
Bernie Sanders. I mean, he's not always wrong, but he's, um, kind of enthusiastic about just taking stuff from those who have it, considerably more than the current understanding of private property (or even taxation) considers acceptable.
>kind of enthusiastic about just taking stuff from those who have it, considerably more than the current understanding of private property (or even taxation) considers acceptable.
Funny how you can use that description for how AI companies used everything for training data.
The obvious counterargument is that the AI labs took virtually all human writing, imagery, music, etc., without regard for licensing or copyright. It's fair to ask what they owe back to the commons they built their models on (and which they are in some sense helping to destroy).
There were long ancestral lines of humans who were very very keen on not redistributing power for many generations.
They were called kings. Cutting their heads off was the best thing to happen to society, ever.
Take the long view. Our particular economic and ideological moment is not worth defending.
This is ahistoric drivel.
China and Russia are simple examples of the fallout of "cutting king's heads".
So awesome in China that they went straight back to a lifetime emperor now.
The French Revolution and its fallout (communism, fascism) killed more humans than any other historic movement, triggering 2 world wars, including numerous genocides.
I think the point is that they took our collective knowledge without asking and are selling it back to us. We should own a substantial portion of it.
(Not sure if this is the right approach, but the general idea seems rather important.)
Yet progressives say that he would be "moderate" in Europe.
Why listen to US progressives' opinions about Europe? Ask Europeans.
...just half?
I would prefer not to
If these companies intend to destroy the fabric of society and jobs and livelihood of everyone, then that leaves us very few choices as a society. This is one of the tamest and most peaceful ones, even if it's just a start. Hopefully Sammy and friends choose wisely.
I would like to see the Big A.I. companies raped and pillaged by the federal government.
That is actually not that bad of an idea huh