Live data from Hacker News

AI Fortunes Are Reviving an Old Debate About Private Power

ai-updates.net

31–39 of 39 posts

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#31

Earlier quoted context omitted.

Do the math on that one. There are 8 billion people. OpenAI revenue -- not profit -- is ~$25B, which is ~$3/year. That's if they have zero expenses for engineers or land or equipment or electricity, which isn't the case now and never will be. Their current profit is a number of substantial magnitude with a minus sign in front of it. The funny thing about all of this is that it's basically taking revenue from the majo…

Its obviously impractical for a number of reasons, I just find it frustrating because based on the last century of IP law that should be what's happening. The big corporations won the fight on file-sharing so they ought to at least be accountable to their own legal arguments.

> Its obviously impractical for a number of reasons, I just find it frustrating because based on the last century of IP law that should be what's happening.

Eh. Most LLM output is significantly transformative. The challenge here is that some of the output is similar enough to a work in the training data that it could reasonably be considered derivative. And evaluating whether it is when you have both of them in front of you isn't even the hard part.

The real problem, from the perspective of LLM users, is that you don't know when it's happening because you haven't seen the original work to notice how similar it is to a particular output, and now maybe you have problems if you start making copies of it.

> The big corporations won the fight on file-sharing so they ought to at least be accountable to their own legal arguments.

Those big corporations are the likes of Sony and Comcast. They're the ones who would prefer to be able to sue OpenAI. But that's also the thing which is trash, because if the money went only to them then it's just propping up the incumbents while spitting on small independent content creators who get nothing. Whereas if you actually included everybody then we're back to ~$3/year and it's not even worth the administrative costs compared to just having them pay ordinary taxes on their profits to the extent that they eventually have any.

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#32

Earlier quoted context omitted.

> Taxes don't really affect entity size. A progressive tax would, no? You could explicitly tax larger entities more.

Using graduated rates for corporations isn't used because it doesn't work. The formal entity size is disconnected from the business size. How many entities is "Google"? Is it one or is Search one company and Android another and YouTube another and so on? If they need to get the size of each unit below some threshold they can just file more LLCs. Just use the size of the parent company, you say? Okay, which is the par…

> Just use the size of the parent company, you say? Okay, which is the parent company, Google or Vanguard?

Could you maybe tie it back to majority control? Alphabet is one company because it actually has a controlling stake over Google, YouTube, Android, etc. Whereas Vanguard is just a minority shareholder in Google.

> And even if you could do it, you would then create a perverse incentive for the government, because then breaking up those large companies into smaller ones would reduce tax revenue. When you should just break them up to begin with and actually solve the problem.

The idea is that you wouldn't need to government to actively break up companies, they'd do it themselves (or get outcompeted). I will admit that this is very much unproven theory though (but I'd certainly be interested in people investigating the theory more formally if nothing else).

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#34

Earlier quoted context omitted.

Its obviously impractical for a number of reasons, I just find it frustrating because based on the last century of IP law that should be what's happening. The big corporations won the fight on file-sharing so they ought to at least be accountable to their own legal arguments.

> Its obviously impractical for a number of reasons, I just find it frustrating because based on the last century of IP law that should be what's happening. Eh. Most LLM output is significantly transformative. The challenge here is that some of the output is similar enough to a work in the training data that it could reasonably be considered derivative. And evaluating whether it is when you have both of them in front…

Microsoft is one of them too and they've been one of the primary funders of OpenAI. And that's to say nothing of the many recent mergers between software corporations and film studios.

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#35
post #4

Earlier quoted context omitted.

More regulation. Specifically requiring worker-owned means of production: re-organize the AI model companies as mutuals and mandate (by law) that their only customers can be individual-contributor-level workers or worker unions. Ditto with robotics.

You realize that the folks with these fortunes are the workers who built the companies, right? If there's one place that's "worked-owned" it's a startup. And the VCs who also ended up with stocks did quite a bit of work along the way-- as well as investing capital. We're not talking about widows with a pension who are just interested in dividends. This isn't some utility.

So everyone past the first dozen people didn't help the company?

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#36

Earlier quoted context omitted.

Using graduated rates for corporations isn't used because it doesn't work. The formal entity size is disconnected from the business size. How many entities is "Google"? Is it one or is Search one company and Android another and YouTube another and so on? If they need to get the size of each unit below some threshold they can just file more LLCs. Just use the size of the parent company, you say? Okay, which is the par…

> Just use the size of the parent company, you say? Okay, which is the parent company, Google or Vanguard? Could you maybe tie it back to majority control? Alphabet is one company because it actually has a controlling stake over Google, YouTube, Android, etc. Whereas Vanguard is just a minority shareholder in Google. > And even if you could do it, you would then create a perverse incentive for the government, because…

> Could you maybe tie it back to majority control?

Then you get three corporations that each own a one third stake in all the things that constitute Alphabet and a variety of investment instruments that allow anyone to buy shares in all three of them at the same time.

Things like that wouldn't work if they had an actual individual majority shareholder, but the larger the company is the less likely that's the case.

> The idea is that you wouldn't need to government to actively break up companies, they'd do it themselves (or get outcompeted).

And then we're back to the original point -- business expenses are deductions so raising the tax rate actually gives them the incentive to spend money expanding rather than using it to pay dividends, so if anything it does the opposite of that.

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#37
post #4

Earlier quoted context omitted.

More regulation. Specifically requiring worker-owned means of production: re-organize the AI model companies as mutuals and mandate (by law) that their only customers can be individual-contributor-level workers or worker unions. Ditto with robotics.

You realize that the folks with these fortunes are the workers who built the companies, right? If there's one place that's "worked-owned" it's a startup. And the VCs who also ended up with stocks did quite a bit of work along the way-- as well as investing capital. We're not talking about widows with a pension who are just interested in dividends. This isn't some utility.

> If there's one place that's "worked-owned" it's a startup.

Sweetie, it is called cooperative, not startup.

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#38
post #4

Earlier quoted context omitted.

More regulation. Specifically requiring worker-owned means of production: re-organize the AI model companies as mutuals and mandate (by law) that their only customers can be individual-contributor-level workers or worker unions. Ditto with robotics.

This doesn't solve any of the problems. It's not a 19th century factory running on leather-belt-harnessed steam engines. There aren't 10,000 workers. There are at most a few dozen "individual-contributor-level workers". You're still concentrating all of that onto a really small number of people. Who, incidentally, have demonstrated time and again that their behavior tends to imitate the big powerful robber barons' be…

> You've really lost the plot.

I think you did.

> There aren't 10,000 workers. There are at most a few dozen "individual-contributor-level workers". You're still concentrating all of that onto a really small number of people. Who, incidentally, have demonstrated time and again that their behavior tends to imitate the big powerful robber barons' behavior of old

You're not clear here, but it sounds like you're talking about employees of OpenAI, Anthropic, etc.

But then you misunderstood me. I was talking about the workers who use the tools those companies make.

In the plan in my comment

* OpenAI, Anthropic, employees would be working for a mutual, as in a customer-owned company. Those "few dozen" robber barons wannabes wouldn't be getting any big payout.

* The workers I was referring to were all the employees who use AI tools to perform a job. They would be the only customers of OpenAI, Anthropic, etc. in their personal capacities. Their employers would not be allowed to buy those tools directly, they'd have to hire a worker to use them, who could then capture a lot of the value instead of the company.

Re: AI Fortunes Are Reviving an Old Debate About Private Power

#39
post #28
post #21

Earlier quoted context omitted.

it seems to be working quite well in the present, if you look at china

...after they abandoned communism ("let some people and some regions get rich first") and liberalized their economy?

They abandoned central planning, not state control of the economy.

Unfortunately, their priority is not worker welfare, but more like national greatness.

And they're performing way better than "capitalist" economies, which all look downright dysfunctional in pretty serious ways (e.g. selling off strategic manufacturing capacity to make a quick buck).

Post reply on HN