49 comments

[ 3.4 ms ] story [ 55.1 ms ] thread
Does OpenAI not have mandatory compliance training which forbids them from giving any government a bribe as a condition of doing business?
I'd settle for Altman and his ilk paying a proper progressive income tax.
This is how they will secure eventual bailout.
Isn't this what people wanted? The public to have access to the gains of these companies?

I don't want public ownership of any private companies but this seems to be what slopulism leads us to

Wait, how does the public benefit from the gains of Open AI with this scheme?
Interesting. But if Sam really believes that the US public should share in the benefits of AI surely the number should be 50% not 5%.
(comment deleted)
Taxes are theft. This is absurd.

There’s zero reason to bail these ding dongs out. Their entire business proposition has been to keep warm by incinerating fresh cash.

How is it a bailout if the government just takes 5% of the shares?
No, taxes are taxes and theft is theft.

Taxes are, in the case of a country like the United States, a legally sanctioned, representative approved (indirectly voter approved) means of supporting the needs of the nation/state as a whole.

Theft is distinguished from say, gifting money by consent. In a democracy mechanisms exist for our representatives to make decisions according to the consent of the people, to an extent.

You presumably live in a democracy so you presumably necessarily agree to abide by the laws of your nation which involve being governed by your democratically elected representatives.

> Altman has also reportedly spoken with the Democratic senator Bernie Sanders in recent weeks. The senator has been pushing for the creation of a sovereign wealth fund overseen by an independent commission and financed through a one-time 50% tax on the stock of the biggest AI companies.

Bernie shooting for the moon here

I know OpenAI has delayed their IPO by a year (i.e. not publicly traded, yet; so: no dividends), but wouldn't it be better for the Government/bottom 95% if instead of taking ownership, they taxed all tech-related stocks 5% every time they're traded – this is a perpetual stream of income, and would likely reduce speculative short-term trading...

source: middle-aged electrician, owns a little stock (and would happily pay trading taxes, either in/out/both); know nothing unrelated to copper; eats crayons

(comment deleted)
But governments cannot be trusted, better give it to an independent entity like they did with the Qatari jet
This feels a bit off.. How is the government supposed to be able to regulate them impartially when they're literally invested in them.

What if a competitive startup startup starts to really take away from OpenAi's profits and then all of a sudden requires some approval for merger with Anthropic for example, I don't know if I would trust the government to be fair in their decision here.

Leaving aside the potential for letting the government(tax payers) hold the bag if there is a collapse.

>This feels a bit off.. How is the government supposed to be able to regulate them impartially when they're literally invested in them.

Was the government impartial to begin with? Or were they stomping things and handing out favorable treatment based on the political whims of the minute?

Seems to me like "hey buddy you own some (but not too much) of this too so play the long game" provides somewhat of a counter incentive.

Will it work? Will it do the opposite and make things worse? Heck if I know.

Wouldn’t the solution to that be for the government to also demand a 5% stake in every AI company? Along the lines of what Sanders wants to do?
You’re completely correct, of course, but I also think it’s worth remembering that the current regulatory environment is also made up of people, so the opportunity for corruption is already there. It does seem preferable to at least have the government putting its thumb on the scales to benefit the country at large, versus money under the table that benefits politicians and bureaucrats.
This is quite literally government control of the means of production, which is the central tenet of the economic philosophy that the right repeatedly claims is their greatest mortal fear: socialism.
Seems to be a very bad mechanism to ensure democratic control of the technology. There must be better ways, even naively assuming that OpenAI is somehow genuine about wanting to broadly share its stake in the future.
what about the behavior of the US government currently has you thinking it would be democratic?
>The proposal would also involve other US AI companies giving a similar stake to the government, the FT reported, although it is not clear yet whether companies such as Anthropic, Google and Meta would agree to the plan.

I can't see Google or Meta shareholders agreeing to this? That said, Google, Meta, and SpaceX are all still founder-controlled using supervoting shares.

In a proper capitalist society the government defines the rules of the market, aligned with the interests of several parties, and then companies compete within that well regulated and fair environment. The only incentive the government should have is to grow the entire market, so that they can collect more tax. There might be minor exceptions to protect key industries like food production or defense, but these should be a small as possible to ensure healthy competition.

It’s something entirely different when the government starts taking a stake in individual companies instead of the market as a whole. This can easily bias the government to pick OpenAI for certain contracts, or enact laws that benefit OpenAI more than its competitors. It reduces competition which hurts the overall economy, and it is an obvious vector for corruption which hurts the efficiency of the government.

It’s great if we can leverage AI to design the next great government system. A 5% stake feels more like a bribe to help push through some of these datacenter projects and enact friendly laws.

AI was already too big to fail, this just locks it in. Talk about a trojan horse.
This smells like an incoming bailout
This feels more communist than communist China. They typically take about 1% in Golden Shares that give them a board seat.
It seems that the US may be in a process of signing itself up for many of the drawbacks of Chinese-style state capitalism (regulatory conflicts of interest, opportunities for politicians to rent-seek) with stakes small enough that the taxpayer will see little real economic benefit.
This.

Though I’d say US is speeding with reimplementation of ruzzian style oligarchy (already happening with open corruption between government and broligarchs) with flavor of Chinese capitalism (this article).

I may be naïve or completely uninformed, but given the federal government’s vast resources, including supercomputers, national laboratories, the NSA, and many talented employees, why does the federal government need OpenAI or Anthropic for that matter when it has the resources to build its own LLM, even one exclusive for government use? The federal government has a long history of technical feats, such as the atomic bomb, the ARPANET, and the moon landing. Couldn’t it build its own state-of-the-art LLM?
All of these examples are Cold War era or before, when defense and aerospace was a much more collectivist venture with far less private sector involvement.

After the Cold War, McKinsey and big 4 consultancies are working in every part of government, ostensibly to make government as efficient as the private sector. The NSA's surveillance program wouldn't have seen the light of day had Booz Allen not had a contract with them.

It would not be without problems and mistakes in execution over time, but I think the US and our NATO allies should nationalize AI research and development in a sweeping manner, and NATO membership ought be revised to hinge on that.

In the US, for example, all intellectual property of OpenAI, Anthropic, et al. would become public domain through custody of the Federal government, probably in an expansion of the NSF. All AI research and development would be required by law to be done in the open: open source code, transparent training data, reproducible models.

Oh, you have NATO allies now? And in the future what… “we leave NATO if you don’t buy enough “tokens”? Go and build a wall.
Donald won’t be POTUS too much longer, hang in there, less than 3 years to go.
With the USA and Israel tightening their intelligence agencies / secret service exchange, and now pulling in OpenAI -- that's a very effective strategy to exercise more worldly dominance
Some advantages this gives OpenAI:

- Altman takes away Musk's power as Trump's favourite tech friend.

- the government won't punish OpenAI too hard, because it makes money when OpenAI does well.

- the government can look at the user's data without any problems.

- OpenAI's competitors are forced to give the government a share in their companies too.

- when OpenAI sells shares to the public, investors will trust it more because the government is involved.

- every American could get a yearly check from OpenAI's profits, so voters will protect OpenAI.

- Sam Altman becomes friends with politicians from all sides, so nobody dares to investigate him.