I’m confused by this news story and the response here. No one seems to understand OpenAI’s corporate structure or non profits at all.
My understanding: OpenAI follows the same model Mozilla does. The nonprofit has owned a for-profit corporation called OpenAI Global, LLC that pays taxes on any revenue that isn’t directly in service of their mission (in a very narrow sense based on judicial precedent) since 2019 [1]. In Mozilla’s case that’s the revenue they make from making Google the default search engine and in OpenAI’s case that’s all their ChatGPT and API revenue. The vast majority (all?) engineers work for the for-profit and always have. The vast majority (all?) revenue goes through the for-profit which pays taxes on that revenue minus the usual business deductions. The only money that goes to the nonprofit tax-free are donations. Everything else is taxed at least once at the for-profit corporation. Almost every nonprofit that raises revenue outside of donations has to be structured more or less this way to pay taxes. They don’t get to just take any taxable revenue stream and declare it tax free.
All OpenAI is doing here is decoupling ownership of the for-profit entity from the nonprofit. They’re allowing the for profit to create more shares and distribute them to entities other than the non-profit. Or am I completely misinformed?
It's about the narrative they tried to create. The spin. It doesn't matter much if they were technically behaving as a for-profit entity previously. What matters is that they wanted the public (and likely, their talent) to _think_ that they weren't even interested in making a profit as this would be a philosophical threat to the notion of any sort of impartial or even hopefully benevolent originator of AGI (a goal which is laid plainly in their mission statement).
As you've realized, this should have been (and was) obvious for a long time. But that doesn't make it any less hypocritical or headline worthy.
>What matters is that they wanted the public (and likely, their talent) to _think_ that they weren't even interested in making a profit as this would be a philosophical threat to the notion of any sort of impartial or even hopefully benevolent originator of AGI (a goal which is laid plainly in their mission statement)
And now they want to cast off any pretense of that former altruistic yolk now that they have a new, better raison d'etre to attract talent: making absolutely unparalleled stacks of cash.
Occam's razor: I think Sam's personal narrative is the correct one. He built a non-profit that took off in a way that he didn't expect it and now a for-profit is the best way to run the lightning they've caught.
In terms of profit, AFAICT, Sam doesn't have designs on building extra large yachts and his own space agency but what he wants is to be the one at the stead of building what he considers is world-changing tech. One could rationally call this power-hungry but one could also rationally call this just helicopter parenting of a tech you've helped built. And for that a for-profit that is allowed to maximize profits to re-invest in the tech is the optimal setup (esp if all the competitors are doing the same)
Is this a different org than when it started? Yes. Was this a dupe from the beginning? I don't think so.
"But why can't he have a more worldly-aligned board looking over his shoulder?"
Because we live in California and have a bad taste for governance by committee or worse: governance by constant non-representative democracy (see: Housing).
If this now completely comes off the wheels, I still think Congressional action can be a stopgap, but atleast for now, this restructure makes sense to me.
It isnt a tax thing or a money thing, its a control and governance thing.
The board of the non-profit fired Altman and then Altman (& MS) rebelled, retook control, & gutted the non-profit board. Then, they stacked the new non-profit board with Altman/MS loyalists and now they're discharging the non-profit.
It's entirely about control. The board has a legally enforceable duty to its charter. That charter is the problem Altman is solving.
The problem is that OpenAI calls itself OpenAI when it's completely sealed off, and calls itself a non-profit when, as you say, almost everything about is for profit. Basically they're whitewashing their image as an organization with noble goals when it's simply yet another profit motivated company. It's fine if that's what they are and want to be, but the lies are bothersome.
There's a now-quintessential HN post format, "Poster criticizing X don't seem to under [spray of random details about X that don't refute the criticism - just cast posts as ignorant]".
In this case, Mozilla as a non-profit owning a for-profit manages to more or less fulfill the non-profit's mission (maintaining an open, alternative browser). OpenAI has been in a hurry to abandon it's non-profit mission for a while and the complex details of its structure doesn't change this.
But unfortunately charities and not-for-profits putting their core business into a company, and then eventually selling it off is not unprecedented. For example, The Raspberry Pi Foundation was a not-for-profit organisation around Raspberry Pi. They formed a LLC for their commercial operations, then gradually sold it off before eventually announcing an IPO: https://www.raspberrypi.org/blog/what-would-an-ipo-mean-for-....
I think it is terrible that not-for-profits are just being used as incubators for companies that eventually take the core mission and stop primarily serving the public interest.
There are of course other examples of charities or not-for-profits that put part of their core operations in a company and don't sell out, instead retaining 100% ownership - for example Mozilla. However, I think there should be some better way for impactful not-for-profits to have some revenue generating aspects in line with their mission (offset by allowing temporary surplus to cover future expenses, or by other expenses).
This might be unpopular, but I don't think Mozilla's behavior is inherently good[1]. And even if it has, I don't think it should be used as a litmus test for other corporate behavior. Every company should stand on its own against scrutiny.
How is it possible to make tax free "donations" for profit making applications? You seem to imply there is nothing nefarious about the setup. Except the non-profit designation doesn't actually perform no social services, instead stand as a business structure to skirt taxation. Change my mind
> How is it possible to make tax free "donations" for profit making applications?
The nonprofit invests the tax free donations into the for-profit. It gets to keep its equity just like any other investor and as long as that equity held by the non-profit, there's no taxable event. If the nonprofit sells its share - since it was closely involved in the creation and management of the for-profit as an active investment - it becomes a taxable event under the unrelated business income tax rules. Until that time, the only "profit making applications" like ChatGPT and the API are run by the for-profit, which - I repeat - pays its taxes.
I genuinely don't understand why people think they've skirted taxation except out of sheer ignorance of how non-profits actually work. A 501(c)(3) is not some magic Monopoly "get-out-of-tax-free" card and the IRS isn't stupid, there's a ton of rules for tax exemption. They'd have a much easier time with tax avoidance if they were an actual for profit corporation with billions of dollars because GAAP rules are a lot more forgiving than non-profit regulations.
Right now, OpenAI, Inc. (California non-profit, lets say the charity) is the sole controlling shareholder of OpenAI Global LLC (Delaware for-profit, lets say the company). So, just to start off with the big picture: the whole enterprise was ultimately under the sole control of the non-profit board, who in turn was obligated to operate in furtherance of "charitable public benefit". This is what the linked article means by "significant governance changes happening behind the scenes," which should hopefully convince you that I'm not making this part up.
To get really specific, this change would mean that they'd no longer be obligated to comply with these CA laws:
The term charitable is used in its generally accepted legal sense and includes relief of the poor, the distressed, or the underprivileged; advancement of religion; advancement of education or science; erecting or maintaining public buildings, monuments, or works; lessening the burdens of government; lessening neighborhood tensions; eliminating prejudice and discrimination; defending human and civil rights secured by law; and combating community deterioration and juvenile delinquency.
... The organization must not be organized or operated for the benefit of private interests, and no part of a section 501(c)(3) organization's net earnings may inure to the benefit of any private shareholder or individual.
I'm personally dubious about the specific claims you made about revenue, but that's hard to find info on, and not the core issue. The core issue was that they were obligated (not just, like, promising) to direct all of their actions towards the public good, and they're abandoning that to instead profit a few shareholders, taking the fruit of their financial and social status with them. They've been making some money for some investors (or losses...), but the non-profit was, legally speaking, only allowed to permit that as a means to an end.
Naturally, this makes it very hard to explain how the nonprofit could give up basically all of its control without breaking its obligations.
All the above covers "why does it feel unfair for a non-profit entity to gift its assets to a for-profit", but I'll briefly cover the more specific issue of "why does it feel unfair for OpenAI in particular to abandon their founding mission". The answer is simple: they explicitly warned us that for-profit pursuit of AGI is dangerous, potentially leading to catastrophic tragedies involving unrelated members of the global public. We're talking "mass casualty event"-level stuff here, and it's really troubling to see the exact same organization change their mind now that they're in a dominant position. Here's the relevant quotes from their founding documents:
OpenAI is a non-profit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. Since our research is free from financial obligations, we can better focus on a positive human impact...
It’s hard to fathom how much human-level AI could benefit society, and it’s equally hard to imagine how much it could damage society if built or used incorrectly. Because of AI’s surprising history, it’s hard to predict when human-level AI might come within reach. When it does, it’ll be important to have a leading research institution which can prioritize a good outcome for all over its own self-interest.
We commit to use any influence we obtain over AGI’s deployment to ensure it is used for the benefit of all, and to avoid enabling uses of AI or AGI that harm humanity or unduly concentrate power. Our primary fiduciary duty is to humanity...
We are concerned about late-stage AGI development becoming a competitive race without time for adequate safety precautions. Therefore, if a value-aligned, safety-conscious project comes close to building AGI before we do, we commit to stop competing with and start assisting this project. We will work out specifics in case-by-case agreements, but a typical triggering condition might be “a better-than-even chance of success in the next two years.”
Sorry for the long reply, and I appreciate the polite + well-researched question! As you can probably guess, this move makes me a little offended and very anxious. For more, look at the posts from the leaders who quit in protest yesterday, namely their CTO.
> I'm personally dubious about the specific claims you made about revenue, but that's hard to find info on, and not the core issue. The core issue was that they were obligated (not just, like, promising) to direct all of their actions towards the public good, and they're abandoning that to instead profit a few shareholders, taking the fruit of their financial and social status with them. They've been making some money for some investors (or losses...), but the non-profit was, legally speaking, only allowed to permit that as a means to an end.
Look at your OpenAI invoices. They're paid to OpenAI LLC, not OpenAI Inc. I can't find confirmation on openai.com what the exact relationship between OpenAI Global LLC and OpenAI LLC is but the former is on their "Our Structure" page and the latter is in their data processing addendum so it's probably the subsidiary in charge of operating the services while Global does training and licenses it downstream. OpenAI Global was the one that made that big $10 billion deal with Microsoft
That obligation is why they had to spin off a for-profit corporation. Courts are very strict in their interpretation of what "unrelated business income" is and the for-profit LLC protects the non-profit's tax exempt status.
> "why does it feel unfair for a non-profit entity to gift its assets to a for-profit"
What assets were gifted, exactly? They created the for-profit shortly after GPT2 (in 2019) and as far as I can tell that's the organization that has developed the IP that's actually making money now.
I honestly don't understand how this isn't in the interest of the nonprofit's mission. It's currently a useless appendage and will never have any real power or resources until either OpenAI is in the black and sending profit up to it, or they can sell OpenAI shares. I don't think the charity has any more claim over GPT4 than Google does, having invented transformers.
If this next round of funding goes through at $100-150 billion valuation, OpenAI Inc will probably be (on paper at least) the second wealthiest charity on the planet after the Novo Nordisk Foundation. This restructuring opens the way for the nonprofit to sell its shares and it's going to be a hell of a lot of money to dedicate towards their mission - instead of watching its subsidiary burn billions of dollars with no end in sight.
What counts as “related” to the charity’s mission is fuzzy but in practice the courts have been rather strict. They don’t have to form for-profit subsidiaries to pay those taxes but it helps to derisk the parent because potential penalties include loss of nonprofit status.
For example, the nonprofit Metropolitan Museum of Modern Art has a for-profit subsidiary that operates the gift shop. National Geographic Society has National Geographic Partners which actually owns the TV channel and publishes the magazine. Harvard and Stanford have the Harvard Management Company and Stanford Management Company to manage their endowments respectively. The Smithsonian Institute has Smithsonian Enterprises. Mayo Clinic => Mayo Clinic Ventures. Even the state owned University of California regents have a bunch of for-profit subsidiaries.
Can anybody explain how this actually works? What happens to all of the non-profit's assets? They can't just give it away for investors to own.
The non-profit could maybe sell its assets to investors, but then what would it do with the money?
I'm sure OpenAI has an explanation, but I really want to hear more details. In the most simple analysis of "non-profit becomes for-profit", there's really no way to square it other than non-profit assets (generated through donations) just being handed to somebody for private ownership.
If the assets were sold to the for profit at a fair price I could see this being legal (even if it shouldn't be). At least in that case the value generated by the non-profit tax free would stay locked up in non-profit land.
The biggest problem with this is that there's basically no chance that the sale price of the non-profit assets is going to be $150 billion, which means that whatever the gap is between the valuation of the assets and the valuation of the company is pure profit derived from the gutting of the non-profit.
If this is allowed, every startup founded from now on should rationally do the same thing. No taxes while growing, then convert to for profit right before you exit.
It’s pretty great if you can manage to have the parent be 501(c)(3). Have all the early investors “donate” 90% of their investment to the 501(c)(3) and invest 10% in the for-profit subsidiary the old-fashioned way. They get a tax deduction, and the parent owns 90% of the subsidiary. Later on, if the business is successful, the parent cashes out at the lowest possible valuation they can pull off with a mostly straight face, and all the investors in the subsidiary end up owning their shares, pro rata, with no dilution from the parent. The parent keeps a bit of cash (and can use it for some other purpose).
Of course the investors do end up owning their shares at a lower basis than they would otherwise, and they end up a bit diluted compared to a straightforward investment, but the investors seem likely to more than make up for this by donating appreciated securities to the 501(c)(3) and by deferring or even completely avoiding the capital gains tax on their for-profit shares.
Obviously everyone needs to consult their lawyer about the probability of civil and/or criminal penalties.
I haven't seen any details, but isn't this a pretty straightforward way of doing it? The non-profit has had majority ownership of the for-profit subsidiary since 2019. The already-for-profit subsidiary has owned all the ChatGPT IP, all the recent models, all the employee relationships, etc etc.
The cleanest way for this to work is the for-profit to just sell more shares at the $150B valuation, diluting the non-profit entity below majority ownership. The for-profit board, which the non-profit could still probably have multiple seats on, would control the real asset, the non-profit would still exist and hold many tens of billions of value. It could further sell its shares in the non-profit and use the proceeds in a way consistent with its mission.
They wouldn't even have to sell that much - I am pretty sure the mega-fundrasing rounds from Microsoft etc brought the non-profit's ownership to just north of 50% anyway.
I don't see how this wouldn't be above board, it's how I assumed it was going to work. It would indeed mean that the entity that controls ChatGPT would now be answerable to shareholders, a majority of which would be profit seeking and a minority of which would be the non-profit with its mission, but non-profits are allowed to invest in for-profits and then sell those shares; all the calls for prosecutions etc seems just like an internet pitchfork mob to me.
You actually don't even need to sell them. Just sign an exclusive, non-revocable license agreement.
Practically the same as selling, but technically not. Non-profit still gets to live up to it's original mission, on paper, but doesn't really do anything internally.
> there's basically no chance that the sale price of the non-profit assets is going to be $150 billion
The non-profit’s asset is the value of OpenAI minus the value of its profit-participation units, i.e. the value of the option above the profit cap. Thus, it must be less than the value of OpenAI. The non-profit owns an option, not OpenAI.
I've actually worked through a similar situation for a prior startup. We were initially funded by a large, hospital system (non-profit) who wanted to foster innovation and a startup mentality. After getting started, it became clear that it was effectively impossible for us to operate like a startup under a non-profit. Namely, traditional funding routes were neigh impossible and the hospital didn't want direct ownership.
It's been many years, but the plan was essentially this:
* The original, non-profit would still exist
* A new, for-profit venture would be created, with the hospital having a board seat and 5% ownership. Can't remember the exact reason behind 5%. I think it was a threshold for certain things becoming a liability for the hospital as they'd be considered "active" owners above 5%. I think this was a healthcare specific issue and unlikely to affect non-profits in other fields.
* The for-profit venture would seek, traditional VC funding. Though, the target investors were primarily in the healthcare space.
* As part of funding, the non-profit would grant exclusive, irrevocable rights of it's IP to that for-profit venture.
* Everyone working for the "startup" would need to sign a new employment contract with the for-profit.
* Viola! You've converted a non-profit into a for-profit business.
I'm fuzzy on a lot of details, but that was the high level architecture of the setup. It's one of those things where the lawyers earn a BOAT LOAD of money to make sure every technicality is accounted for, but everything is just a technicality. The practical outcome is you've converted a non-profit to a for-profit business.
Obviously, this can't happen without the non-profit's approval. From the outside, it seems that Sam has been working internally to align leadership and the board with this outcome.
-----
What will be interesting is how the employees are treated. These types of maneuvers are often an opportunity for companies to drop employees, renegotiate more favorable terms, and reset vesting schedules.
> * As part of funding, the non-profit would grant exclusive, irrevocable rights of it's IP to that for-profit venture.
This is the part that should land people literally in jail. A non-profit should not be able to donate its assets to a for-profit, and if it's the same people running both companies, those people must be sent to prison for tax evasion. There is no other way to preserve the integrity of the "non-profit" status with this giant loophole.
After the non-profit sells its assets, it would either donate the proceeds in a way that would be aligned with the original mission, or continue to exist as a bag of cash, basically.
It seems incredibly convenient that a non-profit's leaders can say "I want equity in a for-profit company, so we will sell our assets to investors (who will hire me) and pass off the proceeds to some other non-profit org run by some other schmuck. This is in the public interest."
Every answer moving forward now will contain embedded ads for Sephora, or something completely unrelated to your prompt...
That money will go into the pockets of a small group of people that claim they own shares in the company... Then the company will pull more people in who invest in it, and they'll all get profits based on continually rising monthly membership fees, for an app that stole content from social media posts and historical documents others have written without issuing credit nor compensating them.
Maybe it’s a hint that the tax rate for small and medium companies should be reduced (or other non tax laws modified based on company size), to copy the advantages of this nonprofit to profit conversion, while taxes for large companies should be increased. It would maybe help make competition more fair and make survival easier for startups.
This is actually a good idea. I say we go even further and stop wasting so much money cleaning up after companies - get rid of the entire legal entity known as a corporation and let investors shoulder the full liability that comes with their ownership stake.
My expertise is in NFP hospitals. Generally, when they convert for for-profit part of that deal is the creation of a foundation funded with assets that are ostensibly to advance the original not for profit mission.
The nonprofit gives all its ownership rights to the for-profit in return for equity. The nonprofit is free to hold the equity and maintain control or sell the equity and use the proceeds for actual charitable purposes.
As long as the money doesn't go into someone's pocket, it's all good (except that Sam Altman is also getting equity but I assume they found a way to justify that.)
OpenAI will eventually be forced to convert from a public charity to a private foundation and will be forced to give away a certain percentage of their assets every year so this solves that problem also.
The significant asset isn't equity, it's control. 51% is much more valuable than 49% when the owned organization is supposedly working towards technology that will completely change how the world works.
Difficult to see how these two stories aren't related.
OpenAI has been one of the most insane business stories in years. I can't wait to read a full book about it that isn't written by either Walter Isaacson or Michael Lewis.
I've only read Michael Lewis's "Liars Poker" which I enjoyed, but perhaps that sort of treatment of OpenAI would make it into more of a drama (which also seems to be somewhat true) and gloss over what the key players were really thinking which is what would really be interesting.
Don't stop at robots.txt blocking. Look through your access logs, and you'll likely find a few IPs generating a huge amount of traffic. Look them up via "whois," then block the entire IP range if it seems like a bot host. There's no reason for cloud providers to browse my personal site, so if they host crawlers, they get blocked.
I wonder how the AI/copyright arguments will play out in court.
"If I read your book and I have a photographic memory and can recall any paragraph do I need to pay you a licensing fee?"
"If I go through your library and count all the times that 'the' is adjacent to 'end' do I need to get your permission to then tell that number to other people?"
Cases where we see the absurdity of copyright in its current form. But either we have it for everyone, including OpenAI, or for no one. Or are perhaps some more equal than others before the law?
We are booing Altman because his bait and switch feels unethical, but many of us saw it coming from a mile away, how does he make this transition to take advantage of the financial system so smoothly? Is there no legal guard for such maneuver, or is he just an insanely good player to circumvent all of them in plain view?
Just cobbling together articles I've read... Once the board failed to fire Altman, this transition was a done-deal, it was just a matter of when. Before the board fired Altman, he had been engaging in a more covert effort to take control - putting together personal attacks on various board members intended to drive them out and hiring people with a similar attitude to and loyalty to himself. When the board fired him, they didn't give any clear reasons and that's most mysterious part of the situation. Apparently they were all concerned about his campaign to take control of the board but they each had slightly different reasons for agreeing. I'd further speculate that the board thought in lawyer-advice terms - say as little as possible to avoid a retaliatory lawsuit.
But the board's lack of communication apparently allowed Altman to demonstrate he was more important to the organization than the formal/legal structure, 90% signed intents to quit and the board backed down. It seemed that Altman simply represented the attitude of the many Silicon Valley tech-people - once you have a chance of money, don't hold back, do everything you can to make it.
It has been running on an honor code, that someone pulling off something so slimy as to funnel money meant for non-profits would just get shunned by society and business. Yet here we are.
Ironically the one person with resources fighting it in a tangible way, even if for spite, is Elon Musk.
Converting to a for-profit changes the tax status of donations. It also voids plausibility for Fair Use exemptions.
I can see large copyright holders lining up with takedowns demanding they revise their originating datasets since there will now be a clear-cut commercial use without license.
I hope I can join in, as a consumer, because there’s a difference between using the IP I contribute to conversations for a non-profit and a commercial enterprise.
I suspect that if you have ever posted copyrightable material online, you will have valid cause to sue them, as they very obviously have incorporated your work for commercial gain. That said, I unfortunately put your chances of winning in court very low.
Any reasonable court would see right through “well we trained it for the public good, but only we can use it directly”. That’s not really a legal loophole as much as an arrogant ploy. IMO, IANAL
> It also voids plausibility for Fair Use exemptions. I can see large copyright holders lining up with takedowns
I thought so for a moment but then again Meta, Anthropic (I just checked and they have a "for profit and public benefit" status whatever that means), Google or that Musk's thing aren't non-profits, are they ? There are lawsuits in motion for sure but with how it stands today I think ai gets off the hook.
This is the most important question, IMO! ChatGPT says that employees and donors would have to show that they were defrauded (lied to), which IMO wouldn’t exactly be hard given the founding documents. But the real power falls to the government, both state (Delaware presumably…?) and federal. It mentions the IRS, but AFAIU the DoJ itself could easily bring litigation based on defrauding the government. Hell, maybe throw the SEC in there!
In a normal situation, the primary people with standing to prevent such a move would be the board members of the non-profit, which makes sense. Luckily for Sam, the employees helped kick out all the dissenters a long time ago.
Genuinely curious because I have no idea how any of this works...
Would the founding documents actually count as proof of a lie? I feel like the defense could easily make the argument that the documents accurately represented their intent at the time, but as time went on they found that it made more sense to change.
It seems like, if the founding documents were to be proof of a lie, you'd have to have corresponding proof that the documents were intentionally written to mislead people.
In the US standing is a specific legal concept about whether you have a valid reason/role to bring up a particular issue. For example most of Donald Trump’s lawsuits around the 2020 election were rejected for a lack of standing rather than on merit (whether the case is any good).
So are they going to give elon equity? He donated millions to the non profit and now they are going to turn around and turn the company into a for-profit based on the work done with that capital.
Elon has allegedly refused equity in OpenAI. He seems to want it to go back to its original mission (which isn't going to happen) or die (which isn't going to happen).
Given that Musk was already worried about this and has a legal team the size of a small army, one would expect that any conditions he wanted applied to the donation would have been made at the time.
he doesn’t need the money, for one. he missed out and didn’t control it, and now he’s jealous that it took off. oh well, world’s richest man and smallest violin.
Readit News
My understanding: OpenAI follows the same model Mozilla does. The nonprofit has owned a for-profit corporation called OpenAI Global, LLC that pays taxes on any revenue that isn’t directly in service of their mission (in a very narrow sense based on judicial precedent) since 2019 [1]. In Mozilla’s case that’s the revenue they make from making Google the default search engine and in OpenAI’s case that’s all their ChatGPT and API revenue. The vast majority (all?) engineers work for the for-profit and always have. The vast majority (all?) revenue goes through the for-profit which pays taxes on that revenue minus the usual business deductions. The only money that goes to the nonprofit tax-free are donations. Everything else is taxed at least once at the for-profit corporation. Almost every nonprofit that raises revenue outside of donations has to be structured more or less this way to pay taxes. They don’t get to just take any taxable revenue stream and declare it tax free.
All OpenAI is doing here is decoupling ownership of the for-profit entity from the nonprofit. They’re allowing the for profit to create more shares and distribute them to entities other than the non-profit. Or am I completely misinformed?
[1] https://en.wikipedia.org/wiki/OpenAI#2019:_Transition_from_n...
As you've realized, this should have been (and was) obvious for a long time. But that doesn't make it any less hypocritical or headline worthy.
And now they want to cast off any pretense of that former altruistic yolk now that they have a new, better raison d'etre to attract talent: making absolutely unparalleled stacks of cash.
In terms of profit, AFAICT, Sam doesn't have designs on building extra large yachts and his own space agency but what he wants is to be the one at the stead of building what he considers is world-changing tech. One could rationally call this power-hungry but one could also rationally call this just helicopter parenting of a tech you've helped built. And for that a for-profit that is allowed to maximize profits to re-invest in the tech is the optimal setup (esp if all the competitors are doing the same)
Is this a different org than when it started? Yes. Was this a dupe from the beginning? I don't think so.
"But why can't he have a more worldly-aligned board looking over his shoulder?"
Because we live in California and have a bad taste for governance by committee or worse: governance by constant non-representative democracy (see: Housing).
If this now completely comes off the wheels, I still think Congressional action can be a stopgap, but atleast for now, this restructure makes sense to me.
The board of the non-profit fired Altman and then Altman (& MS) rebelled, retook control, & gutted the non-profit board. Then, they stacked the new non-profit board with Altman/MS loyalists and now they're discharging the non-profit.
It's entirely about control. The board has a legally enforceable duty to its charter. That charter is the problem Altman is solving.
We worry about the existential threat of AI becoming a paperclip factory while many humans are already.
In this case, Mozilla as a non-profit owning a for-profit manages to more or less fulfill the non-profit's mission (maintaining an open, alternative browser). OpenAI has been in a hurry to abandon it's non-profit mission for a while and the complex details of its structure doesn't change this.
Yes, but going from being controlled by a nonprofit to being controlled by a typical board of shareholders seems like a pretty big change to me.
All? As far as I know this is unprecedented.
But unfortunately charities and not-for-profits putting their core business into a company, and then eventually selling it off is not unprecedented. For example, The Raspberry Pi Foundation was a not-for-profit organisation around Raspberry Pi. They formed a LLC for their commercial operations, then gradually sold it off before eventually announcing an IPO: https://www.raspberrypi.org/blog/what-would-an-ipo-mean-for-....
I think it is terrible that not-for-profits are just being used as incubators for companies that eventually take the core mission and stop primarily serving the public interest.
There are of course other examples of charities or not-for-profits that put part of their core operations in a company and don't sell out, instead retaining 100% ownership - for example Mozilla. However, I think there should be some better way for impactful not-for-profits to have some revenue generating aspects in line with their mission (offset by allowing temporary surplus to cover future expenses, or by other expenses).
[1] "The Mozilla Foundation has no members" https://hacktivis.me/articles/mozilla-foundation-has-no-memb...
The nonprofit invests the tax free donations into the for-profit. It gets to keep its equity just like any other investor and as long as that equity held by the non-profit, there's no taxable event. If the nonprofit sells its share - since it was closely involved in the creation and management of the for-profit as an active investment - it becomes a taxable event under the unrelated business income tax rules. Until that time, the only "profit making applications" like ChatGPT and the API are run by the for-profit, which - I repeat - pays its taxes.
I genuinely don't understand why people think they've skirted taxation except out of sheer ignorance of how non-profits actually work. A 501(c)(3) is not some magic Monopoly "get-out-of-tax-free" card and the IRS isn't stupid, there's a ton of rules for tax exemption. They'd have a much easier time with tax avoidance if they were an actual for profit corporation with billions of dollars because GAAP rules are a lot more forgiving than non-profit regulations.
Right now, OpenAI, Inc. (California non-profit, lets say the charity) is the sole controlling shareholder of OpenAI Global LLC (Delaware for-profit, lets say the company). So, just to start off with the big picture: the whole enterprise was ultimately under the sole control of the non-profit board, who in turn was obligated to operate in furtherance of "charitable public benefit". This is what the linked article means by "significant governance changes happening behind the scenes," which should hopefully convince you that I'm not making this part up.
To get really specific, this change would mean that they'd no longer be obligated to comply with these CA laws:
https://leginfo.legislature.ca.gov/faces/codes_displayText.x...
https://oag.ca.gov/system/files/media/registration-reporting...
And, a little less importantly, comply with the guidelines for "Public Charities" covered by federal code 501(c)(3) (https://www.law.cornell.edu/uscode/text/26/501) covered by this set of articles: https://www.irs.gov/charities-non-profits/charitable-organiz... . The important bits are:
I'm personally dubious about the specific claims you made about revenue, but that's hard to find info on, and not the core issue. The core issue was that they were obligated (not just, like, promising) to direct all of their actions towards the public good, and they're abandoning that to instead profit a few shareholders, taking the fruit of their financial and social status with them. They've been making some money for some investors (or losses...), but the non-profit was, legally speaking, only allowed to permit that as a means to an end.Naturally, this makes it very hard to explain how the nonprofit could give up basically all of its control without breaking its obligations.
All the above covers "why does it feel unfair for a non-profit entity to gift its assets to a for-profit", but I'll briefly cover the more specific issue of "why does it feel unfair for OpenAI in particular to abandon their founding mission". The answer is simple: they explicitly warned us that for-profit pursuit of AGI is dangerous, potentially leading to catastrophic tragedies involving unrelated members of the global public. We're talking "mass casualty event"-level stuff here, and it's really troubling to see the exact same organization change their mind now that they're in a dominant position. Here's the relevant quotes from their founding documents:
From their 2015 founding post: https://openai.com/index/introducing-openai/ From their 2018 charter: https://web.archive.org/web/20230714043611/https://openai.co...Sorry for the long reply, and I appreciate the polite + well-researched question! As you can probably guess, this move makes me a little offended and very anxious. For more, look at the posts from the leaders who quit in protest yesterday, namely their CTO.
Look at your OpenAI invoices. They're paid to OpenAI LLC, not OpenAI Inc. I can't find confirmation on openai.com what the exact relationship between OpenAI Global LLC and OpenAI LLC is but the former is on their "Our Structure" page and the latter is in their data processing addendum so it's probably the subsidiary in charge of operating the services while Global does training and licenses it downstream. OpenAI Global was the one that made that big $10 billion deal with Microsoft
That obligation is why they had to spin off a for-profit corporation. Courts are very strict in their interpretation of what "unrelated business income" is and the for-profit LLC protects the non-profit's tax exempt status.
> "why does it feel unfair for a non-profit entity to gift its assets to a for-profit"
What assets were gifted, exactly? They created the for-profit shortly after GPT2 (in 2019) and as far as I can tell that's the organization that has developed the IP that's actually making money now.
I honestly don't understand how this isn't in the interest of the nonprofit's mission. It's currently a useless appendage and will never have any real power or resources until either OpenAI is in the black and sending profit up to it, or they can sell OpenAI shares. I don't think the charity has any more claim over GPT4 than Google does, having invented transformers.
If this next round of funding goes through at $100-150 billion valuation, OpenAI Inc will probably be (on paper at least) the second wealthiest charity on the planet after the Novo Nordisk Foundation. This restructuring opens the way for the nonprofit to sell its shares and it's going to be a hell of a lot of money to dedicate towards their mission - instead of watching its subsidiary burn billions of dollars with no end in sight.
I don't think that's true? A non-profit can sell products or services, it just can't pay out dividends.
What counts as “related” to the charity’s mission is fuzzy but in practice the courts have been rather strict. They don’t have to form for-profit subsidiaries to pay those taxes but it helps to derisk the parent because potential penalties include loss of nonprofit status.
For example, the nonprofit Metropolitan Museum of Modern Art has a for-profit subsidiary that operates the gift shop. National Geographic Society has National Geographic Partners which actually owns the TV channel and publishes the magazine. Harvard and Stanford have the Harvard Management Company and Stanford Management Company to manage their endowments respectively. The Smithsonian Institute has Smithsonian Enterprises. Mayo Clinic => Mayo Clinic Ventures. Even the state owned University of California regents have a bunch of for-profit subsidiaries.
Deleted Comment
The non-profit could maybe sell its assets to investors, but then what would it do with the money?
I'm sure OpenAI has an explanation, but I really want to hear more details. In the most simple analysis of "non-profit becomes for-profit", there's really no way to square it other than non-profit assets (generated through donations) just being handed to somebody for private ownership.
The biggest problem with this is that there's basically no chance that the sale price of the non-profit assets is going to be $150 billion, which means that whatever the gap is between the valuation of the assets and the valuation of the company is pure profit derived from the gutting of the non-profit.
If this is allowed, every startup founded from now on should rationally do the same thing. No taxes while growing, then convert to for profit right before you exit.
Of course the investors do end up owning their shares at a lower basis than they would otherwise, and they end up a bit diluted compared to a straightforward investment, but the investors seem likely to more than make up for this by donating appreciated securities to the 501(c)(3) and by deferring or even completely avoiding the capital gains tax on their for-profit shares.
Obviously everyone needs to consult their lawyer about the probability of civil and/or criminal penalties.
The cleanest way for this to work is the for-profit to just sell more shares at the $150B valuation, diluting the non-profit entity below majority ownership. The for-profit board, which the non-profit could still probably have multiple seats on, would control the real asset, the non-profit would still exist and hold many tens of billions of value. It could further sell its shares in the non-profit and use the proceeds in a way consistent with its mission.
They wouldn't even have to sell that much - I am pretty sure the mega-fundrasing rounds from Microsoft etc brought the non-profit's ownership to just north of 50% anyway.
I don't see how this wouldn't be above board, it's how I assumed it was going to work. It would indeed mean that the entity that controls ChatGPT would now be answerable to shareholders, a majority of which would be profit seeking and a minority of which would be the non-profit with its mission, but non-profits are allowed to invest in for-profits and then sell those shares; all the calls for prosecutions etc seems just like an internet pitchfork mob to me.
Practically the same as selling, but technically not. Non-profit still gets to live up to it's original mission, on paper, but doesn't really do anything internally.
The non-profit’s asset is the value of OpenAI minus the value of its profit-participation units, i.e. the value of the option above the profit cap. Thus, it must be less than the value of OpenAI. The non-profit owns an option, not OpenAI.
― L. Ron Hubbard
If that's how it works, why wouldn't you start every startup as a non-profit?
Investment is tax deductible, no tax on profits...
Then turn it into a for-profit if/when it becomes successful!
It's been many years, but the plan was essentially this:
* The original, non-profit would still exist
* A new, for-profit venture would be created, with the hospital having a board seat and 5% ownership. Can't remember the exact reason behind 5%. I think it was a threshold for certain things becoming a liability for the hospital as they'd be considered "active" owners above 5%. I think this was a healthcare specific issue and unlikely to affect non-profits in other fields.
* The for-profit venture would seek, traditional VC funding. Though, the target investors were primarily in the healthcare space.
* As part of funding, the non-profit would grant exclusive, irrevocable rights of it's IP to that for-profit venture.
* Everyone working for the "startup" would need to sign a new employment contract with the for-profit.
* Viola! You've converted a non-profit into a for-profit business.
I'm fuzzy on a lot of details, but that was the high level architecture of the setup. It's one of those things where the lawyers earn a BOAT LOAD of money to make sure every technicality is accounted for, but everything is just a technicality. The practical outcome is you've converted a non-profit to a for-profit business.
Obviously, this can't happen without the non-profit's approval. From the outside, it seems that Sam has been working internally to align leadership and the board with this outcome.
-----
What will be interesting is how the employees are treated. These types of maneuvers are often an opportunity for companies to drop employees, renegotiate more favorable terms, and reset vesting schedules.
This is the part that should land people literally in jail. A non-profit should not be able to donate its assets to a for-profit, and if it's the same people running both companies, those people must be sent to prison for tax evasion. There is no other way to preserve the integrity of the "non-profit" status with this giant loophole.
Isn't that fraud/stealing from all the donors? I mean how is that different than just giving money to another business not owned by the non profit?
Every answer moving forward now will contain embedded ads for Sephora, or something completely unrelated to your prompt...
That money will go into the pockets of a small group of people that claim they own shares in the company... Then the company will pull more people in who invest in it, and they'll all get profits based on continually rising monthly membership fees, for an app that stole content from social media posts and historical documents others have written without issuing credit nor compensating them.
As long as the money doesn't go into someone's pocket, it's all good (except that Sam Altman is also getting equity but I assume they found a way to justify that.)
OpenAI will eventually be forced to convert from a public charity to a private foundation and will be forced to give away a certain percentage of their assets every year so this solves that problem also.
https://www.businessinsider.com/sam-altman-openai-note-more-...
OpenAI has been one of the most insane business stories in years. I can't wait to read a full book about it that isn't written by either Walter Isaacson or Michael Lewis.
List of crawlers for those who now want to block: https://platform.openai.com/docs/bots
https://github.com/ai-robots-txt/ai.robots.txt/blob/main/rob...
cloudflare have a button for this:
https://blog.cloudflare.com/declaring-your-aindependence-blo...
"If I read your book and I have a photographic memory and can recall any paragraph do I need to pay you a licensing fee?"
"If I go through your library and count all the times that 'the' is adjacent to 'end' do I need to get your permission to then tell that number to other people?"
But the board's lack of communication apparently allowed Altman to demonstrate he was more important to the organization than the formal/legal structure, 90% signed intents to quit and the board backed down. It seemed that Altman simply represented the attitude of the many Silicon Valley tech-people - once you have a chance of money, don't hold back, do everything you can to make it.
Ironically the one person with resources fighting it in a tangible way, even if for spite, is Elon Musk.
Deleted Comment
I can see large copyright holders lining up with takedowns demanding they revise their originating datasets since there will now be a clear-cut commercial use without license.
I thought so for a moment but then again Meta, Anthropic (I just checked and they have a "for profit and public benefit" status whatever that means), Google or that Musk's thing aren't non-profits, are they ? There are lawsuits in motion for sure but with how it stands today I think ai gets off the hook.
Early hires, who were lured there by the mission?
Donors?
People who were supposed to be served by the non-profit (everyone)?
Some government regulator?
In a normal situation, the primary people with standing to prevent such a move would be the board members of the non-profit, which makes sense. Luckily for Sam, the employees helped kick out all the dissenters a long time ago.
Would the founding documents actually count as proof of a lie? I feel like the defense could easily make the argument that the documents accurately represented their intent at the time, but as time went on they found that it made more sense to change.
It seems like, if the founding documents were to be proof of a lie, you'd have to have corresponding proof that the documents were intentionally written to mislead people.
Whether the case is any good is another matter.
Try to contribute to the conversation, though.
What you say is also untrue, there's a minimum set of requirements that have to be met regarding discovery, etc.
https://x.com/elonmusk/status/1839121268521492975
I can’t tell if Elon is saying that the tweet is misinformed or OpenAI’s behavior is reprehensible