Preferences

What is still bizarre is how the board brought him back. I just don't understand that from any angle. It is very clear that he lied and schemed. So why was there any issue? He had no power once fired. If another company wants to hire him, fine. If he wants to make platitudes on Twitter, who cares? Employees threatened to leave? Let them. I imagine many of them wouldn't, and if they did, you could just rehire and find a proper CEO. It's likely that the employees who would have stayed would have been the ones you want to keep anyway.

So why would you rehire him and then resign? Why not leave him fired and resign?

Why did Microsoft care so much about Altman? They didn't have a board seat, so too bad so sad. They should have had no sway. If they want OpenAI, then they can buy it. Why would they even want Altman to come to their company? What possible value could Altman bring to Microsoft?

Just none of it makes any sense. In fact, the most clear thing is Altman's negative behavior and mode of operation. I remember his very loud, awshucks pronouncements of having no shares in OpenAI. Yea, right. A career VC setup a new company structure or structures where he wouldn't benefit at all financially.

In another timeline, OpenAI firing him, naming an interim CEO, ignoring all else, and then hiring a new CEO would have all gone just fine. I don't know why Microsoft and others made a huge kerfuffle over it. And I also don't know why the board released such a cryptic press release when they could have provided details.

I just can't wrap my head around any of it, even letting in conspiracy theory lines. Lol. It just makes no sense and gives me a feeling that no one involved has any clue about anything, including Satya Nadella.


You had the answer there before you dismissed it -- the board agreed to bring Sam Altman back before resigning because of massive pressure campaign that involved 90%+ of employees threatening to resign, and Microsoft (which has OpenAI's IP rights) offering to hire them all immediately. This included the company's key leadership. The organization was unraveling overnight. This was far past the tipping point of "a few disgruntled employees might leave" - there would have been no OpenAI left to preside over.
When the board fires you for wrongdoing it is like being deported from a foreign country for breaking that country’s laws. You can’t just go against that decision. It would be like telling the government that no, in fact they are the ones who are wrong and no, I didn’t break the law it is in fact their laws which are broken.

When you have irreconcilable differences with the board you could theoretically jump ship and start a rival company. Usually, in practice, that’s impossibly hard and yet Microsoft announced that they intended to do exactly that with funding and stock matching. For some reason that was turned down in favor of staging the equivalent of an in-country coup.

If you get caught breaking the law on vacation abroad and your response to being arrested is to take control of the country in retaliation then you are a very powerful, persuasive, or threatening person indeed.

Indeed, it was stunning. The Board outranked the CEO and had the right to replace him. What they didn't reckon on was that technically, if the entire company is willing to quit on behalf of the CEO, then they as a body outrank the board. And that was that, pretty much.
> It would be like telling the government that no, in fact they are the ones who are wrong and no, I didn’t break the law it is in fact their laws which are broken.

You know a number of people are in that position though.

My first thought went to Samsung's chairman who will break the law and go to prison, but the country will have him back, bending over backwards to somewhat have it make sense.

> It would be like telling the government that no, in fact they are the ones who are wrong and no, I didn’t break the law it is in fact their laws which are broken.

Reminds me of the Wirecard scandal doc on Netflix. Wirecard was so powerful that the financial regulators (BaFin) started targeting journalists and actively defending a publicly traded company.

I answered that scenario though. I'm fine, as a board, letting employees who want to hold the company hostage over a CEO that the board fired walk.

The board's major mistake was not communicating why he was let go.

My guess is that the likely reason why employees threatened to go was that they felt Altman had the best chances of making the for-profit arm's shares skyrocket. As a non-profit company's board, I'd be fine letting those people walk out the front door along with the CEO that was just fired.

It is my understanding that the key personnel who developed the actual technology were not part of the group threatening to leave. It was mainly the group in the for-profit arm that Altman had trojan-horsed into the company structure.

> This included the company's key leadership.

I'm not aware of the machine learning researchers responsible for the core technology threatening to leave. Who were they?

Wait, are you simply unaware of the widely reported circumstances in the company at that time? 95% of employees signed a letter to the board stating that they would leave the company if Sam was not brought back [1]. Ilya Sutskever, who was on the board and voted to remove Sam, changed his mind and signed the letter. The board named Mira Murati interim CEO; then she signed the letter, the board fired her, and hired another outside CEO. Several key researchers resigned outright before letter even went out, including Jakub Pachoki, who replaced Ilya after he left [2] .

I would challenge you to name a researcher who didn't resign or threaten to resign. Remember, they all had a plausible landing spot: They could simply show up at Microsoft with all the same leadership, coworkers, salary, compute, and IP the next day. OpenAI as we know it was over unless and until the board gave in.

[1] https://www.wired.com/story/95-percent-of-openai-employees-t... [2] https://www.theinformation.com/articles/three-senior-openai-...

No, I am not simply unaware. I said my understanding, which appears to be wrong. Thanks for the details. However, I am still confused on which employees of which company. There is the non-profit OpenAI and for-profit OpenAI. Are these articles talking about everything or just the for-profit branch?

I still wager that things would have been different had the board clearly stated their reasons. Doesn't make any sense that they did it months later. The signing of the petition seems mostly group think and political. I would guess that the majority of employees would have followed through.

Lastly, I still can't say any of this makes any sense. Why did the employees even care about Altman? It still seems all very strange to leave your job for someone who doesn't seem to have ever said anything meaningful.

> I'm not aware of the machine learning researchers responsible for the core technology threatening to leave. Who were they?

Ilya Sutskever, Alec Radford, Wojciech Zaremba, Nick Ryder, Mark Chen, ... how many names do you want? >90% of the company threatened to leave.

Sutskever was one of the people who fired Altman to begin with.

Him joining the pile-on when it was already clear how big it is was pretty much surrendering to the mob, and was perceived as such even at the time.

It’s so much weirder than that. Reportedly, Greg Brockman’s wife begged him in tears to change his mind. Ilya presided over their wedding.
>I'd be fine letting those people walk out the front door along with the CEO that was just fired.

Guess you'd be fine being the board over a company that now only consists of a board. Thumbs up.

If the board had shared something equivalent to this article at the time, perhaps there wouldn't have been such an intense internal pressure campaign.

Instead, they did not even attempt to communicate any rationale behind their actions.

It’s not good form to bad mouth someone you’ve fired. They did, in fact, release a statement which aligns (in PR speak) with what Toner is claiming now.

Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI.

https://openai.com/index/openai-announces-leadership-transit...

What they didn't realize is that they were going to have to convince the rank-and-file employees of that. Without a communication strategy to do that, they had basically no leverage.
I agree with this 100%

I even said so at the time.

https://www.hackerneue.com/item?id=38372451#38373194

> It’s not good form to bad mouth someone you’ve fired.

Instead, it's good form to wait 6 months and then do a Business Insider interview on the topic?

Also, "bad mouthing" suggests putting a spin on it. Based on the article, they should have been able to list out several objective facts that supported their decision. Instead, they opened it all up to intense speculation by providing a vague justification and by remaining tight lipped even once it became clear that a lot of clarification was needed.

> This was far past the tipping point of "a few disgruntled employees might leave" - there would have been no OpenAI left to preside over.

But if they had the moral conviction, it seems like this would have been the right choice to make, because it would have diluted Altman's power (unless they trust Nadella even less than they trust Altman?)

They should have called their bluff.
There's not many who have the balls to do that left in the country, let alone California or SF.
>the board agreed to bring Sam Altman back before resigning because of massive pressure campaign that involved 90%+ of employees threatening to resign, and Microsoft...

But, what kind of Wild West is this? It's all so unhinged and strange.

There are no NDAs, non-competes or other impediments? MS just guts OpenAI at its whim?

>there would have been no OpenAI left to preside over.

...If MS can do this, then there's already no OpenAI left to preside over.

The beauty of California is that there are no non competes, for anyone, ever.

It is a very pro employee policy, hence attractive to the best and brightest.

Good to see someone positive about California, given all of the politically-motivated negativity. It's a great state for many reasons.

I mentioned non-competes as one of a universe of things that makes this unusual; including its original non-profit status becoming substantially for-profit, and the apparent ease with which MS can now gut the supposedly controlling non-profit and walk away with everything.

And, I'm guessing even the CA policy on non-competes wasn't conceived with the idea of one company simply "taking" another company at-will, even if that scenario is technically covered.

I am not clear about the intent of your comment. People should be free to sell their labor to whoever they want. And people should be able to buy labor from whoever they want.

All businesses at all times are subject to losing their employees to another employer who is willing to pay enough for them.

My original point had very little to do with selling labor/non-competes, but I think you keyed on that part. Maybe re-reading and omitting that will make it clearer.
Yeah, sadly it looks like it was employees' greed that brought Altman back.
And where does this leave the current board? The CEO gave no mea culpa, and even if he did, actively thwarting board oversight is a clear no-no. How in the world do they expect to perform their duties with a CEO who demonstrated himself to refuse to even provide them the information they need to discharge their duties? As long as they keep this CEO in a position to block their access to information, the board has no credibility.
100% agree. I don’t care about OpenAI or the members of its board at all, one bit.

What I did see was so much incompetence at the one thing I expect a board to be at least okay at. Hiring and firing.

For that alone, I think the board reshuffle was good. Regardless of who you support in all of this.

It clearly wasn't good; the Altman-control system failed, and was replaced with not even the pretense of an Altman-control system.

Sure, the rectification of names is an improvement in a sense; what is actually needed is a working Altman-control system.

> I don't know why Microsoft and others made a huge kerfuffle over it.

Microsoft needs the sold-out version of OpenAI so they can make as much money as possible without anyone making pesky noises about ethics and safety.

yeah MS wants the shield that OpenAI provides.

Outsource all of the risk to a non-profit, but still be able to run it, and snag up all of the researchers if/when something gets ugly.

In deciding to hire him back, Helen Toner said OpenAI lawyers said she and other members of the board could be personally liable if the company lost a lot of money over keeping him out.
I suppose that would be a reasonable reason on a practical level. However, how is a board liable for that? The CEO lies and intentionally withholds information, employees strangely hold the company hostage, and yet it's the board held liable? The board governed the non-profit OpenAI.
I am pretty sure a lawyer could be found that would happily take that case against the board. And that Altman would be quite willing to go that far. At least I would not have staked my professional life and personal economy on it not happening.

PG called it over 15 years ago: "You could parachute him [Sam Altman] into an island full of cannibals and come back in 5 years and he'd be the king".

Which should have been a bluff (board members typically have "director's insurance" to cover exactly this scenario - trying to control a company by threatening to pierce the corporate veil and sue board members individually) but shows you what sort of tactics were being employed by the Altman side. cf. Sutskever's flipping
The answer is simple: the single most valuable asset at OpenAI appears to be Altman bullshitting about science fiction AI, and nothing else. When he disappear, value plumetted, partners evaporated, clients stopped paying, competitors rushed to hire him, etc.

It's like Elon 'Electric Jesus' Musk, without him, they're just selling shitty cars nobody quite like. So he can get paid more than all profit Tesla ever made, because without him, there would be 0 profit anyway.

Talk to Tesla owners: they are surprised by how shit the car is, but they feel like mini Elons. That's probably similar at OpenAI ?

Better at OpenAi then a private team at Microsoft??? I don't know what they were thinking either
Employees are not replacable like that.

Almost the entire company was threatening to quit.

If it was possible to simply replace all of openAI, then you could just do that now, as an outside party.

So the boards choice was to either bring back Sam or watch the entire company go under.

> Almost the entire company was threatening to quit.

This is the part that perplexes me. A CEO being fired is not an unusual occurrence. What about Sam Altman led such a huge number of employees to threaten to follow him out the door? Was it that the board's actions were viewed as internally unjust? Was it Altman's power of persuasion? Was/is Altman viewed by the staff as bringing something irreplaceable to the table in terms of talent, skill or ability?

> What about Sam Altman led such a huge number of employees to threaten to follow him out the door?

My understanding is that a fair amount of it was essentially peer pressure. If you're an employee whose CEO has just been fired for unclear reasons, and someone hands you a chain letter saying "restate him or everyone will quit" and you're told everyone's signing that letter and there's already someone who promised to hire everyone who quits into their current role at equal pay, would you sign it?

My understanding is that it was not so much peer pressure so much as an explicitly coordinated worker action, based on a rational assessment of what their comp packages would look like if the board's plan played out completely.
In my opinion this is most likely.

Scenario A, Sam Altman continues as CEO and the for profit arm of OpenAI continues to call the shots, growth and market share continue to be priority number one.

Scenario B, Sam Altman is fired, OpenAI non-profit board (re?)-asserts more control. Safety, alignment and other things like that take priority over growth and market share.

x% of Employees are Sam believers, and when the remaining x% of ambiguous/non-Sam believers realize the first x% might leave, their PSUs would be worth significantly less, so they sign the letter as well. There is also the peer pressure / fear of retribution factor as well once it becomes likely there is even a chance of Sam being reinstated.

Many employees choose scenario A because it is likely their "profit sharing units" will be worth more than with scenario B. There's a non-zero chance that OpenAI (the for profit arm) eventually joins the ranks of the "FAANG" companies. Those PSUs might be worth millions today, but in the future could be worth "fuck you" levels of money.

I suspect it was all about the money. Those following him viewed him as their best bet for a good financial outcome for themselves.
It’s covered in the podcast that the article is about.

Summarizing, essentially the employees were under the impression either Sama comes back OR OpenAI dissolves and they lose their job.

It’s important to understand that the board oversaw the nonprofit, and their job essentially was to protect the nonprofits mission. Most of the employees were hired by the for-profit company whose mission it is to make money. In theory the for profit was owned and subservient to the non profit but the engineers financial future was tied to the for profit and thus their loyalty.
A CEO being fired, without explanation, at a company that is doing well is highly unusual.
The equity offering was nearing and then the board just fired Altman out of nowhere and had no credible reasons or a narrative to sell to the employees, then they quickly caught on that their comp would go to smokes with this clueless lot at the helm who can't even justify a firing without making a pig's breakfast of the whole affair and demanded the reinstatement of the CEO for their pocket's sake, simple as that.
I would argue it is a demonstration of what Altman is capable of bringing about. He can make things like this 'happen' for him. Caesar, Napoleon, Alexander, and now, our current batch.
The board should have been transparent about the dishonesty, especially around financial stakes, and I suspect the OpenAI employees and MSFT would have reacted differently.
The dishonesty regarding the ChatGPT release sounds like the best decision that Sam or anyone could have made in the history of the company.

The ChatGPT release is what made the AI movement go mainstream. It is why OpenAI is worth ~80 billion dollars.

By Gods am I glad that the board wasn't able to stop ChatGPT from being released.

I thought it wasn't "released" initially in a manner that was meant to be a product.

I seem to recall that the release in late November 2022 was only intended to be a way for volunteers on the Internet to experiment with it and provide feedback.

> It is why OpenAI is worth ~80 billion dollars.

Worth to who?

Well, that was the valuation of their last fund raising round, which included a tender offer to employees. So it was worth that much to investors and to its current employees.
Worth is determined in the same way that it is for every single other thing in the world.

It is "worth" what someone else is willing to pay for it.

and aligning ai is even harder...
Sam Altman made openAI possible. A true CEO like Satya knows without Sam, the rate of progress at OpenAI will be determined by decels like Helen/Jan aka ZERO. It's a disaster for product development, especially when you have openly invited Google to dance.

I know HN leans engineering/safety/reliability/labor/pedantic (like chasing the absolute truth), but at the end of the day, company scales from the likes of Jobs/Musk/Sam/Zuck even it involves deceit or reality distortion field.

Sometimes people just can't handle the truth or don't believe in the vision of visionaries. So, they have to fib a little to the decels and normies. Even Larry / Sergey 'lied' to Eric during Google's growth phase. It's only when they bought normie Sundar that Google became risk averse. And look where it got Google to.

If I have to bet my last $, I'd bet on Elon/Sam/Zuck/Jobs than Helen/Jan/Sundar.

This isn't lying about product philosophy, lying about your financial stake in an OpenAI VC group would be grounds for dismissal from any company.
> “Decel” is used as a dismissive term for tech doomsayers by those who see themselves as members of the burgeoning E/Acc community. [... Who have] the belief that in our current technological age, the powers of innovation and capitalism should be exploited to their extremes to drive radical social change - even at the cost of today’s social order.

I'm sure that--in practice--there will be extremely strong correlation to social-order changes that leave the C-suite richer than before. :p

This is the literal definition of a slippery slope. Letting billionaires, who already wield extreme power, break laws and skirt regulations is the opposite of a free market.
Where do you get this idea that we have a free market? Money begets money, and as the commenter above pointed out, those who have money know how to make it and know that it's the Musk/Zuck/Altmans of the world who get it done.

We don't necessarily have to like it to acknowledge it's very much reality.

The "free market" literally means free of government interference and regulation...

So it actually implies allowing capitalists to wield extreme power, skirt laws and regulations.

I understand the term has gone through some kind of whitewashing to mean "this is the good system (unlike the bad system)", so one might be inclined to think it means something more equitable, but seriously, that was the original meaning.

This item has no comments currently.

Keyboard Shortcuts

Story Lists

j
Next story
k
Previous story
Shift+j
Last story
Shift+k
First story
o Enter
Go to story URL
c
Go to comments
u
Go to author

Navigation

Shift+t
Go to top stories
Shift+n
Go to new stories
Shift+b
Go to best stories
Shift+a
Go to Ask HN
Shift+s
Go to Show HN

Miscellaneous

?
Show this modal