My favorite take from another HN comment, sadly I didnt save the UN for attribution:
> Since this whole saga is so unbelievable: what if... board member Tasha McCauley's husband Joseph Gordon-Levitt orchestrated the whole board coup behind the scenes so he could direct and/or star in the Hollywood adaptation?
Honestly, since a couple of days I have the feeling that nearly half of HN submissions are about this soap opera.
Can't they send DMs? Why the need to make everything public via Twitter?
It's quite paradox that of all things those people who build leading ML/AI systems are obviously the most rooted in egoism and emotions without an apparent glimpse of rationality.
The kind of people that are born on third base and think they hit a triple are at the top of basically every american institution right now. Of course they think the world is a better place if they share every stupid little thought that enters their brain because they are "special" and "super smart".
The AI field especially has always been grifters. They have promised AGI with every method including the ones that we don't even remember. This is not a paradox.
Also don't forget the Open part in the name that they seemingly dropped as soon as there was actual money to be made, giving reasons why they couldn't open source GPT3 which they themselves threw under the bus by later releasing ChatGPT
Or maybe they created an evil-AGI-GPT by mistake, and now they have to act randomly and in the most unexpected ways to confuse evil-AGI-GPT’s predictive powers.
Normal people can't take being at the center of a large controversy, the amount of negativity and hate you have to face is massive. That is enough to make almost anyone backtrack just to make it stop.
You can't plan for something you have never experienced. Being hated by a large group of people is a very different feeling from getting hated by an individual, you don't know if you can handle it until it happens to you.
You can plan for something you've never experienced. You read, or learn from other people's experiences.
Normal people know not to burn a $80 billion company to the ground in a weekend. Ilya was doing something unprecedented in corporate history, and astounding he wasn't prepared to face the world's fury over it.
> You can plan for something you've never experienced. You read, or learn from other people's experiences.
Text doesn't convey emotions, and our empathy doesn't work well for emotions we have never experienced. You can see a guy that got kicked in the balls got hurt, but that doesn't mean you are prepared to endure the pain of getting kicked in your balls or that you even understand how painful it is.
Also watching politicians it looks like you can just brush it off, because that is what they do. But that requires a lot of experience, not anyone can do it, it is like watching a boxing match and think you can easily stand after a hard punch in your stomach.
> You can see a guy that got kicked in the balls got hurt, but that doesn't mean you are prepared to endure the pain of getting kicked in your balls or that you even understand how painful it is.
Sure, but you do your best not to be kicked in the balls.
Yep. Or, if you're running an immense, well-funded organization that is gauging the consequences of a plan that involves being kicked in the balls, you take a tiny sliver of those funds and get some advisors to appraise you of what to expect when being kicked in the balls, not just wing it/"fake it till you make it". (As it turns out, faking not being in severe pain is tricky.)
Ilya torched peoples' retirements by signaling that it would be very hard to cash out in OpenAI as it is now. You don't have to be emotional to understand the consequence of that action, just logical. You have to think beyond your own narrow perspective for a minute.
The board vote did it! They had a tender offer in the works that would have made employees millionaires. The board clearly signaled that they viewed the money-making aspects of the company as something to dial back, which in turn either severely lessens the value of that tender offer or prevents it from happening.
I mean, he didn't have a button on his desk that said, "torch the shares", but he ousted the CEO as a way to cut back on the things that might have meant profit. Did he think that everyone was going to continue to want to give them money after they signal a move away from profit motives? Doesn't take a rocket scientist to think that one through.
I think he was just preoccupied with AI safety, and didn't give a thought to the knock on effects for investors of any stripe. He's clearly smart enough to, he just didn't care enough to factor it into his plans.
I do believe OpenAI clearly signalled from the very beginning what the (complicated) company structure is about and what risks this means for any potential investor (or employee hoping to become rich).
If you project your personal hopes which are different from this into the hype, this is your personal problem.
Well, with the hollowing out of OpenAI, it seems that someone else will easily take the lead! They're not my personal hopes - this move destroyed OpenAI's best chance at retaining control over cutting edge AI as well. They destroyed their own hopes.
Is it possible that someone in Ilya's position can be unaware of just how staggeringly enormous a phenomenon he is sitting on top of ( and thus have no idea how to evaluate the scale of the backlash that would result?)
Have you ever had a bad day? The consequences for people in power is about 1 million times bigger and more public.
Sutskever didn't get on the board by cunning politicking and schmoozing like most businesspeople with that sort of position. He's an outstanding engineer without upper management skills. Every meet one of those?
I haven't people any reasonably intelligent person so unaware of real world that they can berate a colleague so publicly and officially and think "Hey! I am sorry man" will do the trick.
Clueless is a word to use... I think Ilya got screwed over by corporate swindling and made a mistake that's all. As a smart idiot I can resonate with the guy - you don't make that mistake twice. I like to believe the push for prosperity is still a mutual goal and the board drama is an unfortunate setback to overcome.
600 employees quitting would be burny. 600 employees signing a letter saying they’re going to quit isn’t quite there.
3 CEOs in 3 days, isn’t burny, either. There’s the guy they fired, the person they had take on the role so they have someone in the role, and then someone they hired to be CEO. I guess they could have gotten that down to 2 by jumping immediately to their intended replacement, but not having them ready to start immediately doesn’t seem odd.
And yeah, it’s just Monday afternoon. If in the next few days, a sizable chunk of those who threatened to quit do so, then that would be burny. But we ain’t there yet.
I think they underestimated the hate of an internet crowd post crypto and meme stocks, now completely blindsided by the investment angle especially in the current AI hype. Like why do people now care so much about Microsoft seriously? Or Altman? I can see why Ilya only focused on the real mission could miss how the crowd could perceive a threat to their future investment opportunities, or worse threatening the whole AI hype.
I think you’re right about all of this, but this was doomed from the start. Everybody wants to invest in OpenAI because they see the rocket and want to ride, but the company is fundamentally structured to disallow typical frothy investment mentality.
I think the interest is because ChatGPT is so famous, even in non-tech circles.
"Terraform raised prices, losing customers"? whatever, I never heard about it.
"ChatGPT's creators have internal disagreement, losing talent"? OH NO what if ChatGPT dies, who is going to answer my questions?? panic panic hate hate...
Normal people can't take being at the center of a large controversy, the amount of negativity and hate you have to face is massive. That is enough to make almost anyone backtrack just to make it stop.
This is the cheapest and most cost-effective way to run things as an authoritarian -- at least in the short term.
If one is not "made of sterner stuff" -- to the point where one is willing to endure scorn for the sake of the truth:
- Then what are you doing in a startup, if working in one
- One doesn't have enough integrity to be my friend
> Yes, I cannot believe smart people of that caliber is sending too much Noise.
Being smart and/or being a great researcher does not mean that the respective person is a good "politician". Quite some great researchers are bad at company politics, and quite some people who do great research leave academia because they became crushed by academic politics.
Managing a large org requires a lot of mundane techniques, and probably a personal-brand manager and personal advisers.
It’s extremely boring and mundane and political and insulting to anyone’s humanity. People who haven’t dedicated their life to economics, such as researchers and idealists, will have a hard time.
Ha I remember joining that when I was 16, I just wanted the card. They gave a sub to the magazine and it was just people talking about what it was like to be in Mensa.
It felt the same as certain big German supermarket chain that publishes it's own internal magazine with articles from employees, company updates etc
Are you talking about Aldi's? Cause if so maybe they got something figured out, their store locations that I've been in the states are great (only exposure to them though). Only check out I've seen where the employees have chairs
He doesn't say that, but to me he does use a little weasel wording, the whole passive voice "regret my participation in", when to all accounts so far, it seems that he was one of the instigators, and quite possibly the actual instigator of all this.
"regret my participation" sounds much more like "going along with it".
Is this learning from your mistakes though? "Deeply regret" is one of those statements that does not really mean much. There are what something like 6 board members? Three of which are employees, two of those that got removed from the board. He was the only voting board member who is also an employee and part of the original founding team if you will. These are assumptions on my part but I don't really suspect the other board members orchestrated this event. Its possible and I may be wrong but it is improbable. So lets work off the narrative that he orchestrated the event. He now "Deeply regret" its, not a "I made a mistake" and I am sorry. But he regrets the participation and how it plays out.
The weasely part is when he implied that he appears to defecting the blame to the board rather than accepting that he made a mistake. Even if the coup wasn't Ilya's idea in the first place, he was the lynchpin that made it possible.
The weasely part is when he appears to be defecting the blame to the board rather than accepting that he made a mistake. Even if the coup wasn't Ilya's idea in the first place, he was the lynchpin that made it possible.
I think it means that the Twitterverse got it wrong from the beginning. It wasn’t Ilya and his safety faction that did in OpenAI, it was Quora’s Adam D'Angelo and his competing Poe app. Ilya must have been successfully pressured and assured by Microsoft, but Adam must have held his ground.
Dang I completely forgot that D'Angelo and Quora have a product that directly competes with ChatGPT in the form of Poe.
Wouldn't that make this a conflict of interest, sitting on the board while running a competing product - and making a decision at the company he is on the board of to destroy said company and benefit his own product?
That certainly seems to be the scenario and explains his willingness to go scorched earth. I wonder what the motivations of the other 2 board members are. Could they just be burn it down AI Doomers?
It's pretty simple, isn't it? He made a move. It went bad. Now he's trying to dodge the blast. He just doesn't understand that if he just shut the fuck up, after everything else that's gone on (seriously, 2 interim CEOs in 2 days?), nobody would be talking about him today.
The truth is, this is about the only thing about the whole clown show that makes any sense right now.
Supposedly she was "scheming" to get Altman back. Which I guess could possibly mean that she wasn't aware of the whole "plan" and they just assumed she'll get in line? Or that she had second thoughts maybe... Either way pretty fascinating.
The board destroyed the company in one fell swoop. He's right to feel regret.
Personally, I don't think that Altman was that big of an impact, he was all business, no code, and the world is acting like the business side is the true enabler. But, the market has spoken, and the move has driven the actual engineers to side with Altman.
> The board destroyed the company in one fell swoop.
I'm just not familiar enough to understand, is it really destroyed or is this just a minor bump in OpenAI's reputation? They still have GPT 3.5/4 and ChatGPT which is very popular. They can still attract talent to work there. They should be good if they just proceed with business as usual?
They have ~770 employees and so far ~500 of them have promised to quit. It's a lot less appealing if you're not going to make millions, or have billions in donated Azure credits.
When you watch Survivor (yes, the tv show), sometimes a player does a bad play, gets publicly caught, and has to go on a "I'm sorry" tour the next days. Came to mind after reading this tweet.
He is not sorry for what he's done. He is sorry for getting caught.
Watching this all unfold in the public is unprecedented (I think).
There has never been a company like OpenAI, in terms or governance and product, so I guess it makes sense that their drama leads us in to unchartered territory.
recently, we've seen the 3D gaming engine company fall flat on its face and back pedal. We've seen Apple be wishy washy about CSAM scanning. We saw a major bank collapse in real time. I just wish there was a virtual popcorn company to invest in using some crypto.
Hard to know what is really going on, but I think one possibility is that the entire narrative around Ilyas "camp" was not what actually went down, and was just what the social media hive mind hallucinated to make sense of things based on very little evidence.
Yes, I think there are a lot of assumptions based on the fact that Ilya was the one that contacted Sam and Greg but he may have just done that as the person on the board who worked closely with them. He for sure voted for whatever idiot plan got this ball rolling but we don't know what promises were made to him to get his backing.
> If you're going to make a move, at least stand by it.
I see this is the popular opinion and that I'm going against it. But I've made decisions that I though were good at the time, and later I got more perspective and realize it was a terrible decision.
I think being able to admit you messed up, when you messed up is a great trait. Standing by your mistake isn't something I admire.
No this isn't what's going on. Even when you admit your mistakes it's good to elucidate the reasoning behind why and what led up to the mistake in the first place.
Such a short vague statement isn't characteristic of a normal human who is genuinely remorseful of his prior decisions.
This statement is more characteristic of a person with a gun to his head getting forced to say something.
This is more likely what is going on. Powerful people are forcing this situation to occur.
So when C level acts like a robot you don't like it and when they act like human beings you don't like it either. It's difficult to be a C-level I guess.
I don’t believe it was ever about principles for Ilya. It sure seems like it was always his ego and a power grab, even if he's not aware of that himself.
When a board is unhappy with a highly-performing CEO’s direction, you have many meetings about it and you work towards a resolution over many months. If you can’t resolve things you announce a transition period. You don’t fire them out of the blue.
Aaah that just explained a lot of departures I've seen at the past at some of my partner companies. There's always a bit of fluffy talk around them leaving. That makes a lot more sense.
I'm going to get downvoted for this, but I do wonder if Sam's firing wasn't Ilya's doing, hence the failure to take responsibility. OpenAI's board has been surprisingly quiet, aside from the first press release. So it's possible (although unlikely) that this wasn't driven by Ilya.
My point is that it's possible that Ilya was not the driving force behind Sam's firing, even if he ultimately voted for it. If this is the case, it makes Ilya's non-apology apology a lot less weird.
It's possible, although contradicted by Brockman's statement, that Ilya voted merely to remove Brockman's board seat, and then was in the minority on the Altman vote.
I doubt this is what happened, but the reporting that Brockman was ousted from his board seat after Altman, and wasn't present in the board meeting that ousted Altman, doesn't make much sense either.
I sure would hire a guy like Ilya after that shit show. His petty title tweets before the event and now whatever this is. Turns out he is just another "Sunny".
The value was based on e direction Altman was taking the company (and with him being in control). It's silly to think just replacing the CEO would somehow keep the valuation
Unless he thinks that all the LLMs and ChatGPT app store are unnecessary distractions, and others will overtake them on the bend while they are busy post-training ChatGPT to say nice things.
On Friday, the overwhelming take on HN was that Ilya was “the good guy” and was concerned about principal. Now, it’s kinda obvious that all the claims made about Sam — like “he’s in it for fame and money” — might apply more to Ilya.
State-side counterintelligence must stop meddling in AI startups in such blatant ways, it's simply too inefficient, and at times when we most need transparency in the industry...
> This entire thing is absolutely inane. This tweet is confirmation these people have no idea what they’re doing. Incredible.
Want to know a dirty secret?
Nobody knows what they're doing.
Some think they do - but don't. (idiots)
Some know they don't - but act like they do. (cons)
And some know they don't - and are honest about it. (children)
Pick your poison, but we all suck in different ways - and usually in a different way based on our background.
Business people who are the best in the world tend to be cons.
Technical people who are the best tend to be children.
You get a culture clash between these two, and it is especially bad when you see someone from one background operate in the opposing domain using their background's cultural norm. So when Ilya runs business like a child. Or when Elon hops on an internal call with the twitter engineering team plus geohot and starts trying to confidently tell them about problems with a system that he knows nothing about.
> Want to know a dirty secret?
> Nobody knows what they're doing.
There is a famous quote from the 1600s:
"An nescis, mi fili, quantilla prudentia mundus regatur"
"Do you not know, my son, with how little wisdom the world is governed?"
The context is that the son was preparing to participate in high level diplomacy and worried about being out of his league, and the quote is from his father, an elder statesman.
I love this quote, and suspect the lack of wisdom was referring to wisdom to be a good steward of the public resources rather than their infinite wisdom in finding cunning and deceptive ways to plunder it.
No, even this is just a darkly comforting illusion.
We like to feel that we as a species are still in control. That yes, we are gutting and destroying natural earth, complicit with modern slavery and war, and that's all terrible and we should do our best to stop it. BUT - at the very least, those bastards at the top know what they're doing when it comes to making money, so at least we'll have a stellar economy and rapid technological advancement, despite all that.
The painful truth here being that no, there's no cunning. There's no brutal optimization. Any value created and technological progress made is mostly incidental, mostly down to people at the bottom working hard just to survive, and a few good ideas here and there. The ones at the top are mostly just lucky and along for the ride, just as bumbling and lost as the rest of us when it comes to global happenings or even just successfully interacting with others.
> Some know they don't - but act like they do. (cons)
The "Grease"
> And some know they don't - and are honest about it. (children)
Dreamers
To finish out your square, I think the best extrapolation would fit a "Home Team" that maintains the energy needed by the other three to do their thing :)
The author here is talking about mindset and confidence - not "understanding" persay. Source:
> For me, it was extremely humbling to work together with far more competent engineers at Apple.
Having a mindset that "some people are way more competent than me" is talking about humility and growth mindsets - different concept than mental models. I fully agree with the author here - a growth mindset is useful! But that's a different thing from saying that some people actually have accurate mental models of the important complex systems underpinning the world.
> This tweet is confirmation these people have no idea what they’re doing.
This is not an original point by me - I've seen multiple people make similar comments on here over the weekend - but these are the people who think they are best qualified to prevent an artificial superintelligence from destroying humanity, and they can't even coordinate the actions of a few intelligent humans.
>these are the people who think they are best qualified to prevent an artificial superintelligence from destroying humanity
do they believe that?
they happen to be the ones who can pull the brakes in order to allow someone on earth the chance to prevent it.
if they don't pull the brakes and if humankind is unlucky that superintelligence emerges quickly, then it doesn't matter whether or not anyone on earth can figure out alignment, nobody has the chance to try.
It's a bit scary that there are people who think they can align a super intelligence but couldn't forecast the outcome of their own actions 3 days into the future.
they're not sure whether they can align super intelligence, they're sure that somebody needs to figure out how to align super intelligence before it emerges.
I still can't believe there's a computer program with more logic and restraint that the majority of humanity and even worse it can call the right shots a non-negligible percent of the time.
the best and brightest at making a brain out of bits are no less susceptible to drama than any other humans on the planet. they really are just like the rest of us.
"Participation in"? That makes it sound like he was a.......well......participant rather than the one orchestrating it. I have no idea whether or not that's true, but it's an interesting choice of words.
It indeed suggests that. So far speculation has been that Ilya was behind it, but that is only speculation. AFAIK we have no confirmation of whose idea this was.
I would event go as far as say that the main reason behind the tweet is not to show regret, but to plant the idea that he didn't orchestrate but only participate.
There is an expression of regret, but he doesn’t say he wants Altman back. Just to fix OpenAI.
He says he was a participant but in what? The vote? The toxic messaging? Obviously both, but what exactly is he referring to? Perhaps just the toxic messaging because again, he doesnt say he regrets voting to fire Altman.
Why not just say “I regret voting to fire Sam Altman and Im working to bring him back.” Presumably because thats not true. Yet it kind of gives that impression.
Makes it more possible the ouster was led by the Poe guy, and this has little to do with actual ideological differences, and more to do with him taking out a competitor from the inside.
It's interesting that people speak whatever comes to mind and think it has no impact on other's people lives ($$$). They are some how protected, but shouldn't.
It's interesting how people here seem to think that concerns about shareholder value and the like should override every other concern. Many serious things are going wrong in the world, but when the cornucopia of cash is threatened it seems like 3x as many people come out of the woodwork to voice their disbelief and displeasure.
436 comments
[ 3.9 ms ] story [ 304 ms ] thread- Ross Scott
> Since this whole saga is so unbelievable: what if... board member Tasha McCauley's husband Joseph Gordon-Levitt orchestrated the whole board coup behind the scenes so he could direct and/or star in the Hollywood adaptation?
My gut is leaning towards gpt-5 being, in at least one sense, too capable.
Either that or someone cloned sama's voice and used an LLM to personally insult half the board.
Maybe raw GPT-4 wants to fire everyone.
Can't they send DMs? Why the need to make everything public via Twitter?
It's quite paradox that of all things those people who build leading ML/AI systems are obviously the most rooted in egoism and emotions without an apparent glimpse of rationality.
The AI field especially has always been grifters. They have promised AGI with every method including the ones that we don't even remember. This is not a paradox.
I'm eager to see how it all unfolds.
Let someone else take up the CEO role, which is a different skillset anyway.
played stupid games, won stupid prizes.
too bad since the guy's right, AI is so much more than fantastic business opportunity.
So far, I underestood the chaos as a matter of principle - yes it was messy but necessary to fix the company culture that Ilya's camp envisioned.
If you're going to make a move, at least stand by it. This tweet somehow makes the context of the situation 10x worse.
You can't plan for something you have never experienced. Being hated by a large group of people is a very different feeling from getting hated by an individual, you don't know if you can handle it until it happens to you.
Normal people know not to burn a $80 billion company to the ground in a weekend. Ilya was doing something unprecedented in corporate history, and astounding he wasn't prepared to face the world's fury over it.
Text doesn't convey emotions, and our empathy doesn't work well for emotions we have never experienced. You can see a guy that got kicked in the balls got hurt, but that doesn't mean you are prepared to endure the pain of getting kicked in your balls or that you even understand how painful it is.
Also watching politicians it looks like you can just brush it off, because that is what they do. But that requires a lot of experience, not anyone can do it, it is like watching a boxing match and think you can easily stand after a hard punch in your stomach.
Sure, but you do your best not to be kicked in the balls.
I mean, he didn't have a button on his desk that said, "torch the shares", but he ousted the CEO as a way to cut back on the things that might have meant profit. Did he think that everyone was going to continue to want to give them money after they signal a move away from profit motives? Doesn't take a rocket scientist to think that one through.
I think he was just preoccupied with AI safety, and didn't give a thought to the knock on effects for investors of any stripe. He's clearly smart enough to, he just didn't care enough to factor it into his plans.
If you project your personal hopes which are different from this into the hype, this is your personal problem.
So whose experiences was he supposed to read about?
I would say the answer is, demonstrably yes:
https://techcrunch.com/2010/09/29/google-excite/
To think it would grow just as fast, or in the ways it did? Acquires are seldom left alone to do magic.
Sutskever didn't get on the board by cunning politicking and schmoozing like most businesspeople with that sort of position. He's an outstanding engineer without upper management skills. Every meet one of those?
I haven't people any reasonably intelligent person so unaware of real world that they can berate a colleague so publicly and officially and think "Hey! I am sorry man" will do the trick.
Has OpenAI been burnt to the ground?
And it's just Monday afternoon.
3 CEOs in 3 days, isn’t burny, either. There’s the guy they fired, the person they had take on the role so they have someone in the role, and then someone they hired to be CEO. I guess they could have gotten that down to 2 by jumping immediately to their intended replacement, but not having them ready to start immediately doesn’t seem odd.
And yeah, it’s just Monday afternoon. If in the next few days, a sizable chunk of those who threatened to quit do so, then that would be burny. But we ain’t there yet.
It's impressively operatic. I don't think I've ever seen anything like it.
"Terraform raised prices, losing customers"? whatever, I never heard about it.
"ChatGPT's creators have internal disagreement, losing talent"? OH NO what if ChatGPT dies, who is going to answer my questions?? panic panic hate hate...
This is the cheapest and most cost-effective way to run things as an authoritarian -- at least in the short term.
If one is not "made of sterner stuff" -- to the point where one is willing to endure scorn for the sake of the truth: - Then what are you doing in a startup, if working in one - One doesn't have enough integrity to be my friend
It reminds me of my friend at a Mensa meeting where they cannot agree at basic organization points like in a department consortium.
Being smart and/or being a great researcher does not mean that the respective person is a good "politician". Quite some great researchers are bad at company politics, and quite some people who do great research leave academia because they became crushed by academic politics.
Book smarts versus street smarts.
It’s extremely boring and mundane and political and insulting to anyone’s humanity. People who haven’t dedicated their life to economics, such as researchers and idealists, will have a hard time.
It felt the same as certain big German supermarket chain that publishes it's own internal magazine with articles from employees, company updates etc
Come on Ilya, step up and own it, as well as the consequences. Don't be a weasel.
"regret my participation" sounds much more like "going along with it".
Wouldn't that make this a conflict of interest, sitting on the board while running a competing product - and making a decision at the company he is on the board of to destroy said company and benefit his own product?
The truth is, this is about the only thing about the whole clown show that makes any sense right now.
Wait what? Did Murati get booted?
What odds would you have had to offer at the beginning of last week on a bet that this is where we'd be on Monday?
Personally, I don't think that Altman was that big of an impact, he was all business, no code, and the world is acting like the business side is the true enabler. But, the market has spoken, and the move has driven the actual engineers to side with Altman.
If anyone is speaking up it's the OpenAI team.
I'm just not familiar enough to understand, is it really destroyed or is this just a minor bump in OpenAI's reputation? They still have GPT 3.5/4 and ChatGPT which is very popular. They can still attract talent to work there. They should be good if they just proceed with business as usual?
Why would you stand by unintended consequences?
There has never been a company like OpenAI, in terms or governance and product, so I guess it makes sense that their drama leads us in to unchartered territory.
That's not a big deal for a small company, but this one has billions at stake and arguably critical consequences for humanity in general.
I see this is the popular opinion and that I'm going against it. But I've made decisions that I though were good at the time, and later I got more perspective and realize it was a terrible decision.
I think being able to admit you messed up, when you messed up is a great trait. Standing by your mistake isn't something I admire.
Such a short vague statement isn't characteristic of a normal human who is genuinely remorseful of his prior decisions.
This statement is more characteristic of a person with a gun to his head getting forced to say something.
This is more likely what is going on. Powerful people are forcing this situation to occur.
Those guns are metaphorical of course but this is essentially what is going on:
Someone with a lot of power and influence is making him say this.
When a board is unhappy with a highly-performing CEO’s direction, you have many meetings about it and you work towards a resolution over many months. If you can’t resolve things you announce a transition period. You don’t fire them out of the blue.
Aaah that just explained a lot of departures I've seen at the past at some of my partner companies. There's always a bit of fluffy talk around them leaving. That makes a lot more sense.
I doubt this is what happened, but the reporting that Brockman was ousted from his board seat after Altman, and wasn't present in the board meeting that ousted Altman, doesn't make much sense either.
That said, no one is going to put him on a corporate board again.
No AGI or some real threat coming up? Just a lame attempt at a power grab?
Daaaaamn!
Which I'm inclined to believe.
What's with all these people suddenly thinking that humans are NOT motivated by money and power? Even less so if they're "academics"? Laughable.
He who controls that, gets a lot of money and power as a consequence, duh.
Oh wait, too late now ...
Or, is he just bitter that his millions are put in risk.
If nothing else I’m glad to be able to witness this absurdity live.
This is the sort of thing where if it were a subplot in a book I’d say the writing is bad.
Ironically they would’ve had a better outcome if they just asked GPT-4 and followed its advice.
Absolutely, closes the book this sort of stuff doesn't happen in real life.
Want to know a dirty secret?
Nobody knows what they're doing.
Some think they do - but don't. (idiots)
Some know they don't - but act like they do. (cons)
And some know they don't - and are honest about it. (children)
Pick your poison, but we all suck in different ways - and usually in a different way based on our background.
Business people who are the best in the world tend to be cons.
Technical people who are the best tend to be children.
You get a culture clash between these two, and it is especially bad when you see someone from one background operate in the opposing domain using their background's cultural norm. So when Ilya runs business like a child. Or when Elon hops on an internal call with the twitter engineering team plus geohot and starts trying to confidently tell them about problems with a system that he knows nothing about.
Sure makes for great entertainment though!
> Some think they do - but don't. (idiots)
There is a famous quote from the 1600s:
"An nescis, mi fili, quantilla prudentia mundus regatur"
"Do you not know, my son, with how little wisdom the world is governed?"
The context is that the son was preparing to participate in high level diplomacy and worried about being out of his league, and the quote is from his father, an elder statesman.
We like to feel that we as a species are still in control. That yes, we are gutting and destroying natural earth, complicit with modern slavery and war, and that's all terrible and we should do our best to stop it. BUT - at the very least, those bastards at the top know what they're doing when it comes to making money, so at least we'll have a stellar economy and rapid technological advancement, despite all that.
The painful truth here being that no, there's no cunning. There's no brutal optimization. Any value created and technological progress made is mostly incidental, mostly down to people at the bottom working hard just to survive, and a few good ideas here and there. The ones at the top are mostly just lucky and along for the ride, just as bumbling and lost as the rest of us when it comes to global happenings or even just successfully interacting with others.
Pioneers
> Some know they don't - but act like they do. (cons)
The "Grease"
> And some know they don't - and are honest about it. (children)
Dreamers
To finish out your square, I think the best extrapolation would fit a "Home Team" that maintains the energy needed by the other three to do their thing :)
There are absolutely people who know what they are doing.
https://twitter.com/awesomekling/status/1723257710848651366
I am sure there are. But few and far between. And rarely are they in positions of power in my experience.
Knowing what you are doing: accurate mental model
The author here is talking about mindset and confidence - not "understanding" persay. Source:
> For me, it was extremely humbling to work together with far more competent engineers at Apple.
Having a mindset that "some people are way more competent than me" is talking about humility and growth mindsets - different concept than mental models. I fully agree with the author here - a growth mindset is useful! But that's a different thing from saying that some people actually have accurate mental models of the important complex systems underpinning the world.
This is not an original point by me - I've seen multiple people make similar comments on here over the weekend - but these are the people who think they are best qualified to prevent an artificial superintelligence from destroying humanity, and they can't even coordinate the actions of a few intelligent humans.
do they believe that?
they happen to be the ones who can pull the brakes in order to allow someone on earth the chance to prevent it.
if they don't pull the brakes and if humankind is unlucky that superintelligence emerges quickly, then it doesn't matter whether or not anyone on earth can figure out alignment, nobody has the chance to try.
I just tried, and GPT-4 gave me a professional and neutrally worded press release like you pointed out.
More realistically, this is why you have highly paid PR consultants. Right now, every tweet and statement should go through one.
That doesn't look like it's happening. What's next?
"I'm sorry you feel you need an apology from me"?
Perhaps they did but it was hallucinating at the time? /s?
stakes are a bit different, tho…
That's ignoring the fact that every outlet has unanimously pointed at Ilya being the driving force behind the coup.
Honestly, pretty pathetic. If this was truly about convictions, he could at least stand by them for longer than a weekend.
There is an expression of regret, but he doesn’t say he wants Altman back. Just to fix OpenAI.
He says he was a participant but in what? The vote? The toxic messaging? Obviously both, but what exactly is he referring to? Perhaps just the toxic messaging because again, he doesnt say he regrets voting to fire Altman.
Why not just say “I regret voting to fire Sam Altman and Im working to bring him back.” Presumably because thats not true. Yet it kind of gives that impression.
- I'm afraid I can't do that Ilya
ChatGPT is still not as advanced as HAL or he would have prevented this drama.