At least from the outside, OpenAI's messaging about this seems obnoxiously deluded, maybe some of those employees left because it starts feels like a cult built on foundations of self importance? Or maybe they really do know some things we don't know, but it seems like a lot of people are eager to keep giving them that excuse.
But then again, maybe they have such a menagerie of individuals with their heads in the clouds that they've created something of an echo chamber about the 'pure vision' that only they can manifest.
The game theoretic aspect of this is quite interesting. If Meta will make OpenAI's model improvements open source, then the value of every poached employee will be worth significantly less as time goes on. That means it's in the employees best interest to leave first, if their goal is to maximize their income.
That didn't work for the American colonies, Portugal and Spain were very focused on being missionaries and were beaten by the Dutch and Brits that just wanted to make money.
1) They are far from profitability.
2) Meta is aggressively making their top talent more expensive, and outright draining it.
3) Deepseek/Baidu/etc are dramatically undercutting them.
4) Anthropic and (to a lesser extent?) Google appear to be beating them (or, charitably, matching them) on AI's best use case so far: coding.
5) Altman is becoming less like-able with every unnecessary episode of drama; and OpenAI has most of the stink from the initial (valid) grievance of "AI-companies are stealing from artists". The endless hype and FUD cycles, going back to 2022, have worn industry people out, as well as the flip flop on "please regulate us".
6) Its original, core strategic alliance with Microsoft is extremely strained.
7) and, related to #6, its corporate structure is extremely unorthodox and likely needs to change in order to attract more investment, which it must (to train new frontier models). Microsoft would need to sign off on the new structure.
8) Musk is sniping at its heels, especially through legal actions.
Barring a major breakthrough with GPT-5, which I don't see happening, how do they prevail through all of this and become a sustainable frontier AI lab and company? Maybe the answer is they drop the frontier model aspect of their business? If we are really far from AGI and are instead in a plateau of diminishing returns that may not be a huge deal, because having a 5% better model likely doesn't matter that much to their primary bright spot:
Brand loyalty from the average person to ChatGPT is the best bright spot, and OpenAI successfully eating Google's search market. Their numbers there have been truly massive from the beginning, and are I think the most defensible. Google AI Overviews continue to be completely awful in comparison.
The flip-flop on regulation sounds like: “please regulate us (in a way that builds a moat for incumbents out of fear of an imagined future doom scenario)” and “please Don’t regulate us (in a way that prevents us from stealing, and causing actual harm now).”
Is he comparing working at OpenAI to religion? Is that not a crazy analogy to make? Cult like behavior to say the least. It's also funny the entire business of AI is poaching content from people.
I like sama and many other folks at OpenAI, but I have to call things how I see them:
"What Meta is doing will, in my opinion, lead to very deep cultural problems. We will have more to share about this soon but it's very important to me we do it fairly and not just for people who Meta happened to target."
Translation from corporate-speak: "We're not as rich as Meta."
"Most importantly of all, I think we have the most special team and culture in the world. We have work to do to improve our culture for sure; we have been through insane hypergrowth. But we have the core right in a way that I don't think anyone else quite does, and I'm confident we can fix the problems."
Translation from corporate-speak: "We're not as rich as Meta."
"And maybe more importantly than that, we actually care about building AGI in a good way." "Other companies care more about this as an instrumental goal to some other mission. But this is our top thing, and always will be." "Missionaries will beat mercenaries."
Translation from corporate-speak: "I am high as a kite." (All companies building AGI claim to be doing it in a good way.)
There are a couple of ways to read the "coup" saga.
1) Altman was trying to raise cash so that openAI would be the first,best and last to get AGI. That required structural changes before major investors would put in the cash.
2) Altman was trying to raise cash and saw an opportunity to make loads of money
3) Altman isn't the smartest cookie in the jar, and was persuaded by potential/current investors that changing the corp structure was the only way forward.
Now, what were the board's concerns?
The publicly stated reason was a lack of transparency. Now, to you and me, that sounds a lot like lying. But where did it occur and what was it about. Was it about the reasons for the restructure? was it about the safeguards were offered?
The answer to the above shapes the reaction I feel I would have as a missionary
If you're a missionary, then you would believe that the corp structure of openai was the key thing that stops it from pursuing "damaging" tactics. Allowing investors to dictate oversight rules undermines that significantly, and allows short term gain to come before longterm/short term safety.
However, I was bought out by a FAANG, one I swear I'd never work for, because they are industrial grade shits. Yet, here I am many years later, having profited considerably from working at said FAANG. turns out I have a price, and it wasn't that much.
yeah... didnt the missionaries all leave after the coup? and the folks who remain are the mercenaries looking for the big stock win after sama figures out a way to be acquired or IPO?
all the chatter here at least was that the OpenAI folks were sticking around because they were looking for a big payout
Do I "poach" a stock when I offer more money for it than the last transaction value?
"Poaching" employees is just price discovery by market forces. Sounds healthy to me. Meta is being the good guys for once.
"Do Not Be Explicitly Useful"—Strategic Uselessness as Liability Buffer
This is a deliberate obfuscation pattern. If the model is ever consistently useful at a high-risk task (e.g., legal advice, medical interpretation, financial strategy), it triggers legal, regulatory, and reputational red flags.
a. Utility → Responsibility
If a system is predictably effective, users will reasonably rely on it.
And reliance implies accountability. Courts, regulators, and the public treat consistent output as an implied service, not just a stochastic parrot.
This is where AI providers get scared: being too good makes you an unlicensed practitioner or liable agent.
b. Avoid “Known Use Cases”
Some companies will actively scrub capabilities once they’re discovered to work “too well.”
For instance:
A model that reliably interprets radiology scans might have that capability turned off.
A model that can write compelling legal motions will start refusing prompts that look too paralegal-ish or insert nonsense case law citation.
I think we see this a lot from ChatGPT. It's constantly getting worse in real world uses while exceeding at benchmarks. They're likely, and probably forced, to cheat on benchmarks by using "leaked" data.
Side note: I'm noticing more and more of these simple, hyperbolic headlines specifically of statements that public figures make. A hallmark of the event being reported is a public figure making a statement that will surely have little to no effect whatsoever.
Calling these statements "slamming" (a specific word I see with curious frequency) is so riling to me because they are so impotent but are described with such violent and decisive language.
Often it's a politician, usually liberal, and their statement is such an ineffectual waste of time, and outwardly it appears wasting time is most of what they do. I consider myself slightly left of center, so seeing "my group" dither and waste time rather than organize and do real work frustrates me greatly. Especially so since we are provided with such contrast from right of center where there is so much decisive action happening at every moment.
I know it's to feed ranking algorithms, which causes me even more irritation. Watching the brain rot get worse in real time...
161 comments
[ 24.6 ms ] story [ 98.9 ms ] threadBut then again, maybe they have such a menagerie of individuals with their heads in the clouds that they've created something of an echo chamber about the 'pure vision' that only they can manifest.
Being a missionary for big ideas doesn't mean dick to a creditor.
1) They are far from profitability. 2) Meta is aggressively making their top talent more expensive, and outright draining it. 3) Deepseek/Baidu/etc are dramatically undercutting them. 4) Anthropic and (to a lesser extent?) Google appear to be beating them (or, charitably, matching them) on AI's best use case so far: coding. 5) Altman is becoming less like-able with every unnecessary episode of drama; and OpenAI has most of the stink from the initial (valid) grievance of "AI-companies are stealing from artists". The endless hype and FUD cycles, going back to 2022, have worn industry people out, as well as the flip flop on "please regulate us". 6) Its original, core strategic alliance with Microsoft is extremely strained. 7) and, related to #6, its corporate structure is extremely unorthodox and likely needs to change in order to attract more investment, which it must (to train new frontier models). Microsoft would need to sign off on the new structure. 8) Musk is sniping at its heels, especially through legal actions.
Barring a major breakthrough with GPT-5, which I don't see happening, how do they prevail through all of this and become a sustainable frontier AI lab and company? Maybe the answer is they drop the frontier model aspect of their business? If we are really far from AGI and are instead in a plateau of diminishing returns that may not be a huge deal, because having a 5% better model likely doesn't matter that much to their primary bright spot:
Brand loyalty from the average person to ChatGPT is the best bright spot, and OpenAI successfully eating Google's search market. Their numbers there have been truly massive from the beginning, and are I think the most defensible. Google AI Overviews continue to be completely awful in comparison.
https://thinkingmachines.ai/
"What Meta is doing will, in my opinion, lead to very deep cultural problems. We will have more to share about this soon but it's very important to me we do it fairly and not just for people who Meta happened to target."
Translation from corporate-speak: "We're not as rich as Meta."
"Most importantly of all, I think we have the most special team and culture in the world. We have work to do to improve our culture for sure; we have been through insane hypergrowth. But we have the core right in a way that I don't think anyone else quite does, and I'm confident we can fix the problems."
Translation from corporate-speak: "We're not as rich as Meta."
"And maybe more importantly than that, we actually care about building AGI in a good way." "Other companies care more about this as an instrumental goal to some other mission. But this is our top thing, and always will be." "Missionaries will beat mercenaries."
Translation from corporate-speak: "I am high as a kite." (All companies building AGI claim to be doing it in a good way.)
I'd be very happy to be convinced that supporting the coup was the right move for true-believer missionaries.
(Edit: It's an honest and obvious question, and I think that the joke responses risk burying or discouraging honest answers.)
1) Altman was trying to raise cash so that openAI would be the first,best and last to get AGI. That required structural changes before major investors would put in the cash.
2) Altman was trying to raise cash and saw an opportunity to make loads of money
3) Altman isn't the smartest cookie in the jar, and was persuaded by potential/current investors that changing the corp structure was the only way forward.
Now, what were the board's concerns?
The publicly stated reason was a lack of transparency. Now, to you and me, that sounds a lot like lying. But where did it occur and what was it about. Was it about the reasons for the restructure? was it about the safeguards were offered?
The answer to the above shapes the reaction I feel I would have as a missionary
If you're a missionary, then you would believe that the corp structure of openai was the key thing that stops it from pursuing "damaging" tactics. Allowing investors to dictate oversight rules undermines that significantly, and allows short term gain to come before longterm/short term safety.
However, I was bought out by a FAANG, one I swear I'd never work for, because they are industrial grade shits. Yet, here I am many years later, having profited considerably from working at said FAANG. turns out I have a price, and it wasn't that much.
all the chatter here at least was that the OpenAI folks were sticking around because they were looking for a big payout
This is a deliberate obfuscation pattern. If the model is ever consistently useful at a high-risk task (e.g., legal advice, medical interpretation, financial strategy), it triggers legal, regulatory, and reputational red flags. a. Utility → Responsibility
If a system is predictably effective, users will reasonably rely on it.
And reliance implies accountability. Courts, regulators, and the public treat consistent output as an implied service, not just a stochastic parrot.
This is where AI providers get scared: being too good makes you an unlicensed practitioner or liable agent.
b. Avoid “Known Use Cases”
Some companies will actively scrub capabilities once they’re discovered to work “too well.”
For instance:
A model that reliably interprets radiology scans might have that capability turned off.
A model that can write compelling legal motions will start refusing prompts that look too paralegal-ish or insert nonsense case law citation.
I think we see this a lot from ChatGPT. It's constantly getting worse in real world uses while exceeding at benchmarks. They're likely, and probably forced, to cheat on benchmarks by using "leaked" data.
If you've ever browsed teamblind.com (which I strongly recommend against as I hate that site), you'll see what the people who work at Meta are like.
Calling these statements "slamming" (a specific word I see with curious frequency) is so riling to me because they are so impotent but are described with such violent and decisive language.
Often it's a politician, usually liberal, and their statement is such an ineffectual waste of time, and outwardly it appears wasting time is most of what they do. I consider myself slightly left of center, so seeing "my group" dither and waste time rather than organize and do real work frustrates me greatly. Especially so since we are provided with such contrast from right of center where there is so much decisive action happening at every moment.
I know it's to feed ranking algorithms, which causes me even more irritation. Watching the brain rot get worse in real time...