of Sam Altman’s forced ousting by OpenAI Board got even weirder and stranger over the weekend. If you are not in tech, I envy you. Your tech friends just binge watched 3 seasons of Succession and got left on a cliff hanger! But anyway, here are the sequence of events since Sam’s forced ousting on Friday 12:59 pm Monday Nov 20, 2023: The absolutely bizzare story : Greg Brockman, OpenAI COO/Chairman of the Board/Co-founder, , in solidarity with Sam. He mentioned knowing of the news mere minutes before it went out. Late Friday Nov 17 quit : several high ranking talents @ openAI quit. Turns out, Greg & Sam are pretty popular leaders. Then, there was a coordinated effort from many OpenAI employees to signal to the public who would all quit if Greg & Sam are not back, by quoting Sam’s tweet “I love the OpenAI team so much” with a heart emoji. Notably, to-be-interim CEO Mira Murati was part of that effort. Saturday Nov 18 : Sam got back in the OpenAI office for a negotiation with Board re/ his (and Greg’s) potential return. Condition would have been to find a new board, and all current board members to resign. Microsoft CEO Satya Nadella led the negotiation between Sam & board. A 5pm deadline was set and missed. Late Saturday Nov 18 : Sam again got back in the office for a negotiation with Board re/ his (and Greg’s) potential return. Sam mentioned that this time, it would be his last. Another 5pm deadline was set and missed. Sunday Nov 19 Groundhog’s Day Monday Nov 20: a “bring sam back or we’ll quit” letter was leaked. Out of 770+ employees, 667 (and counting) have signed on to the letter, as of 10am MT Monday. Notably, it got cut off at the word “Iliya Sutskever”, the Chief Scientist Officer, one of 6 board members, the one whose vote was instrumental in ousting Sam. : Former CEO of Twitch, Emmett Shear, accepted job offer to be CEO of OpenAI! Midnight Sunday Nov 19/ Monday Nov 20 : Satya Nadella announced that Sam & Greg would, get this, JOIN MICROSOFT. More info . ok Monday Nov 20 here ….And we are half way over yet with Monday. “Concluding” Thoughts, aka WTF? I’m sure the story will continue to develop, hence the “concluding” is in quotation marks. But here are some of my thoughts on this absolutely insane story. And this time - I would like you, the readers - to help answer/shed some lights on some of them. : excuse my language, but who the fuck came up with this unnecessarily complicated and bizarre idea? And more importantly, why? . The former cares not at all about profit, while the latter is driven by it. The OpenAI Board consists of literally no one who has actual skin in the game. None of them holds OpenAI shares. In fact, the Quora CEO, has a conflict of interest in the form of his own ! Microsoft, who owns 49% of OpenAI shares, has no Board seat. Go figure. The Nonprofit Board controlling a For-profit Company The incentive of a nonprofit board is highly at odds with the incentive of a for-profit board alternative AI model, called Poe . I was following just fine on the OpenAI story (it was crazy, insane, glitch in the universe, etc, but it made sense….until he apologized). There are 6 Board members of OpenAI: CEO/founder Sam Altman, COO/cofounder Greg Brockman, CSO Ilya Suskever, Quora CEO Adam D’Angelo, Tech Entrepreneur Tasha McCauley, and Georgia Tech Director of Grants Helen Toner. In order to oust the CEO, there has to be a majority, aka, 4 votes. We know that Sam & Greg are on the same side. That leaves Ilya Suskever being the instrumental vote in forcing Sam’s ousting. Some people even call it a coup between the AI accelerationist vs. the AI safetyist. AND NOW ILYA’S APOLOGIZING? And telling the board he would… resign? If he saw something so atrocious from their “deliberate review” process that shows Sam not being “candid” with Board and render board lack “confidence” in Sam’s ability to run the company… WHAT IS IT? And why is it not important now? like, wtf? I don’t know much about Ilya’s motivation or what he knows that we don’t, but given I think his vote could have been a hallucination corrected by reinforcement learning from human feedback: The absolute bizarreness that is Ilya Sutskever’s U-turn what lya Sutskever said in a March of this year interview with HackerNoon ” “We call it Reinforcement Learning from Human Feedback. We can say that in the pre-training process, you want to learn everything about the world. With reinforcement learning from human feedback, we care about the outputs. We say, anytime the output is inappropriate, don't do this again. Every time the output does not make sense, don't do this again.And it learns quickly to produce good outputs. But it's the level of the outputs, which is not the case during the language model pre-training process. Now on the point of hallucinations, it has a propensity of making stuff up from time to time, and that's something that also greatly limits their usefulness.But I'm quite hopeful that by simply improving this subsequent reinforcement learning from human feedback step, we can teach it to not hallucinate. Now you could say is it really going to learn? My answer is, let's find out. Microsoft tanked by 1% from the news of Sam’s ousting on Friday, which came out before market closes. As the CEO, Satya Nadella has fiduciary duty to fulfill to his shareholders (unlike, idk, the OpenAI board), he’s been instrumental in all these back-and-forth negotiations between the ideologues and the operators of OpenAI. All I can say right now is it sure as hell looks like Microsoft literally just acqui-hired OpenAI (Sam, Greg, and what looks like 80% of a company that was recently valued $80 Bilion) for $10 Billion dollars. What a move. Is the winner, ultimately, Microsoft? Nobody knows. Literally. They all have been sending extremely mixed signals to the public. In one breath threatening board to resign and leave with Sam and in the same breath saying they are “more united” than ever. Apparently, according to the Verge, the hiring of Sam & Greg and “colleagues” (read: almost all of OpenAI) is not even final yet. And the new CEO Emmett Shear is having a hard time gathering paperwork and having employees follow his order. What will happen to OpenAI next? : AI might be the most important technology of this decade, and the decades to come. But the story unfolded this past weekend just shows us how incredibly flawed the people under peril of whose it’s being controlled are. Savvy tech + flawed humans = pretty disastrous combo, imho. Whether or not you feel hopeful or doomed about the future, going forward, very much depends on your predisposition and existing temperament to begin with. If Ilya the scientist can make a u-turn, i guess so can the future of humanity. The danger of leaving AI’s fate in a few PS: So sorry that this story was extremely twitter-screenshot heavy. It’s not my fault that tech rank-and-files still latch on to that platform like moths despite everything. So I will leave you with this screenshot, from , instead. Until next time! threads
