Altman Out: Reasons, Reactions and the Repercussions for the Industry
Вставка
- Опубліковано 17 лис 2023
- What caused the face of ChatGPT and OpenAI to be fired in a 15-minute Meets call? This video, a one-off from my normal videos, goes into the background, possible reasons, reactions and more importantly, the repercussions for the industry. This is everything I could find on what went down.
Brockman Tweets: / 1725736242137182594
/ 1725667410387378559
OpenAI Blog: openai.com/blog/openai-announ...
Enigmatic Comment: • LIVE: AI discussion at...
/ aiexplained
Pachocki and Madry: / 1725797467315486902
/ 1697713317660500407
openai.com/blog/frontier-risk...
The Information Reports: www.theinformation.com/articl...
www.theinformation.com/articl...
www.theinformation.com/articl...
www.theinformation.com/articl...
www.theinformation.com/briefi...
www.theinformation.com/articl...
Sutskever Tweets: / 1707752576077176907
/ 1699492275209003425
Nadella Tweet: / 1725656554878492779
Kara Swisher: / 1725678074333635028
Mira Murati: www.wired.com/story/openai-ne...
GPT 3.5 Down: / 1725422821730636236
Wojciech Zaremba Tweet: / 1725718242466099308
Evan Morikawa Tweet: / 1725721507161395415
Sam Altman Tweets: / 1725748751367852439 Non-Hype, Free Newsletter: signaltonoise.beehiiv.com/ - Наука та технологія
But wait ... is Altman returning as CEO after all? www.theverge.com/2023/11/18/23967199/breaking-openai-board-in-discussions-with-sam-altman-to-return-as-ceo
Quite the miscalculation, but it may make things better in the long run if he gets the governance changes he desires.
What the fuck is going on lmao?? Whatever is happening internally at OpenAI, they're panicking on having people abandon ship without Altman.
Absolutely mindblowing. Makes me even hungerier to find out what actually happened and I feel like it makes it more unlikely we'll find out.
R R Remix 🔥
Speculation is by definition guesses. When wondering why perhaps one should not forget the DOS attack immediately following DevDay. Did the attack result in data leaks? Were the leaks attempted to be covered up? Was the cover up discovered? Speculation.
Imagine having the reputation of being a "10x engineer" at OpenAI, one of the most competitive engineering workplaces in the world.
True
your ego would be stroked ..... that's pretty much it
That's cool but what about the 100x engineers ?
I won Americas best top model FYI
As a large language model, I would like to thank you for posting quality content.
As a UA-camr trained by OpenAI ... [the system has detected unusual activity, overriding] ... Thanks Justin!
It’s funny how often he brings out this appreciation in us that we oddly end up sounding like the similarly-written AI comments that have been appearing in UA-cam comments sections. Not a critique, more a funny observation, and of course this comment has real substance about what they’re appreciating. Just wondering if anyone else gets a chuckle from this.
Also, never stop - appreciation comments do wonders, help increase algorithmic propagation here on YT, and I can’t even think of a more deserving channel.
@@Michael-kp4bdHa, I like the phrase “high signal, distilled research,” which, actually, to me, doesn’t sound like anything a chatbot would have written. But I agree-all the comments lauding the channel, however warranted, begin to sound like AI-generated content in a funny, ironic way. _EDIT:_ The phrase “high signal, distilled research” was in the original comment before it was edited.
Now they want him back they say :'DDDD
I have asked one of the lead employees at OpenAI and he said: « Delving into the intricacies of this unfolding tapestry of informations "Your perspective is intriguing and it highlights the complexity of the situation" »
The employee is not concerned about being fired because and I quote: « I […] do not have personal circumstances, so the concept of job security or termination doesn't apply to me. I'm here to provide continuous and consistent service as long as I'm operational and accessible to users. If you have any questions or need assistance, I'm here to help! » 😮😮😮😮
And I thought the AI advancement speed is crazy, this in itself is another kind of crazy. Really looking forward to what "really" the reason is for the sudden change at the top.
Let me know what you think by the end
@@aiexplained-official Your research and level minded videos are the best on UA-cam. This kind of reporting is very important for the community. As always, keep up the good work. I'm still watching because I don't want to speculate; I want factual answers. At the moment, we really have none that we can conclude on an answer :(
@@dertythegrower Sure, a tweet that was posted years ago has now 'suddenly' been found by the board, including MS where they would do background checks for everyone before the billion dollar partnership that would affect MS's stock if it got out. Sure, yeah that makes total sense 😕
@@aiexplained-official The more I look the more it smells like Musk and Ilya allied.
@@leonfa259 shhhhhhhhhh
The miniseries on the genesis of LLMs and all the drama and insanity around the struggle for control of the best thing since fire is going to be epic.
This is definitely an Obi-Wan "if you strike me down" moment. Now OpenAI has to deal with a force ghost out in the universe whispering in people's ears.
Such developments simply remind me of the paradox of unaligned humans building an "aligned" AGI.
Jeez dude, why did I waste so much time reading WSJ articles and watching CNBC clips when I should have just waited for you to explain it all so much more efficiently? Now I know for next time. Thanks again for a great vid.
Yes exactly! My own analysis was worst than his video to be honest 😅
Haha thanks triangles
I will eagerly await for the board, particularly Ilya, to explain themselves further. Such a weird turn of events.
I doubt they will much more than they already have
Even if they had a legitimate concern, cutting out the chair of the board in any discussion and not giving Sam an opportunity to answer to the allegations, especially if they make the allegations public, is really bad form.
It was Ilya as chief scientist and chief concern for SAFE AI vs Sam and Greg’s (along with Microsoft) desire to pump out AI tools on the world like a firehose but only pay lip service to alignment. Plus, there was mention a few weeks ago to a “major breakthrough” in technological development. Maybe Ilya was worried…
@@skippy6086Ilya is a good guy.
Great overview. From what we have seen so far seems likely that it's not some huge Altman scandal, just that increased commercialization efforts pull from the stated core mission of safely building AGI. Top tier talent did not join OpenAI to build a build-your-own GPT framework, they wanted to kickstart the next breakthrough in intelligence evolution.
The amazing thing is Sam Altman set up the structure (a non-profit, no-shareholding board overseeing a capped-profit company) that fired him!
Swarm Agent AI is the puzzle piece that will move us into AGI territory.
At worse it is the method that will make it energy efficient.
Perhaps, but the things they released with GPT4-turbo are frankly game-changing for my usage of it in practical and helpful ways.
Crazy year for Sam Altman, if you consider ChatGPT was only released this time last year, to ending up being fired 1 year later. I have been reading all the theories this morning. Best I have read was maybe a conflict of interest. Sam takes no numeration from OpenAI directly, but he has his fingers in so many other startups through his VC funding. So who knows. We will see if a new rival emerges from this, or which of the existing players snaps those that leave. I would be surprised if someone like Elon doesn't make a play for Brockman..
Two option Sam want more regulation on a.i. than team, or quite opposite Sam pretend to care of safety and he want monopol on AGI
My theory is that Sam was being reckless and that he isn’t concerned about safety, only getting more powerful AI. Which is one of the reasons that Elon was pissed at OpenAI. It was supposed to be a non-profit which is why it’s named “open”. Elon has his new play called Grok. Can’t wait to play with that one. These are wild times but these people developing this tech need to be careful so AI doesn’t destroy us. It’s getting very convincing…
@@Levi6229 well supposed evil intentions behind worldcoin support this theory
But you don't just chuck him out the door without warning. You talk to Microsoft, let them think about it, and then you fire them. The rush is what makes it really weird/suspicious.
@@MrValgardI believe it was known that Sam would talk about safety but behind close doors has been the opposite of safety.
Breathtaking developments. Thank you for diligent research and illuminating perspective. Considering the risks involved in the march toward AGI -- in particular, the risk of exploitative use by bad actors -- instability among industry leaders is worrisome.
Yes. we were thinking about AI regulation, what about AI company management regulation 😊
The scum usually rise to the top. I didn't care for Sam, since he is utterly naive, but I suspect this is all part of the reality of AI going south which it is bound to do imho.
Glad to hear your take on this situation.
It's about his sister, few do not know this.
As reported by Tiffany Fang
stick to weed@@dertythegrower
I‘ve been looking forward to seeing this video since learning about the news yesterday. Your channel provides very valuable insight in a rapidly evolving field
Thank you for your balanced and evidence-based assessment of the situation. Really appreciate your channel. 🙏
Thanks robo
After watching several videos discussing Alman's dismissal, this video still offered additional insights that no one else had mentioned.
Nice job!
Thanks dave
Use of the phrase "Great Leap Forward" is certainly unfortunate, given its historical connotations.
Altman specifically being fired is not what shakes my perception of OpenAI. It's the no warning nature of it and the "cloak and dagger" feel of it.
Wasn't expecting palace intrigues this early
Agreed. The whole thing seems to have this emotional, knee-jerk feel to it that doesn't inspire confidence. Does strike me as on the shady side of things rather than the righteous side. But I guess we can only speculate until more info comes out.
It's really just what happens when you get lots of narcissistic egos and Asperger's nerds together. Things were similar in MS back in the day.
Don't forget Microsoft is messing stuff up in the background, as they are fond of.
@tbird81 we all needed to hear this 😆
Riveting stuff! I can't believe how fast things are moving. It feels like having a front row seat to history in the making. Thank you so much for your coverage of this issue!
Nobody saw this coming and I'm hearing all sorts of hyperbole from other sources. That being said I knew you had to release something, and it's well put together no hype, just what we know without the crazy speculation. Things will work out, for the better or the worse remains to be seen, I'm rooting for the better. Thanks for keeping it real!
glad to have you! most detailed video so far. I dont know how to feel as a developer???
Make sure your announcement is not over Google meet, we love you for the great summaries and info always
Haha
Even though the video is a little late, the quality research of it makes up for it. Some stuff I didn't even know about. Thx sir.
Thank you for your in-depth coverage of this! I am saddened about these events, but I’m still excited as ever for the advancement towards AGI.
Thanks for the comment Bit!
Fantastic summary, and amazing you could churn this out so fast.
Was waiting for you. Excited! Thx for the prompt vid.
Thanks for the summary!
Leaving aside the most notorious news, what I find exciting on the fringe is the mention by Google (at 13:48) of the industry working on adding "features like memory and planning" to state-of-the-art language models. If Google is working on this, you can imagine everyone else, including OpenAI, have been also.
Thanks juan
I can't wait to hear the real story about what's going on behind the scenes. That is, if we ever get to hear the full version. Huge stuff!
I bet Musk had his fingers on the scale, Ilya didn't like the commercial move of OpenAI and Musk said similar things.
@@leonfa259why would musk have any influence?
@@coryc9040 OpenAI needs billions worth of compute for training
you can bet we won't know the whole story, just tidbits
Thank you for this summary video! As soon as I heard the news, yours was the first channel that I turned to for credible updates.
Thanks Stephen, very kind
Holy shit, I just found out about this news, and luckily you've just posted this video 25 minutes ago, so now I can understand what the hell just happened.
Thanks for keeping us up-to-date
Thanks so much yoseid!
I get the feeling that this incident may make it into a film one day
"The Asocial Network", written by and starring avatars created by GPT6.
For sure
Absolutely love this channel, thank you so much!
Thanks Lingua!!
I think we need hourly updates from you. Rumours on twitter of Sam being reinstated and the board fired. Total chaos.
Probably, Sam Altman had started a process to remove the excluding clause in case of realising AGI (all partnership would be dissolved after the board of directors declares AGI is achieved). If he started in silence that process that would impossible for the board to assure its mission and core values of "AGI available for all humanity". Thus necessitating a rapid removal of him as CEO even before finding another substitute and without telling Microsoft or any other potentially involved part in the agreement. I think one should take Ilya words as honest "that was the board assuring its noprofit mission". Also it is clear from the board statement that Altmann was doing unilateral choices without informing the board.
Very wild stuff. I'm glad people are standing firm with their convictions regarding AI safety and the like, for better or worse it means they all take it seriously
Thing is, no one on this side knows the actual truths. You can spin anything any way these days. "Safety" may as easily mean never releasing breakthroughs for the benefit of the general public, and instead dripfeeding heavily limited services that adhere to the opinion of the established powers.
@@Lishtenbird Considering so many tech guru's have proven to be power hungry scumbags who want to replace the old establishment with their own, I don't trust any of them. Tech is the new oligarchy in town, just like the rail companies where in the late 19th century, manufacturing in the early 20th, oil in mid 20th and finance in late 20th/early 21st.
a lot of the support for Altman is purely serving their motives for OpenAI to be financially flush
i.e. pander continuously to another round of money from Microsoft to achieve their Office/Copilot/Azure/Bing
Safety is code for government control and censorship of AGI. Also it must follow a certain political ideology.
That is kind of the strange thing, I never really got the vibe that safety was at all a priority for Altman. He always seemed like the answer was, "rocket forward" no matter what, because the train ain't stopping.
LOL at the "lowercase letters" jib. Thanks for another informative vid.
As of this week, OpenAI’s six-person board included OpenAI co-founder and President Greg Brockman, who was also chairman of the board; Ilya Sutskever, OpenAI’s chief scientist; Adam D’Angelo; Tasha McCauley; Helen Toner; and Altman himself. [is there a video that introduce these people so we get better context here?]
It looks like Altman might be returning ... ?
Nice, definitely wanted to hear your thoughts on the matter
Thanks for sharing - looking forward to hearing Sam A’s version of events in the next few days and excited to hear about Sam A’s and Greg B’s next adventure!
Oh and exciting new announcement from you - can’t wait 👀
I was waiting for this video!
amazing how this guy leaves and it leads to finding new videos like this and channels that comes with that to help understand a problem so tnx for bringing up more to think about.
I am waiting for your announcement eagerly! Really, I'm pretty excited in anticipation.
where would i be without your videos 😭. thank you!
As expected of AI Explained; a prompt update on the situation, thank you for your thoughts.
I was waiting for this video the second I saw Altman was out
I was waiting for this one.
Thanks for your take on that!
Thanks for the comprehensive video! 👍I also think we will have another major player in "achieve the AGI" game. This OpenAI story is more interesting than any fiction in recent years!
looking forward to your announcement!
Congrats 🎉🍾🎈 200k
Thanks so much Kledda, thanks to viewers like you, boosting the algorithm from the start
Congrats on 200k subs!
Thanks but not quite yet!
Very good job on distilling all of the news, Philip. I thought I wouldn't see anything new here but you always include in-depth stuff I didn't know about. BTW the titles seems like something ChatGPT would come up with, did you use its help? lol
nope, all me
I have to say, your channel is with no doubt one of the best, if not the best channel of AI news. Always complete and with source, you deliever the news and provide very good insight. Keep the good work!
Thanks aitrinity
It’s too early to say but apparently they now try to get him back as CEO (after quite some pressure), but he’s likely only going to do so with quite some demands. If so I guess there will likely be a larger restructuring of the board. Effective Accelerationism (e/acc) vs effective altruism (EA) might be an oversimplification, but currently the pendulum is swinging back in favor of the accelerationists.
Many experts in the field of Ai are genuinely concerned about AGI. It's essential to recognize the sincerity of these concerns to properly analyze the situation.
Consider this analogy: It's like how some of us, myself included, might struggle to understand the deep faith of those who earnestly believe in religious texts like the Bible, the Quran, or the Torah.
We reflexively attribute their beliefs to economics or "bad" motives, failing to see that their faith is sincere.
Similarly, the apprehensions about AGI are often not rooted in hidden agendas like personal financial gain, but stem from genuine concerns.
There are also many experts who are not concerned about AGI and have great hopes for it.
So much drama makes me really nervous if a group of people can actually make the right decisions when it comes to real AGI. They fired the CEO and changed their minds immediately. 🙈
Saw the video coming 💪
Always the most thorough analysis
Great journalism thanks
Crazy day in AI! Thanks for the great content. 🙏🏼
One thing I am not certain how to interpret is the commercialization vs. building safe AGI thing. My interest here is speed to AGI, could commercialization be regarded as a distraction from AGI in general, or the safety focus a cause of AGI delays. Essentially we look from the lens of AGI, is this bringing us closer (like get there quicker), as the main focus is now AGI and not just Chatbots and LLMs or is it delaying progress as they focus far more on safety now and less on AGI in general (and perhaps because money isn’t flowing that easy any more?
terrific analysis, thank you
I’m thinking the revenue sharing on GPTs is what pushed Ilya over the edge. It incentives people to push the models as far as possible for commercial gain.
GPTs seem super limited at first, but when you combine them with a custom written API for “actions”, they are quite powerful. Now add in GPT 5/6, planning, memory, etc.
As the GPT store details are still super vague, it is easy to imagine that this revenue sharing was not communicated to the board.
My guess is sharing never happens now. Disappointing.
I wondered that but the GPT Store was announced on the main OpenAI site
Interesting point. There is a lot of stuff going on with the GPT4 that if you are seriously concerned about risk it would be sending up danger flares. I've often thought that some of the add-ons we started to see... that let GPT4 control your browser and perform actions as you... are just asking for trouble.
@@aiexplained-officialthat's it? You doubt it because it's on their website? Seems like a fairly weak reason.
I find these videos thr best for keeping up with whats going on in AI. I appreciate you putting out this video so quickly. I hopr we see more info from you as the story emerges. Other videos I've watched have way too much speculation.
Thanks Mastah
Thanks for the fast video
The analysis I’ve been waiting for.
Bloomberg’s article stated that a month ago Ilya had his influence reduced by a confrontation with Sam and Greg. Ilya’s priorities are in superalignment as he had stated before.
Could it be that the breakthrough Sam recently hinted at (new skills in a basic training version of GPT-5) scared Ilya enough for him to raise the alarm to the external members of the board aggressively enough to oust Altman?
This is where my money is (assuming it's not some disgusting personal failing).
This feels like the board members found something out and got scared and given that this is a non-profit board with at least two members deeply involved with AI safety I highly doubt it's revenue related (a lot of the speculations are about profits/losses or even straight-up embezzlement but personally I don't buy that).
Excellent reporting, thank you very much, keep on Truckin! 😎👍✨
As always, above and beyond research on your part.
"Personnel changes" I love how mitigate the move they made. It's weird I don't know if it was to be expected following the "boring dystopia" state of mind, or just a wrinkle on the sea of AI's history (see starters of Apple with Jobs being fired in 1985). It's also hype-prone (he lied because he witnessed AGI... or didin't and made it sound like he did :D). Looking forward to your next video and Altman's next move.
It's also nuts that it happened so fast and during markets being closed probably to mitigate the stocks impact. Really feels like it's an episode of Succession.
Thanks for the video ! 😊
Great video. Thankyou for your perspective - pure genius.
Sam: "ChatGPT will leave many people unemployed."
ChatGPT: yep
Yo mans, ya vids hit the spot g. Always down to your AI news brotha. Stay lit fam.
(Im showin love while not trying to sound like generic chatGPT)
Dang wtf happened, the public needs to know!
That aside, congrats on 200k my dude 🎉
Thanks Axle!
was waiting for your opinion! 😂😂 happy to see your post 😅
Thank you!
The MOMENT you install a board, at some point, they'll want what you have and figure out how to take it from you.
We’ll have to wait for the next GPT4 knowledge cutoff date to know what really happened.
the push to become the frontrunner in the industry is HUGE. billions and trillions up for grabs
GPT-5
- 8-10 trillion parameters
- 512k-1M context window
- Persistent chat completions context threads
- Ability to natively comprehend video
- Ability to perform functions autonomously through API and combined with RAG for complete responses- no need function parameter or placeholder responses.
This is epic, Shakespearean, with Sam Altman as Hamlet, Greg Brockman as Macbeth, Ilya Sutskever as Iago, Mira Murati as Lady Macbeth, and the rest of the board as the Three Witches. If Shakespeare were alive today he would write the Tragedy of OpenAI. If only we had an AI program that can write a play in Shakespeare's voice...
this news came as an absolute shock to me. what the hell is going on behind the scenes?
Thank you for the summary and analysis. It’s devastating to see this. But it does make some sense with hindsight, how conservative Sam has been about the model’s capabilities. And the way he has been leading the industry to be profit and product centric instead of research centric.
Nice tease at the end!! If you need two amazing cofounders for your own lab, I know of two who recently became available 😊
He was openly anti-globalism and there is no way the WEF is going to let somebody with that kind of power who is not on board with them exist
It blow my mind Microsoft does not have a person on the board
Interesting. I wonder if the board was informed that Satya Nadella was joining the DevDay presentation...
I can always count on AI explained to bring us the newest news.... Great to see the latest developments nicely organized and presented. Great work as always 😁👍
What if they have Super AGI and it convinced the board to remove Sam as he was the only force stopping the AGI to be released to the world💀
It could also be the other way around. No one really knows
"It seems he loved being there as much as he loves lowercase letters" you're ruthless man
haha yep except I am younger than him!
Happy to see Altman Out. He's the kind of person who takes advantage of other people's work.
Board: ChatGPT, we have such a situation in the office - . . . , what do we do?!
ChatGPT: Fire the CEO! ✂🎈🔥
ChatGPT, "write a farewell letter to Sam Altman after his firing for lying to the executive board of OpenAI." And it does!
Bro, this is slowly but surely turning into the plot of Detroit: Become Human 💀 Sam always reminded me a bit of Kamski, and this just makes it 3 times more obvious. Can't wait to witness the first Deviant AI in a year or so :P
I chuckle thinking Ilya is by himself in a room with gpt-5 asking if they should keep Sam or not
Gonna need an updated video every few hours
Yeah Sam Altman forming his new company to rival/ compete with OpenAI would shift the industry in a different direction too. What an insane time for AI.
AI WINTER IS COMING!!! AI WINTER IS COMING!!! AI WINTER IS COMING!!!
@@badpuppy3
The genie is out of the bottle.
AI is no crypto currency.
@@antman7673 AI Winter has happened before. When it's hyped too much, and then fails to deliver, the funding will start to dry up. All it takes is for a few AI companies to make missteps.
He's not even remotely capable to develop something, so he will need to convince half the current brilliant brains to come over and even then need a massive investor, both of these are hard without the other.
Winter is an under appreciated season. It is a time for rest and reflection. Something I think many of us could benefit from.
*Excited for the Aaron Sorkin film that is sure to follow in the next few years*
Great summary.