A CEO typically builds up a network of his people within the org and if he falls hard they are next on the chopping block. Same deal as with dictators. "Dozens" sounds like about right amount for a large org.
So having Altman's loyalists leave is probably exactly what Sutskever wants? Still, what do they actually want? It seems a bit overly dramatic for such an organisation.
I think it's pretty obvious after reading it why people who were really committed to that Charter weren't happy with the direction that Sam was taking the company.
Let's see whether Satya Nadella's bet on that risk will pay or not. Chance is a "biased random" in the real world. Let's see whether his bias is strong enough to make a difference.
Are you talking about OpenAI or about Sam Altman's hypothetical new company? OpenAI already had the best technology fully developed and in production when Microsoft invested in them. I believe "cargo cult" means something quite different to how you're using it. It's not "cargo cult" to consider someone's CV when you hire them for a new job. Sam Altman ran a successful AI company before and he most likely can do it ag…
> Are you talking about OpenAI or about Sam Altman's hypothetical new company?
About him and Greg joining to Microsoft.
> I believe "cargo cult" means something quite different to how you're using it.
I don't think so.
Tribes believed that building wooden air strips or planes would bring the goods they have seen during wartime.
People believe that bringing Altman will bring the same thing (OpenAI as is) exactly where it's left off.
Altman is just tip of the iceberg. Might have some catalyst inside him, but he's not the research itself or the researcher himself.
The whole affair has been bizarre. Professionals discuss all this stuff out of the public eye and execute when they have a plan. Every single move from all parties here has been run and gun and looks like amateur hour. These people do business like the characters on Succession.
I think trump and elon sort of made it ok to showcase the shit online. Like how YC got involved with fighting other VCs on twitter. I have huge respect for YC and never thought they would do that, but here we are. Again a stark reminder that all of these guys from Ray Dalio to the run of the mill SF VC are all just normal, twisted people who don't know much better about anything and merely had a run of good luck. Sto…
Or in other words, there’s no downside to narcissism when you can broadcast it and accumulate loyal followers who will ply you with money and fame
I still cannot process what’s happened to one of the most prominent and hyped companies of the past year in just one weekend. If it’s true that Altman won’t return to OpenAI (or alternatively: that the current board won’t step down) then where does that leave OpenAI? Microsoft can’t be happy, as evidenced by reporting that Nadella was acting as mediator to bring him back. Does OpenAI survive this? Will be super inter…
No, OpenAI will not survive as a company with more than one shareholder. At the end of the day, MSFT has a fiduciary duty to its own shareholders. MSFT has set certain expectations for its own financial performance based on its agreements with OpenAI and MSFT shares traded based on those expectations. Now OpenAI has sustained a hemorrhage of its leadership that negotiated those agreements, including a public admissio…
“Sorry, we are reporting a write down of $10 billion due to potential misrepresentations of commercial intent that occurred in our OpenAI portfolio.”
Things you will never hear Satya Nadella say. Way more likely he will coordinate to unify as much of their workers as he can to continue on as a subsidiary, with the rest left to go work something out with other players crazy/desperate enough to trust them.
Oh wow, she’s definitely taken sides. I wonder if Elon will be paying their legal bills. She alludes to shadowy machinations. This reaches an end, then gets wackier, and repeats.
Kara Swisher is so unhinged, she thinks Emmett Shear, founder and CEO of Twitch, is "being generous here, less than impressive in comparison to" Mira Murati...
Noticed that. She did soften that stance later. She seemed oddly invested.
In time of Windows, around let's say mid 1990s, people thought Windows is irreplaceable. Now turns out Linux is the workhorse everywhere for running workloads or consuming content. Almost every programming language (other than Microsoft's own SDKs) gets developed on Linux, has first class support for Linux and Windows is always an afterthought. It has gone to that extent that to lure developers, Microsoft has to embe…
Linux isn't the workhorse in any business that isn't tech based. The dev bubble here is pretty strong. I've done IT for a couple MSPs now so I've seen 100s of different tech stacks. No one uses Linux for anything. ESXi for the hypervisors, various version of Windows server, and M365 for everything else. Graphics/marketing uses Macs sometimes but other than that, it's all Windows/MS. Seeing a Linux VM is exceeding rar…
True. You'll find Windows XP based terminals on many industrial machines. Its pervasive but outnumbered where "running the workloads" comea into picture.
The dev bubble is not that small. This very website is I'm pretty sure not served from Windows.
Other than stack overflow or few handful of exceptions, very little is actually served from Windows if I'm not wrong.
Honest question: Other than 1) Microsoft and 2) anyone building a product with the OpenAI api 3) OpenAI employees… …is OpenAI crashing a burning a big deal? This seems rather over hyped… everyone has an opinion, everyone cares because OpenAI has a high profile. …but really , alternatives to chatGPT exist now, and most people will be, really… not affected by this in any meaningful degree. Isn’t breaking the strangle h…
I've not found anything that really competes with GPT4, and that's been released for some time.
> Isn’t breaking the strangle hold on AI what everyone wanted with open source models last week?
By other things getting better, not by stalling the leader of the pack.
This is exactly why you would want people on the board who understand the technology. Unless they have some other technology that we don't know about, that maybe brought all this on, a GPT is not a clear path to AGI. That is a technical thing that to understand seems to be beyond most people without real experience in the field. It is certainly beyond the understanding of some dude that lucked into a great training s…
>Unless they have some other technology that we don't know about, that maybe brought all this on, a GPT is not a clear path to AGI. So Ilya Sutskever, one of the most distinguished ML researchers of his generation does not understand the technology ? The same guy who's been on record saying LLMs are enough for AGI ?
To be clear, he thinks that LLMs are probably a general architecture, and thus capable of reaching AGI in principle with enormous amounts of compute, data, and work. He thinks for cost and economics reasons it's much more feasible to build or train other parts and have them work together, because that's much cheaper in terms of compute. As an example, with a big enough model, enough work, and the right mix of data you could probably have an LMM interpret speech just as well as Whisper can. But how much work does it take to make that happen without losing other capabilities? How efficient is the resulting huge model? Is the end result better than having the text/intelligence segment separate from the speech and hearing segment? The answer could be yes, depending, but it could also be no. Basically his beliefs are that it's complicated and it's not really a "Can X architecture do this" question but a "How cheap is this architecture to accomplish this task" question.
The contrast between the media's depiction of today's events with this outcome demonstrates how much Sam and his aligned interests can influence a narrative. The vast majority of reporting indicated that Sam held all the leverage, the board members were about to be fired, and a new board would be appointed which would be much better aligned with Sam. These accounts took special care to portray the current board as in…
The fact that this outcome occurred proves the board's resolve & determination (stubbornness?) but not much else. some important questions: - if Ilya had 4:2, why not just sit Sam down and work all this out in private? - why has the board been completely unable to explain themselves to OAI employees? to the public? - why not take a more neutral "parting ways" tone? - latest reporting suggests that the board is doing…
Human character doesn't change that much, let's say that in his heart Sam is more of of a move fast and commercialize type of leader and Ilya doesn't want that. Do you think them sitting down and talking about it is really going to change things that much? People are who they are. If anything they've probably had these exact conversations many times already. Not every relationship is fixable, sometimes people are just incompatible.
As for the board's silence to the public, this should be obvious. Talking about their thinking/plans/reasons for firing Sam exposes them to all kinds of risk both legally and otherwise. The safe move is to stay quiet in public and continue talks with the relevant stakeholders (Microsoft, Sam + loyalists) in private
This is exactly why you would want people on the board who understand the technology. Unless they have some other technology that we don't know about, that maybe brought all this on, a GPT is not a clear path to AGI. That is a technical thing that to understand seems to be beyond most people without real experience in the field. It is certainly beyond the understanding of some dude that lucked into a great training s…
>Unless they have some other technology that we don't know about, that maybe brought all this on, a GPT is not a clear path to AGI. So Ilya Sutskever, one of the most distinguished ML researchers of his generation does not understand the technology ? The same guy who's been on record saying LLMs are enough for AGI ?
Sorry, I am not including Ilya when I say not understand the technology.
In fact, he is exactly the type to be on the board.
He is not the one saying 'slow down we might accidentally invent an AGI that takes over the world'. As you say, he says, LLMS are not a path to a world dominating AGI.