Mira has done some great interviews, some with Sam. Also, interesting interviews with Ilya. I like her a lot, and Ilya too, based on what I’ve seen them say in public. I’ve consumed dozens of hours of content with them both, and firing Sam seems to align with their priorities if they are truly worried about safety issues. Surprising though, my impression was that Sam was the one pushing the safety agenda forward. Sam’s talk at APEC was very interesting as well…
My hot take
Sam wanted to push things out QUICK before the EO and Bletchley blah blah took real effect and OpenAI was shackled by overregulation. “Too late now, weve just got to work with what weve got!” Kind of situation.
There are 125 replies with an estimated read time of 16 minutes.
OpenAI announced a significant leadership transition - Sam Altman, the company’s CEO, is set to leave the firm, and the board of directors has appointed Mira Murati, OpenAI’s Chief Technology Officer, as the interim CEO. This change in management resulted from a review process by the board that concluded Altman was not consistently candid in his communications, thereby affecting the board’s functionality.ref
This abrupt departure has sparked various reactions within the forum, ranging from surprise to concerns over the organization’s future. Some forum members have speculated about undisclosed board dynamics and motivations, while others have expressed apprehension about the potential impact on their businesses and the AI industry in general.refrefref Some have speculated that this shakeup could signal OpenAI’s strategy to go public, but there has been no official confirmation to support this.ref
Forum participants called for greater clarity on the reasons behind Altman’s departure, urging the board to provide a transparent explanation to all its stakeholders, especially the supporters and paid users.refref Altman’s departure also saw several forum members express support for his vision and leadership style, worrying that his exit could lead to a divergence from OpenAI’s mission to ensure that artificial general intelligence benefits all of humanity.refref
In an additional twist, Greg Brockman, the company’s Chairman, has decided to step down from his role but will be remaining in the company, now reporting to the CEO.ref It was suggested that depending on the severity of the issues Altman was allegedly deceptive about, the board may have been legally obligated to act swiftly.ref
Despite the leadership shakeup, OpenAI has reiterated its commitment to its mission and has stated its confidence in Murati’s ability to effectively lead the organization in the interim while a permanent CEO is sourced. The forum will continue to keep a vigilant eye on the developments and their potential impact on OpenAI and AI at large.ref
Summarized with AI on Nov 18 2023
AI used: gpt-4-32k
I really felt that it was Satya Nadella saying, “how do I think its coming along? You have the most powerful piece of technology mankind has ever developed. Im coming along for that ride.”
Does anyone know what Sam Altman’s view on monetization was? Did he have less of an appetite for maximizing monetization than the board wanted? If the board and he didn’t align on this, it could have been a big issue.
My opinion is it has something to do with the regulatory process of AI and Sams participation in the process. I speculate that Sam said or agreed to something beyond the bounds of the board’s agenda. Role in OpenAI and Public Perception : Sam Altman was not just the CEO but also the face of OpenAI. He was widely recognized for his contributions to advancing generative AI technology. Over the past several months, Altman has actively engaged with regulators and global leaders to shape the response to AI, participating in U.S. congressional hearings and meeting with international figures like President Joe Biden and German Chancellor Olaf Scholz
It’s my belief that the direction of regulation Sam advised to public officials is in contradiction to the board’s perception of AIs future.
I suspect that Sam will bounce back heavily with a team that follows him from Open AI. I look forward to the developments of both structures and hope the new structure will breed a competitive nature that is more dispensable to the public’s needs.
AI will always be available to all; is what I get from the “ousting” of Sam.
Greg following, is a sign of things not understood by me yet and I’m curious to why it’s not discussed more. I find it fascinating that he would just quite his invention. Obviously, there is much more to the scope of these developments than we are aware of.
I’m wondering how partners are taking this. Obviously, Microsoft has deep concerns but they probably have sway and options to reset the table.
What about companies like CustomGPT - how does @alden feel about this obvious disruption in momentum as they grow their platforms on OpenAI’s foundation?
Altman was instrumental in commercializing OpenAI. If the company is leaning in a different direction - more about safety, attenuating fear, and reshaping the organization to slow commercial progress - does that energize partners to fill the void created by this new direction?
Is it possible that this seemingly undesirable turn of events may benefit partners?
“The story” is a lot of recursive groupthink without basis. The only thing to go off of is the OpenAI blog post written from their own position with ambiguous wording.
If anything, it could be the whole course the company has been following as motivation, degrading models and being dishonest about it, up to devday being one man’s “one more thing” Steve Jobs emulation. A disconnect between the Musky YCombinator Stripe undergraduates looking to hang as many billables onto their VC buddy’s services (even where I type), vs the machine learning pioneers. The world tour of chats with leaders. Or one particular transgression that will not become public.
All is just midnight speculation until we see signs of a course change.
It’s a complete misstep they should be making general AI technology like they said they would, I was hearing behind the scenes claims of “just niche down and you’ll be fine.” But I think releasing GPT’s is basically a slap in the face to that.
If someone effectively manages to brand an API wrapper than good for them. I just don’t think its their job to create niche tech but rather create the general purpose tech. That’s what I always thought it would be. Releasing GPT’s is basically a step in the direction of like “No actually we’re coming for your necks”. It just makes it really hard for anyone to build a business with their tech if they can’t even keep up with the speed of which their own tech is moving. This happened to plug-in developers, all their efforts are just for nothing now?
“API For Commercial Use” Yeah okay. How are we supposed to use it for commercial use? How am I supposed to build products if next year they’re going to make it even easier to build an even more complicated product that isn’t just a chatbot? They can’t just come for small businesses necks like that, IMO.
The writing on the wall to me is…“Yeah you THINK you have your MOAT, but we’re coming for you”.
I just am not feeling strongly in investing in the tech, because without knowing their plans and what the future is - anyone could just be wasting a ton of time building a product that’s going to be wiped out. It’s just too chaotic and unstable to me.
You might think right now “Oh okay whatever I’m building has a really strong Moat its complicated it has RAG and multiple GPT instances and using Google AI tools and yada yada.” But how do you know in a years time all your effort will be for naught? because “GPT builder 2.0” comes out and it just doesn’t matter anymore theyve captured the entire market.
Gotta think of it as day-trading, in and out. Exponential growth means something CRAZY for development, and therefore expectations of creating a lasting company based on current tools makes no sense. It’ll be a lot of small movers on the periphery but they are the foam on the wave. The foam shouldnt think itll last past the next wave breaker.
There ARE niche enough ideas (I think) that it makes sense to build now with the current tech and can always be updated. But it seems like you’re saying you have to basically build and sell as quick as possible to a market that doesn’t quite understand what it is yet
Dude, agreed! Okay. If Sam did something horrendous/out of line why would other people be leaving as well? Why was nobody informed? Complete subterfuge. They even told both Sam & Greg at a random, separate time. It’s like breaking up with your girlfriend over text man, you just don’t do em dirty like that
I think the only truth here is that there were diverging paths of interest.
Company direction (the GPT Store, for example) was likely to nuke the value prop of various startups building atop the same stack. And apparently it was at odds with the core mission the board and remaining management is committed to.
As unnerving as this is, for those building applications here a strategic return to foundation development may turn out to be very good news, leaving us to build what might otherwise have been built by some (imagined) OAI Consumer Products Group or the OAI Enterprise Division.
The unnerving part is the impression of instability, the shock and awe of enormous management changes with sparse explanation.
Would say, keep building, architect to keep your options open, and ship, while hoping that the OAI keel evens out quickly and our collective and individual investments here pay off.
Disagree. If OpenAI doesn’t build a GPT store someone else will, and it will be filled with ads and possibly sell information. Just as it happened with the plugin store.
Yeah, I think that’ll be the case.
Right now only big tech has the resources to develop and deploy these LLMs, the LLM capabilities will be locked behind those said big tech. They are the wave. Most people will be on those platforms because of those vast capabilities.
The wave has the ability to pick up all of the projects in the ocean and use them to improve its own significant capabilities, and eveutually all those other projects will merge together under those LLMs.
So theres big opportunity in between LLM production and product release cycles to make things that arent included in those big models, but at an ever-increasing pace, all of those opportunities will be gobbled up and become one with the wave. I think small business can get good at surfing, but the exponential nature of development means that movement will have to become more and more quick.
