The OpenAI Debacle  ## Metadata - Author: [[Changelog Master Feed]] - Full Title: The OpenAI Debacle - Category: #podcasts - URL: https://share.snipd.com/episode/7f4d6fe1-f6ad-40f5-9674-b39f25df9a65 ## Document Notes ## Highlights - **OpenAI's Early Robotics Research** - Early OpenAI work explored reinforcement learning with robotics, like robotic arms reacting to perturbations. - These experiments illustrated foundational AI research with real-world physical interactions. Transcript: Speaker 1 I don't know if you remember at that time they were doing these robotic arms and they would have stuff giraffes and they would hit the robotic arms with the stuff giraffes, which is kind Of comical, but it was like a perturbation. They were trying to create kind of robust reinforcement learning to control robotics. And that was kind of, I need to look back at the episode. I think that was basically what we had talked about back then. ([Time 0:04:57](https://share.snipd.com/snip/e5f33c72-d9c8-4a9b-837c-24376cf159ff)) - **OpenAI's Mission and Founding** - OpenAI began as a nonprofit aiming to safely build and deploy beneficial AGI for humanity. - Its founding group included visionaries like Sam Altman and Elon Musk, emphasizing safety and impact. Transcript: Speaker 1 And there were various people involved in that, various funding groups. We mentioned Wojciec, but of course, Sam Altman, who has been in the news a lot, Elon Musk, Greg Brockman and others that were part of that original group when the organization was founded In December of 2015. Speaker 2 I want to explicitly point out it was set up as a nonprofit at that point in time. That's correct. Which somewhat changed as we'll talk about. Speaker 1 Yeah, which maybe was part of the tension that led to last week. Indeed. Yeah. So it kind of had this initial board. Sam Altman was part of that. Of course, he was in the news a lot this last week. Prior to OpenAI, Sam was the president of Y Combinator from 2014 until 2019. And I know that there was some things in the news kind of trying to imply certain things about why he was let go or fired or left Y Combinator in 2019 and trying to tie those forward. I don't know that that was totally coherent for me in terms of what it was. But anyway, that's kind of his past in this startup venture backed world. I think that would be the key thing to highlight there is he's coming from this sort of venture backed startup world IPO, raise a bunch of rounds of funding and kind of big tech mindset, I guess. Yeah. Speaker 2 I agree with you. I think that was the very beginning in my sense of it at least of kind of that tension is that you it's a very different culture from a nonprofit effort in terms of the kind of the way you run Your business and such as that. Speaker 1 So, you know, one kind of key other commercial player in this would be Microsoft who has invested at this point over $10 billion in OpenAI global LLC. So that will be an important maybe distinction here in a minute. But Microsoft is a big player on ([Time 0:06:56](https://share.snipd.com/snip/cc2d7240-f415-47e0-b4c6-8e87b8345684)) - **Complex OpenAI Corporate Structure** - OpenAI's structure includes a nonprofit parent with a for-profit LLC subsidiary to balance mission and investment needs. - This convoluted setup created tensions as it merged nonprofit ideals with startup funding dynamics. Transcript: Speaker 2 We're not quite there, but we were, you know, when 2019 we were operating when all this stuff was happening and I remember it and they've sort of created they have the parent company, Which is nonprofit and the short of it is they have this LLC, which is a for-profit entity, which is a subsidiary of the nonprofit, but all of the because of the way they're operating now, All of the funding, the investment, everything has gone into the LLC. And yet you have this group of disconnected investors from the daily operations operating at the nonprofit level above. So it's two entities that were never really intended to operate together in a direct legal manner, at least from an intent standpoint. Speaker 1 And somehow the attorneys have made this all work so that it is legal. Yeah. When I was looking at like key takeaways in terms of what we're doing here, which is sort of a ([Time 0:09:11](https://share.snipd.com/snip/e64a6fd9-e393-4ef9-9b68-ec4a89378a8e)) - **Avoid Complex Corporate Structures** - Avoid creating convoluted corporate structures; they often cause confusion and operational difficulties. - Simplify organizational setups to maintain clarity and alignment between mission and operations. Transcript: Speaker 1 Have made this all work so that it is legal. Yeah. When I was looking at like key takeaways in terms of what we're doing here, which is sort of a retrospective on this crazy week of events, one of the things which is not really related to OpenAI, but if you're in business, it was a key takeaway is like, don't create convoluted corporate structures. No. It's not not going to help anybody. So there you go. That's a non-AI tip. If you're in the process of structuring some complicated ([Time 0:09:58](https://share.snipd.com/snip/d0bcba18-3455-4fb0-8658-d53f46286c04)) - **Mission Versus Startup Culture Clash** - OpenAI faced a push between pursuing its humanitarian mission and adopting the fast-paced startup culture to secure funding and scale. - This tension contributed to internal conflicts, impacting leadership and direction. Transcript: Speaker 2 And so they did that. They got the new thing. They got the initial infusion of a billion. And then the year after that, they got 10 billion more from Microsoft. But by doing that, you had the driving forward on how fast can we get there and be the leader and do this and that entire startup kind of culture contending with the our mission says we're Going to do this for humanity. And we're going to do this safely and you can see the they're crashing together and have been the last week or so. Yeah. ([Time 0:12:03](https://share.snipd.com/snip/b6e62ab2-072f-4457-be16-0c5f58e8ebb2)) - **OpenAI's Shift in Model Release Strategy** - GPT-3 marked a shift from open model releases to gated API access, reflecting safety concerns and commercial strategy. - This change signaled OpenAI's balancing act between open research and controlled product rollout. Transcript: Speaker 1 But that was in February of 2019. Still something that I think was basically just being looked at by people like us. Yeah, we had episode 32 of our own podcast was covering that. Oh, there you go. Episode 32, GPT-2, episode 32. It's interesting now, Chris, when I'm helping people learn how to fine-tune a language model or something like that, I'll often use GPT-2. And it's interesting to see because now this would be considered a very small model. And not only that, but it's open, right? So OpenAI was by its very nature and stated aims going to be very open with its research and models and IP and all of that. So GPT-2 is on hugging face. And it's a great model to use even now to kind of figure out how to fine-tune language models. But it's just interesting to compare that to, for example, GPT-3.5 or GPT-4. I'll never see those models in terms of downloading the weights and all that. It became a different org. Yes. And with their own reasons for doing that, and that's their decision. But yeah, in 2020, then, so a year later, OpenAI announced GPT-3, which was a language model that was more extensively trained on this sort of huge corpus of content from the public internet. And so it was a lot of scraped data and all of that. And I think this is really where you started to see some pretty crazy outputs from these language models, what people might have thought not possible. They started to see this with GPT-3. This is also where you kind of see this shift within OpenAI from releasing GPT-2 as a model to releasing GPT-3 as an API with a more gated release, a slower release. I was even reading in some of kind of the lead-up and debriefing from the chaos of last week that you could interpret some of this slower release of GPT-3 as an API and a product to really This tension that you're seeing between a startup kind of mentality, wanting to fail fast, release things fast, learn from things in public. And this kind of more shielded, nonprofit, good of humanity side that is wanting to do things maybe in a more slow way that ensures safety and releases things not with harmful outputs, That sort of thing. So you see this kind of starting to really collide and come together, I think, around GPT-3. GPT-2, GPT-3 were language models, caused a big stir in the AI world. ([Time 0:19:18](https://share.snipd.com/snip/566e7d6c-e0c4-493a-9db1-a64cbcdc5152)) - **Sam Altman's Sudden Firing** - Sam Altman was abruptly fired following a keynote, shocking the AI community and Microsoft. - The board had pre-planned leadership changes without broad stakeholder communication, causing upheaval. Transcript: Speaker 2 And the first thing I saw was OpenAI had fired Sam. Speaker 1 And was like, which was crazy because it was like right after his keynote. I know a day. A day later. Speaker 2 It was a day after his keynote. Yeah. Speaker 1 Yeah. It was just mind boggling. Yeah. So apparently he got a text on Thursday night from one of the co-founders asking him to join a Google Meet on Friday. And they had already kind of tapped the CTO Mira Marotti as the next CEO. And on Friday, literally Microsoft learns of Altman's firing, quote, a minute before the world does again. So it was on Friday that we heard it actually not Thursday, but yes. Yeah. Yeah. Exactly. So it happened while it got into motion on Thursday and then some of us heard about it on Friday. And apparently also Microsoft, which was just, you know, so I knew about the kind of convoluted corporate structure and all of that. But it was just mind boggling to me that Microsoft was not informed about this given their investment in the for-profit entity, which I think they have 49% stake in, if I recall correctly. Correct. Yeah. It's something like that. I don't remember the exact figure, but certainly they've invested over $10 billion with plans to invest $10 plus more billion. And of course they've integrated OpenAI, chat GPT, et cetera into Azure, into Bing, into Microsoft Office 365. And so yeah, it just sort of boggles the mind that this happened in the way that it did. So then on the 17th, then OpenAI releases a statement that Sam Altman was ousted. ([Time 0:26:12](https://share.snipd.com/snip/0fbb5de9-3467-4ea0-92c5-6fe4488dc193)) - **Employee Uprising at OpenAI** - OpenAI employees expressed strong dissent internally, even using explicit language on Slack to protest leadership decisions. - Reports indicated up to 95% of staff threatened to leave without Sam Altman's return. Transcript: Speaker 2 But the OpenAI employees were reported to on their internal Slack. We're using, I'll gently say, FU and showing the middle finger on their Slack. Apparently the employees had had enough. By the time we got to mid weekend and going from Saturday into Sunday, the employees started finding their voices. We'll hear about next. Speaker 1 And there were reports basically that up to 95% of the employees within OpenAI, we're going to depart OpenAI if some deal wasn't struck to have Sam Altman return as CEO, which obviously Would be the end of OpenAI, at least as we know it. ([Time 0:30:50](https://share.snipd.com/snip/e6249538-941f-4edb-9f69-249cc97f0a00)) - **Rethinking AI Dependency Risks** - The OpenAI crisis revealed overreliance risks on a single AI provider in industry strategies. - Organizations are now rethinking diversity and backup options for foundational AI models. Transcript: Speaker 2 If only there were services that could help people find a variety of different models and not be entirely dependent on a single family prediction card. Speaker 1 I didn't say that. Yeah, yeah, sorry. This is a point we'll make later, but I think it does. We even a few weeks ago had some comments about with the executive order, is there a framing up of the market around these kind of foundation models because of the regulatory burdens That would start starting to be put on them, right? And I think that this kind of opens up the field a little bit more, regardless of what has played out with OpenAI. I think people are really wrestling with the fact of, hey, what else is out there? And of course you've got amazing players in that space from Mistral to Mosaic ML to Meta and what they're doing with Lama2. And people are really finding, of course, we're helping, prediction guard helping people build with these models. But a lot of people are finding a lot of success with these models. And I got a number of messages during this chaos about, hey, we weren't thinking about trying open models or models outside of the GPT family prior to all of this craziness going on. But now we're wrestling with that, not necessarily over the fence yet in terms of a strategy around that. But it's definitely caused people to think a little bit more about this kind of market. Speaker 2 Yeah, having backup through diversity, if you will, in terms of model selection, is now going to be in the corporate consideration for any significant organization going forward. We've ([Time 0:32:26](https://share.snipd.com/snip/c62f8b72-75f9-4115-9513-8ffa38cf6234)) - **Board Overhaul and Leadership Restoration** - OpenAI's resolution involved replacing its entire board and reinstating Sam Altman, signaling a shift towards traditional Silicon Valley leadership. - This raised questions about continued adherence to its original nonprofit mission. Transcript: Speaker 1 I think that's a real question to ask. Speaker 2 Yeah, if I was channeling the OpenAI marketing team, I would expect that they would say, absolutely, that still exists. That's a overriding mission that I would also suggest that the marketplace is probably not convinced of that at this point in time. Yeah, for sure. Before we go on, I just want to point out one thing. Another big, big retro consideration that will not go quickly is the ability for the voice of the employees in a unified stand in such a corporate. This was the biggest AI company in the world in the sense of mind share and what they're doing models. And the employees made this happen. When they said you're not going to have an organization, if you continue down this path, that made a big difference. And I think that's another thing that we will see play out in organizations going forward. ([Time 0:35:04](https://share.snipd.com/snip/8de397b3-8f67-46dd-bc3e-f705da9b3a61)) - **Use Diverse AI Solutions** - Explore alternative enterprise AI models for increased control and reduced single point failure risk. - Balance the ease of managed AI services with the resilience of self-hosted or diverse solutions. Transcript: Speaker 1 There's definitely from my perspective, some takeaways, of course, I'm actively leading a company that is providing hopefully trustworthy LOM APIs in enterprise environments That, you know, can be self hosted. So my big takeaway, and I think what I was busy doing all week was answering people's questions about, Hey, if I don't have open AI, what is there? And it turns out that there's a lot. So, you know, it doesn't necessarily have to be our APIs, but there's a lot of options around using enterprise models in a safe, secure environment that can be deployed under your control. And yes, there's kind of this balance to take around like, Hey, open AI's APIs are really good because they're a managed service. ([Time 0:40:26](https://share.snipd.com/snip/a5f24fb8-8e7d-402b-807f-b1fdd046f6ed)) - **AI Risk Management Industry Emerges** - AI risk management emerged as a formal industry field due to OpenAI’s recent turmoil. - Corporations now recognize AI governance as a pivotal aspect of strategic risk assessment. Transcript: Speaker 2 Provide enterprise solutions. Yeah, AI risk management as an industry field has been born. That's what's happened here. Yeah. AI isn't just the data science team that is going and either using APIs or in some cases using their creating their models, you now have risk management as a formal corporate concept That everyone will be adopting and will go through all the the ranks of every organization. So an entire new industry has been born out of this. Speaker 1 Yes. Another point that I saw being made is, this is kind of a wake up call to regulators as well. You have companies testifying before Congress to tell them how maybe they should be regulated. But this is kind of a wake up call that hey, you know, maybe these companies aren't so good at regulating themselves that often. So ([Time 0:42:02](https://share.snipd.com/snip/f09aafbc-f4be-4d55-a68c-21a0fd208d9d)) - **AGI Moves Into Public Consciousness** - Artificial General Intelligence (AGI) shifted from a distant concept to a serious, real concern within just a few years. - Public awareness and discourse now treat AGI as a legitimate technological and societal issue. Transcript: Speaker 2 If we go back just a few years on this podcast and not very far, if you really think about it, when the subject of artificial general intelligence, AGI would come up, we didn't take it too Seriously. You know, it was one of those things that wasn't wasn't really practical AI for us at the time. It was one of those someday, maybe, you know, kind of get their things. And I think that what we've seen in a couple of different waves, we saw it with chat GPT coming into being in the capabilities and the fact that it hit the public's consciousness so intensely. And then, you know, the concern, no matter what Q star ends up being, regardless of what the outcome is, Thanksgiving dinners were actually with people with genuine concern. Whether it's general or not, it's general in a certain way. There you go. And the power of it and what that would mean for their own lives, even if they had nothing to do with the AI industry. And so what we've seen change in the large is that the notion of artificial general intelligence is entirely legitimate to ponder, to be concerned about, to either fear or be excited About, or maybe a bit of all of the above. But ([Time 0:43:40](https://share.snipd.com/snip/289f6349-c239-4b9b-8246-c97676a43fcf))