EVERY SPOKEN WORD
50 min read · 10,013 words- 0:00 – 9:56
Recapping the OpenAI saga
- SGSarah Guo
(music plays) Hi, No Priors listeners. Time for a host-only episode. This week, Elad and I are gonna talk about what's going on at OpenAI, of course, video, Q*, uh, what might be next in research, and some predictions. Okay, Elad. We have to start with the saga from this past week. What is your take on the outcome and the second-order effects?
- EGElad Gil
From a second-order effect perspective, um, this seems like overall positive news for everybody involved, so... It looks like on the OpenAI side, they're back to being in a really positive, stable situation. I think they still have, like, the leading model in GPT-4. Um, they've reworked the board, which seems like a positive thing, so, you know, imagine if this had happened two years from now or three years from now, et cetera. So it, it seems like it would net increase the stability of the company and governance and a few things like that, or the nonprofit and the company and governance. So as an external viewer, it seemed like a, uh, painful thing to go through. But the flip side of it is, it seems like they're moving forward and moving ahead in a positive way. Um, and then in parallel, I think it may have ramifications to other areas, um, that we can talk about if useful, but, like, what are the second-order aspects of this? But it'd be great to hear what you think.
- SGSarah Guo
Yeah, I think the first, uh, obvious lesson is that governance matters, right? And this isn't an area where I think most companies are that experimental, but I think a lot of entrepreneurs are likely to think twice about placing their destiny in the hands of groups with explicitly mixed incentives now. I'd say generally, nonprofit governance is... Not every organization, but as a class, known to be kind of abysmal, right? Because performance is hard to measure objectively, and so it often ends up more about politics and specific relationships and status gains than outcomes. The clarity around how much, you know, any board matters was kind of a, a wake-up call for people. The second lesson that a lot of people will take away, or I think should, from this whole saga is that money matters, right? The factors of production are labor and capital, and compute is the AI-specific form of capital. Microsoft holds the compute here, and that clearly mattered. This is amazingly well-managed, uh, and supported by Satya and Kevin. And then the class of, like, really special labor here, the team rebelled, and the board obviously underestimated the level of support that Sam and Greg had from them. And then, then I think one thing that is often unsaid, because it's a little bit less idealistic, is that a lot of OpenAI people were very upset this last week about not just the destruction of the mission, which I think was absolutely, like, genuine, but also destruction of the value they'd built and been promised a piece of with the $86 billion, uh, tender offer. It's just a reminder that labor and capital are, like, leverage. They're stakeholders, and there's no, there's no free lunch or control without skin in the game, and I, I think there, um, likely shouldn't be from the view of many of the people involved in this.
- EGElad Gil
Yeah, I think you're raising an important meta point, which is basically, what are the incentives that different organizations have in place? And ignoring OpenAI, I think there's a lot of boards which added board members for reasons that were politically motivated or motivated by different regulations getting passed that forced certain board changes, et cetera, and you also see that in executive teams. And I think it's really important for people to go back and rethink, "Okay, who should be on my board, and what- why? What are they representing relative to the board? What expertise do they provide? Or what insights are they bringing, or what strategic views are they bringing?" Same with your executive team. And also, what are their incentives? And, you know, there's this, um, th- this view that's kind of been moving around Silicon Valley in terms of the professional managerial class, right? People who have alliance not to the organizations they work at, but external incentives. And those external incentives could be speaking at TED, or going to Davos, or getting kudos or an award from a specific organization versus doing what's actually their duty as a, as a representative of the various shareholders of a company in the context of the corporation. And so there are these fiduciary duties that may be being breached by, um, other incentives for different actors who've been added over the last, you know, five, 10 years to boards, to executive teams, et cetera. And I think it's really worth rethinking, like, "Who do I want on board and why?" And it also comes back to some of the companies that have been resetting things relative to politics. I think Shopify did a great job, for example, of saying, "We're a performance-based culture. We're focused on, um, you know, a very specific mission. We don't want that mission to creep. We're about... We're not a family," you know? (laughs) Like, if your uncle shows up drunk and does something bad, you forgive him. If a board member shows up and does that, then, you know, you don't want them on your board, right? They're being i- irresponsible. There's also that broader context of like, how do you want to think about alignment, incentives, culture, motivations, and, you know, is this a good moment in time to sort of pause and rethink some of those things relative to, to your own company?
- SGSarah Guo
Yeah. One friend at OpenAI who, uh, I guess publicly declared that this reignited their belief in clear incentives and some, um, and good intent in capitalist structures, that has actually seemed somewhat radical in, uh, in many Silicon Valley companies over the last few years, but, uh, I think that is going to get rethought when you see what happens when there are unclear or misaligned incentives.
- EGElad Gil
Yeah, there, there's two actually related quotes to that. There's one which is something which I'm gonna get wrong, which is something along the lines of, like, capitalism is the best way to take care of people that you don't know. You know, it's the means of actually growing the pie in many cases and providing for, for others through the sort of incentive of markets. But the other one is a Charlie Mungerism, and unfortunately, Charlie Munger passed away earlier today, and obviously he was sort of a giant of industry, and he had this great saying, which was, "Anytime I think I understand the importance of incentives, I realize that I'm underestimating the importance of incentives."
- SGSarah Guo
If we just-Think about what the other second-order, like, more commercial effects are. I do think that, uh, there is an interest in owning models more, in open source models, and in, um, at least understanding, like, reliance on a single vendor. What do you think here?
- EGElad Gil
Yeah, I think, um, there's a couple people who have built solutions during the last week that, for example, Braintrust now has a AI proxy where you can use the OpenAI SDK to effectively query multiple different, uh, models, including Mistral, um, and LLaMA through Perplexity, as well as a variety of other things, GPT-4, GPT-3.5, um, I think potentially Anthropic. And so it just allows you to be able to both load balance your, uh, queries or prompts, but also interchange models more easily so that you can actually look at performance across them. I think Qima similarly has done something over the last week that they've released that helps with some of the proxying and other things. And so I think there's solutions like that that have started to be accelerated in a market that would have happened inevitably, right? I think everybody... The journey that I see people often take is they'll prototype on GPT-4. They'll look at how good it is, and then they'll either keep it on GPT-4, particularly if they need, like, advanced channel logic or other things, or if they need very high throughput and performance and low cost, then sometimes they'll either switch to GPT-3.5 or they'll see if they can fine-tune something, right? And that's only people with enormous scale. Like, I don't see very many fine-tunes happening (laughs) in general, unless, you know, somebody has enormous scale and/or proprietary data that they just don't want to get out, right? So they'll fine-tune Mistral or LLaMA or something. So already, I think people were thinking about that, and then that means you need to build an orchestration layer. You may need the proxy, you may need a variety of things, so.
- SGSarah Guo
The dimensions that people were evaluating an LLM provider on or whether or not they wanted to, um, control or, uh, host or fine-tune themselves just became more clear, right? Um, where, like, reliability, um, became more important, but the reliability, latency, um, cost control, capability questions were sort of naturally there. And to be clear, like, OpenAI leads on capability in many areas, in unique capability in, in some, right? Like code generation, GPT-4V, right? You can do amazing things with that and people, uh, should go build on those tools. I think the ecosystem will mature and OpenAI is a great partner, but, uh, I think the questions are just much more obvious for anybody relying on these models now.
- EGElad Gil
Yeah, and I think honestly, a lot of the bigger enterprises I knew, uh, always wanted to make sure that if they really needed to that they could second-source something, so I don't think this is a new thing. In other words, uh, you know, one could argue that no matter what OpenAI does, there will always be at least one or two other suppliers or vendors or partners for advanced LLMs simply because the market always wants an alternative, even just for negotiation leverage. And so if you look at other markets, for example, in the router world, one of the main reasons Juniper exists is because everybody wants to make sure that they can push on Cisco for pricing, and so they always want to have a second source. That's why Juniper is always 10 to 20% the size of Cisco, right? It's just second sourcing. Or AMD versus Intel for a very long period of time. So I think often markets will end up with other players just because big enterprises always want to have that option if they need it, even if it isn't as good. And if anything, I think OpenAI kind of emerges more stable through this in ways that people didn't expect simply because there's gonna be more stability at the board level in a way that people didn't understand perhaps that there could have been instability, right? This is a strengthening event and a focusing event for the company, at least, you know, from what I can tell
- NANarrator
(instrumental music)
- EGElad Gil
Do you want to talk about Pika and video?
- 9:56 – 16:14
AI video products
- EGElad Gil
- SGSarah Guo
Yeah.
- EGElad Gil
Perhaps?
- SGSarah Guo
Yeah. There are a couple really amazing launches happening in the video space. What's the cause for this? Like, we, we suddenly have text-to-video generation and avatar cloning in different ways. What do you think is gonna happen in this space?
- EGElad Gil
Interesting shift has been happening because basically if you go back a year and a half, Midjourney launched, Stable Diffusion came out, DALLE-2 came out, and there's a whole wave of people saying that they were gonna go build on diffusion models, and ImageNet was, like, the thing that everybody was gonna go do for, like, two months. And then ChatGPT came out, and then everybody was like, "Oh my God, I need to go work on LLMs and language and natural language and NLP and all this stuff." And so the entrepreneurial (laughs) ecosystem went through this sort of zigzag where everybody was going to do image gen, and a bunch of companies started going down that direction, and then the, the LLM stuff really kind of was substantiated through ChatGPT, and then most people went that way. And a handful of founders stuck around on the diffusion model side, and diffusion models, uh, you know, are really popping up. And obviously there's, like, image transformer and a bunch of other stuff, but they're mainly being used for image gen, for video, and for audio actually. And so there's a wave of people who have continued to work and crank on this, um, and they're starting to come out with really interesting products. For example, Pika is a great example where it was two, um, Stanford PhD students who've been working on diffusion models for some time, and they made this, you know, really amazing creative, um, text-to-video engine. There are other companies, like, HeyGen or Synesthesia or others that are doing, you know, "Let me use these diffusion models to, to clone an avatar, to generate an avatar of a person so that they can either go into the metaverse a la Zuck, or alternatively, they can use it for marketing purposes, they can use it for internal training, they can use it for all sorts of applications." And then there's some really cool, like, audio-based things coming out too which I think are starting off more sort of tools to create music or to simulate voice in the context of, um, a, a soundtrack or you, like, make EDM and you wanna add voice to it, right? And you can just now kind of do some really interesting things there. So it seems like there's this really interesting renaissance, um, that's happening in part due to diffusion model work and in part due to a, due to a handful of founders not getting distracted by LLMs, which are super exciting (laughs) obviously, um, but wanting to do things in video. It's a really exciting trend, and I'm, I'm guessing the success of some of these companies...And their, their traction and growth is gonna pull more people over to, to work in this area again. I think it was just an area of less emphasis for the last year, relative to language.
- SGSarah Guo
One of my favorite dynamics that happens in sort of technology ecosystems is that once people show that something is possible, like, a lot of talent floods in, right? And you kind of get... Or y- you get a lot of competition, but you also get, um, innovation coming in waves. That could be with Mistral developing open-source models that are, uh, actually interesting from a reasoning perspective at relatively small parameter size. Or it could be Demi and Chenling and the Pika team creating text-to-video generation models that are really interesting quite efficiently from a training perspective. And, uh, I know we're both investors here, but I've seen a huge wave of people interested, as you said, in media diffusion of different kinds now that they know it's possible.
- EGElad Gil
And it has real benefits, right? Because it's, um, it's very cheap to do.
- SGSarah Guo
Mm-hmm.
- EGElad Gil
And from a dataset perspective, the data is reasonably straightforward to get. I mean, it's hard to get, but it's not as hard as, you know, the entire internet and transcribing voice from videos and all the rest of it. The original, um, Stable Diffusion model supposedly was trained on, like, 600K of GPU. And these models cost in the millions to train, not tens of millions, at least initially, right? And so that's another big difference relative to the really big foundation models and language models and all the rest, right? And so you can actually imagine that in the language world, you're gonna have a lot more platforms that people build on. In the diffusion model world, uh, image, video, et cetera, you're gonna have more people kind of grow their own, right? And people should still potentially try things on Stable Diffusion first just to test it out. It's back to the, you know, no GP before product market fit, but they can still train their own model in a very economic way relative to, like, the amount of money a startup would raise. So I think it is, it is a more accessible thing in some sense, unless you just go and build on somebody else's LLM, which people should do for most things initially as well.
- SGSarah Guo
Yeah, one of the things I- I think is really interesting about this space as well is we've actually had leading researchers, like, say, you know, "We're still very, very early in video. Video generation is so hard." It's data-intensive. The data's like... As you said, it's not the whole internet, but it's problematic in that a lot of the training has happened on short clips. People aren't sure how to caption. It's expensive to generate. You have, like, complex, like, sliding window approaches and others to, um, try to deal with the, like, temporal coherence problem. There are many more unknowns about how to, uh, progress this type of product technically.
- EGElad Gil
Do you mean video specifically, or-
- SGSarah Guo
For video specifically.
- EGElad Gil
Yeah, 'cause the teams for all these things are actually quite small, right? The Pika team is reasonably small. The Midjourney team for a long time was tiny.
- SGSarah Guo
Mm-hmm.
- EGElad Gil
And so I actually think this is a good example where you can do a lot with very small teams. And to your point, there's all sorts of technical challenges. But the reality is you can get to the cutting edge with, like, a handful of people in these fields, which isn't necessarily true as much for, you know, other types of models, or certain types of models at least. So I- I do think it's striking, um, how few people you actually need to do something really interesting here. And to your point, there's other challenges coming. In three, four years, maybe it becomes harder.
- SGSarah Guo
I agree with you. You don't need more than a handful of people. Like, there's empirical evidence now. Maybe a slightly different point, which is instead of, like, having to have a certain size of team to go deliver an LLM at scale, there are more degrees of freedom in how you would innovate technically in this area. Um, and there's more disagreement on, like, how to progress, and I think that's actually just interesting for startups.
- EGElad Gil
Yeah, I think they should just use Q* .
- SGSarah Guo
(laughs) That's probably-
- EGElad Gil
So I think that's, that's the main solution to most problems, I feel, in, uh, in AI today.
- SGSarah Guo
Okay, well, do you want to give me some investment advice given Q*? Should I go home?
- EGElad Gil
Yeah, you should do mainly Q* centric companies. And so, you know, if they're doing Q*, you do it. If they're not doing Q*, you don't do it. So that's
- 16:14 – 19:47
Moving from Diffusion Models to LLMs
- EGElad Gil
one big piece of advice.
- SGSarah Guo
I think one of the things I would like to go back to and, um, talk about is your point of view on, like, "Hey, a bunch of people moved away from diffusion models to LLMs."
- EGElad Gil
Yeah.
- SGSarah Guo
One of the reasons that people moved away from diffusion models to LLMs is because there's a lot more, uh, sort of text and code in enterprises. That is obvious, right?
- EGElad Gil
Mm-hmm.
- SGSarah Guo
Working with images and video and audio felt more verticalized, like, "Where are the B2B use cases?" And I- I think what we're seeing increasingly is the creative fields are, uh, are actually pretty commercial, right? So one of the things I'm most inspired by, and I think there's a lot of money in, is if you look at Midjourney, one of the busiest, like, biggest knocks on them, uh, from investors or naysayers early on was, "Well, like, how many people want to make images?" That's not a hobby. That's not a social network. Like, what percent of the population are artists? And this was clearly wrong just in terms of the scale that Midjourney has already reached. There's probably two pieces here, right? Like, one, these tools like Pika and HeyGen and audio generation and Midjourney, they make the pie bigger for creative fields, especially since if you look at Pika or HeyGen, they're really focused on all creators, rather than just, like, the film industry professional. Um, and, like, if you go look at the, uh, Midjourney use cases and then I suspect the Pika and HeyGen use cases over times, they're very commercial, right? Like, a lot of the things that you named or that people are experimenting with are really about communication, marketing, and advertising.
- EGElad Gil
Yeah, I think if you just look at it as market cap of incumbent, right? Adobe's a 280, $300 billion company. Like, that's huge, right? (laughs) And so I don't think the creative world is small, right? I think a lot of creator economy companies have failed in the past, which is a different thing, e- e- depending on how you think of the creator economy, right, in terms of...... you know, celebrity-based marketing or whatever. But if you actually look at enterprises, obviously they use enormous amounts of imagery and video and other things to reach with and interact and brand and, you know, associate with their customers. And you look internally, you, you need to create imagery for slides or for other things and communication, and, you know. And so to some extent, you're kind of looking at different proxies and you say, "Okay, well, what are some of the proxies on the, um, on the text-based side?" And you could say, well, you add up Microsoft and, you know, a few other companies like that and you're kind of getting a rough proxy for some form of text. Not really, but you know, I'm just simplifying things dramatically. And then you add up Adobe and a few other companies and that's your proxy for, for image gen, right? And so I think, um, both are big. And then the question I think always with the diffusion model based companies was, where are the biggest application areas? And the application areas also were a little bit driven by, well, what, where will incumbents play a role? Where will you get blocked by other companies in the ecosystem versus, you know, it's a natural new greenfield thing? Most things aren't truly new capabilities. Most things are just like making certain things dramatically easier. There's the, the duality of that. There's the market expansion and more people can do this thing. And then there's a value contraction. Hey, you can do this at a tenth or a hundredth the cost. And so often in markets like this, you see both of those things happen at the same time. You're simultaneously growing the market and shrinking
- 19:47 – 26:00
The beneficial margins of AI investing
- EGElad Gil
it.
- SGSarah Guo
Yeah, I mean, you have the, um, the, um, famous, uh, strategy, like your margin is my opportunity, right?
- EGElad Gil
Uh, I was gonna say, I mean, well, these are actually higher margin things. I think really what you're doing, if you look at the sort of, um... Uh, my understanding, I, I need to look up these numbers again, but I think it's something like software spend is like, I don't know, I'm making it up, half a trillion dollars, $500 billion a year. And then services spend is like three to five trillion a year, right? And so really what you're doing is you're taking services revenue, which is very people-intensive and low margin, and you're converting it into higher margin software revenue, but less of it. So maybe you take that five trillion and you turn it into two trillion. But it's 80% margin two trillion versus 30% margin, right? The margin dollars actually expand. And you see that in other industries. That's kind of what Anduril is doing in defense, right? They're taking a cost plus model, right? You buy a drone from Lockheed Martin for a million dollars and you, you get paid by the government 5% cost plus, which means you get 5% as your margin so you make 50K off of it. And Anduril will sell the same drone or, you know, a better drone for $100,000 with 50% margin. I'm making up the margin, right? But that's 50K. And so you're making the same margin on a tenth of the price. And so I think one of the ways I think about Anduril as a company is they're taking very bad low margin revenue from other defense companies and turning it into higher margin, healthier revenue, right?
- SGSarah Guo
Yeah. Uh, I'm gonna edit the quote and just say like, "Your ASP is my opportunity," right? There's a, like a democratization that happened in the like latter half of the SaaS revolution, or really most of the SaaS revolution, which is instead of there's 100 very large enterprises that have some sort of CRM because it costs X dollars to, um, deploy and implement Siebel. Then you have, you know, tens of thousands of companies who can buy Salesforce and then companies that figure out how to efficiently distribute, um, SMB SaaS on the internet, even though that, that is still hard. I, I think one of the things that is, um, interesting here is let's just take video generation as the example, the ASP of, you know, hundreds to thousands of dollars an hour make it single digit dollars to generate per minute and expand the audience to many more people, right? So that's the democratization that is happening. Somebody told me a joke the other day, like somebody really negative on AI investing. There's only five businesses in AI that have breakout traction right now. Foundation models, waifus, mid-journey, copilot and inference platforms. I think there's both truth in it and like also it's not that funny of a joke because you're like, it's true. Like there is a set of things that people are figuring out that are really early, many of which feel pretty different than like yes, it's, it's useful to look at the incumbent vendors like Adobe, but you're not fighting really the video editing software spend. You're eating into the production spend.
- EGElad Gil
Sure. The early internet version of this, by the way, is there's only five things you do on the internet. You go to Yahoo! to look for links. You buy Pez dispensers in an auction on eBay. You buy some books on Amazon. And then there's probably like some, like two bullshitty companies that you would have quoted as like high grit things, right? So if you were looking at the internet circa '96, '97, whatever, you probably would have had a pretty short list of real use cases and then a bunch of stuff you just thought was kind of dumb, right? And you'd be like, "Look, you're not changing anything." Like you're still using Microsoft Office and you're still using whatever shrink wrapped software, you're still watching TV, right? And so I feel like we're kind of in that era of AI. The stuff that's going to happen right now is the really easy low hanging fruit, and then a bunch of dumb things are going to get built that aren't going to work. And dumb is not meant in a pejorative way. It just means like it's very hard to tell what's actually a good idea in a new market like this. And, um, that was true of the internet, and that was true of mobile, and that was true of cloud. There's a lot of these like waves where there's a bunch of stuff that gets built, right? Um, so I think it's the same thing, right? It's a very positive sign. This has been such massive traction in such a short period of time for so many companies, if you think about it. It's actually kind of amazing. So I'd actually take the other side of that, but I, I totally get the point.
- SGSarah Guo
Yeah. Well, with the AI focus fund, I agree with you. ............................
- EGElad Gil
Are you gonna change the name of your fund? You should call it like ConvictionStar or something.
- SGSarah Guo
ConvictionStar? Yeah.
- EGElad Gil
But spelled with a Q, like Conviction.
- SGSarah Guo
Yeah. Co- ConvictionStar. Okay, LPs, you heard it here first, ConvictionStar.
- EGElad Gil
Very exciting.
- SGSarah Guo
Yeah.
- EGElad Gil
I, I should send you a T-shirt.
- SGSarah Guo
Thank you. Please do. That'll actually be-
- EGElad Gil
... ............................ -the swag for No Priors season one. If you're a guest, you're gonna get the No Priors tequila and then a ConvictionStar T-shirt. (laughs) Yeah, that's very good. I'm very excited about the tequila. Anybody with a podcast has to have a tequila. Uh, we could actually call the tequila ConvictionStar with a Q.
- SGSarah Guo
(laughs) Okay.
- EGElad Gil
That'd be pretty amazing. I'm serious.
- SGSarah Guo
Okay.
- EGElad Gil
It could be like a Q-shaped bottle. You know how they have like the really cool bottles for different things?
- SGSarah Guo
Elad Gil, guys, our, our, uh, No Priors branding-
- EGElad Gil
Brand marketer. Yeah. I think I'm moving to LA and starting the brand.
- SGSarah Guo
If you haven't yet been a guest, please write in to the show and, you know, for the low, low price of one GPU, we'll ship you a bottle.
- EGElad Gil
Yeah, we're looking for a brand marketer to join the team too.
- SGSarah Guo
We're not actually.
- EGElad Gil
So if you work for MrBeast, just call me.
- SGSarah Guo
Elad's gonna do it. Okay, Elad, thank you for joining me on No Priors.
- EGElad Gil
Thank you for joining me.
- SGSarah Guo
And I look forward to getting my ConvictionStar T-shirt and tequila.
- EGElad Gil
Exciting.
- SGSarah Guo
Find us on Twitter @NoPriorsPod. Subscribe to our YouTube channel if you want to see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week. And sign up for emails or find transcripts for every episode at no-priors.com.
Episode duration: 26:01
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode FSHXE5r1mQY
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome