Reid Hoffman & Saam Motamedi | AI's Transformative Power

Hi everyone. Welcome to Grey Matter, the podcast from Greylock, where we share stories from company builders and business leaders. I'm Heather Mack, head of editorial at Greylock. Today, Greylock General partners Sam Motamedi and Reid Hoffman talk with Umi Mehta, who is the global head of tech, private equity and venture capital investing at Morgan Stanley. In this conversation, they discuss the current investing environment in the context of artificial intelligence technology, specifically its impact on pretty much every industry. Across the enterprise and consumer sectors, AI has evolved rapidly in recent years, and that pace of advancement is only getting faster. Som and Reed explain how AI is becoming an enabling platform technology, much the same as previous waves of tech transitions like mobile and cloud did. That means it's starting to impact nearly every sector invested in, including cybersecurity, fintech, commerce and consumer networks. This conversation was recorded at Morgan Stanley's annual TMT conference, which draws an audience of more than 3,000 investors and members of tech companies. On that note, the audio quality isn't the highest, but each speaker is clear. And if you have any difficulty hearing some of the parts, rest assured that you can find all of the conversation in the transcript, which is linked in the show notes. If you aren't already a subscriber to Gray Matter, you can sign up wherever you get your podcasts. Now, here's Umi Mehta from Morgan Stanley with SA Motamehti and Reid Hoffman.

Thanks everybody for joining.

I'm Umi Mehta from Morgan Stanley.

Before we jump in on what I.

Know is everybody's favorite topic, AI, I.

Want to spend a few minutes and.

Just talk about the overall VC ecosystem. The tone has changed, the investor sentiment has changed, what people are looking for has changed. Just kind of give me a little sense of where you think we are in the VC ecosystem today. So why don't you kick it off?

I'll kick it off and it's good timing. We actually just came off our limited partner meeting last week. I think it's no surprise to anyone here that we're in a reset moment in the venture capital industry. From the limited partner perspective, I think they're going to be a lot more discerning going forward on the managers they work with. They're going to orient to people of a track record of actually building enduring businesses that can scale over time. On the company side, we're also seeing a significant reset moment. You have a number of companies that got ahead of their skis in terms of capital, raise the cost structures they're operating their businesses at. We're seeing that normalize. We're seeing a lot of restructuring going on in mid to late stage tech and I think that work is just beginning and it's going to play out second half of this year and next year in particular. On the positive side, I would note early stage continues to be very robust and it might surprise people here that we are as active as we've ever been at Greylock. I think this week we had three companies into our full partner meeting on Monday. A lot of them have something to do with AI, which we're going to talk a lot more about in a bit. You know, we're investing early, we're investing with 10 year time horizons and we continue to be very active.

Yep. And I think, you know, part of the thing is, while obviously when you're looking at these things and there's obviously, obviously we're in a large language model AI moment, which Sam and I wrote about last year, and a bunch of other things, it doesn't mean you stop looking at marketplaces, networks of people. One of the things that Greylock has been decades, one of the leading firms on is various enterprise SaaS, cloud security, et cetera. So you keep doing that. Of course, you're also always asking in a new platform, okay, what's the new platform mean for this? What does AI as a new platform mean? And you're always asking that question, both of your current portfolio and also of course, prospective portfolio.

So I heard some sub verticals there that you've historically invested in outside of AI. Are there other new verticals that the VC community and Greylock in particular is.

Interested in on the enterprise side? You know, it's interesting because AI is an enabling technology wave and it's shifting every category that we invest in. So, for example, cyber security is an area that we have a prolific history in and we continue to invest in quite actively. And, and as you see the perimeter shift to the cloud and increasingly to home and to mobile, that whole stack is being reinvented. And so that's one example. We continue to look at systems of engagement, systems of record. On the application side, there's new themes and infrastructure. So at a high 30,000 foot view, like the categories are the same, but I think for the first time in a long time, they're going through very material disruption and a lot of these markets are up for grabs.

The other thing I would add to that is in addition to the kind of classic areas that we look at, which is on the consumer side, tends to also be, you know, marketplaces, Airbnb, you know a bunch of others networks, LinkedIn, Facebook, etc. Some of the stuff is also pure play AI stuff, right? So in terms of new categories, so like, you know, first thing since LinkedIn, I co founded a company Inflection last year, some led an investment in Adept, which is also like, has a very heavy enterprise side, but is not a classic category, it doesn't fit in the category. So there's some stuff also where that, that transition happens. And so, and I think one of the things we will see from AI as a platform technology is I think some new categories will emerge and we're looking at a lot of seed and series A deals, trying to figure out which will be those, what will be the value creators. Because. Because the short answer is now more or less, if you go to a cocktail party that's for the real estate industry and you ask what they're talking about and the answer is ChatGPT. So it's like, okay, everyone's talking about this, it's not news. But the question is, now that everyone's gaze is on it, what does the next couple years look like?

I was trying to use the first five minutes to not talk about AI.

We've failed.

But let's just go to AI. So from our count, there were 800 companies kind of identified as AI companies. The growth has 10x in terms of just volume, both in terms of funding itself. And so I guess individually, I'd love to hear from both of you what makes you so confident, so certain that this is the platform shift and it's as big as you are saying and thinking and investing behind.

So first someone, Tom and I wrote a piece six months ago, now something along that line that said every profession will have a copilot within five years. Really within two to five years, every profession will have a copilot that will be between useful and essential for what you're doing. Once you think that, like that doesn't even get to the well, how does this change infrastructure, how does this change services? Or how does this change productivity software? Or how does this change like the on and on and on security, on and on and on, even that that's changing industries. And then you obviously have, how is the product reconstituted as well. And so part of the reason why this is a wave that couldn't happen with the Internet, couldn't happen without mobile, couldn't happen without cloud, but it's building upon all three, it's the crescendo of them and that's the reason why the impact is very Very broad. Like when individuals come up to me and say, well, how should I start pointing, planning my. My AI strategy? Yeah. The answer is start playing with it. It will evolve rapidly. And. And one of the mistakes people make is they say, well, I play with chat gbt. I know exactly what it is. Well, actually, in fact, things will be a little different three months from now. Things will be a little bit different six months from now.

Keep holding that.

Don't.

Don't give it away.

Yeah. And what's more, when you look at what's being built on the APIs, even right now, it's like, well, it goes anything from new ways of constantly how you're doing marketing to how you're doing sales, how you're doing customer service. And you think, well, those are fundamental to how companies operate. And that's part of the reason why it's a really interesting, transformative moment. And the thing to start playing is not to think it's fixed right now. But like you said, well, what do I think people won't get is that the slowness of adoption will be a little bit like, most people still don't really fully know how to use search in order to solve all their problems. Like, there's still a lot of cases where people don't realize you can actually solve problem X by using search, usually to solve a problem they don't understand, is you type in six words or 10 words. Right. As a way of actually kind of getting there. It's the same thing is like, you say, well, what should I think about this question? Well, can I put it in as a prompt that will get me something useful as my output? So literally, I was talking to a friend and he's like, well, I'm going to resign my job. And so I wanted to chatgpt. I typed in give me a resignation letter with the following three characteristics. And he's like, oh, my God, that's a lot better than I would have done. He's like, yep. And that took three seconds.

Sam, before you go, I want to double click on the copilot concept. And so we have a lot of investors kind of in the room. Give me the range of applications and tools that could be a copilot for this audience.

So for this audience, think, for example, questions of you say, well, okay, what are the current trends in Technology X? What are the current challenges that Technology X is having? What companies should I also look at if I'm investigating? Company like, this is all kind of analytic due diligence questions. I think, like, one of the things is it's a research assistant that by the way, sometimes hallucinates and sometimes gets it wrong. You get it. That's one of the things. You learn and get things. But it delivers immediately. Right? Like as opposed to the hey, go, go, please do this, come back in two or three days. It's what would be. If I'm thinking about company X and I'm thinking about how it's going to transform, like say the market for databases, what are the other companies that would be really interesting in this thing? And if I was considering it to be this new kind of object oriented database or database intersected with quantum computing, what would be the other things that I would look at in terms of doing and what quite or I meeting with an entrepreneur? What question should I ask now, right now, by the way? Most of the like, having done all of these things, as an investor, everything I'm telling you, I have actually sat in front of the prompts and done which questions are too vanilla right? So far, GBD is not good at generating the question to ask. Like if I were to say I'm interviewing you, what questions should I ask? And say, well, ask about the markets and ask about, you know, it's like.

Okay, yeah, they weren't that good.

I asked, yes, exactly. Yes, exactly. Not good at what questions should I ask. But it's like an instant Wikipedia page on whatever kind of prompt you can think of that has some kind of general characteristics. Now the other one to be a little bit wary of is the training of these things is to be as interesting as possible. So when you're asking kind of theory of like overview, synthetic, et cetera, it actually hits it really well. If I said give me Umi's biography, it might invent something, you know, that PhD from Cambridge University, you know, like, okay, so when you go to the specific, you have to understand like which areas is it, is it good on? So anything in that arena and it's only beginning to scratch it.

I'd add two things just around like, you know, your initial question, which is what's the motivation for the optimism and the excitement? So one is we've been investing in AI for a long time at Greylock and in this new wave of AI for several years now. The thing that continues to surprise even us is how quickly and how significantly these models are improving. And so it's just hard to fully have intuition for exponential growth. And I think this is one of the areas where 2 quarters from now, 4 quarters from now, 8 quarters from now, everyone in this Room is going to be surprised by what these models can do and the applications they enable. That's one thing I'd note. The second is we're still early days in terms of modalities.

So the way most of us have.

Interacted with these models is primarily via text with things like ChatGPT and Image with things like Dall E and stable diffusion. But later this year, next year, we're going to see many more modalities come out. And just, you know, Reid referenced our example in Adept. For those of you who aren't familiar, Adept is a model to take actions. So instead of generating text, Adept can use something like Jeff was on stage right before us. Use something like Twilio to take actions on behalf of the end user. And you begin to think about, okay, if you have these generative models that can be interacted with in natural language and use any tool that we as humans can use, what types of things does that unlock? And that's just one example. But I think these additional modalities as they come out, many, many more applications are going to get built.

And to give you an example of like, the applications that will be live, like basically this year, like another of our portfolio companies, I'm on the board of CODA is take notes in the meeting, give a summary, give you what your action items are, the triggers. Go to oh, now, now send an email to UMI about this. Right? Et cetera, inform. So and so all of that stuff now becomes as the assistant, all becomes doable. And that's the reason why there's a stack of companies, because that's a, you know, essentially an application company and productivity.

Well, I think you're bringing up a great point because I do think there are kind of theories in the market that it's about elimination, not enhancement. And so you just described the enhancement of the position and the job and the to dos as opposed to elimination. There's also some themes around just critical thinking. And so when I think of the limitations around critical thinking, will we have to be less critical in the way that we do our work? Will these kind of LLMs kind of catch up to us? Like, give me the critical thinking process.

Let me take this one first. So like, for example, take ChatGPT comes out and the education establishment has a collective aneurysm going, oh my God, we can't assign s essays anymore. It's like, we've got to run this still. Like it's the 1940s, right? Which is when we had all this stuff. So I sat down for three minutes to think about how could I say GBT and teach essay writing One version of critical thinking. Here's two. One is to say your essay is the argument's for and against. And then you have to make both arguments and then you have to conclude as to what's there. That's something that you can use be using GPT to help you with, but you have to essentially still get to that conclusion or say you don't want to do the, you know, pro side, antagonist side. Well, okay, so I'm an English teacher and say the assignment is, you know, Jane Austen and colonialism. I go to whatever the current ChatGPT is. I generate five different essays of my prompts. I hand out those essays to the students saying, I've done this with gbt. This is the D plus. You have to hand in something better so it sets the new benchmark. They go and you go use GPT. And by the way, I can use GPT to help grade and all the rest. But if you just typed in a prompt saying give me a essay about Jane Austen and British colonialism in Africa, you're going to get one of the five essays. You have to figure out how to make the essay better than what you got. And that's the reason I do that as a lead in using the education thing is critical thinking is still be there. It's still going to be amplifying. You're looking at this thing like when I was saying what I learned for having been using GBT for, you know, a year and a half now in various ways is I learned which things I go, oh, that I don't want to just rely on.

I want to cross check.

How many times do you use GPT in a day?

I'd say at least four and sometimes 20. Okay, specific. Yes.

The thing I would add is it's remarkable how effective ChatGPT is at, at Socratic reasoning. And so if like actually I use it this way, which is if you're thinking through a problem as a thought partner to help you reason through it, it's actually very, very effective.

Hence again, critical thinking. I think the anti critical thinking thing is a lack of thought on the part of the critics. It's clearly just use it as an amplifier and you still you go further down that path.

Sam, you mentioned there's been massive adoption. But when does it become ubiquitous within certain sectors? And like what's the adoption curve kind of look like? Which sectors will be affected the fastest? Maybe which ones won't be at least until there's some other version out Comments on some of those things.

Sure.

So I'd start by saying it's already ubiquitous. So you know, if we're using Google Search, parts of Google Search are publicly discussed as being powered by large models. Those of us use different social media applications that have recommendation systems are powered by large models. A number of the existing existing application companies that have data gravity. So whether it's a salesforce with their announcement of Einstein GPT or Workday and things they'll do on top of HR data, I actually think this is proliferating very, very quickly. The other thing that's remarkable about these companies and a data point I can share from the early stage perspective is because these tools really feel like magic. Like the first time I used ChatGPT, it certainly felt like magic to me. And the level of value they deliver end users is a step function over prior generations of software. The velocity with which they get adopted is also unlike anything we've seen. So for example, just to give you one example of a concrete case like we incubated a company called Tome at Greylock. Tome is a generative AI way to tell stories or to build presentation decks. This company reached a million monthly active users faster than any productivity tool we've seen.

The analyst and associates in here look very worried right now, by the way.

That's a copilot for analysts.

It's a copilot. No, it's. But for example, so I'm starting an additional podcast called Possible, which is how do we imagine what future is possible? How do we build toward the technology? I sat down in front of Tome and I said, write me a deck about a new podcast by Reid Hoffman, Possible that says technology is between 30 and 80% of the solution of any scale thing. The problem problems can include criminal justice, economic justice, climate, et cetera. Will involve AI. Give me a deck. In five minutes, I had the beginnings of a workable deck with 10 slides, graphics, text, et cetera. It was not the deck I would stop with, but I was there.

You saved a week or you saved a couple days.

Yes.

Yeah.

The other thing I would add is just so, for example, one area Reid and I talk about is customer service. So if you think about areas inside the enterprise that are more of a cost center, where there's more of an orientation towards efficiency and where often the types of things happening are more repetitive and I pick on the contact center. But you can look at different categories of kind of work to be done that today might be serviced by BPOs or different automation players, those areas lend themselves really well to augmentation and eventually for some subsets, automation with these technologies. So I think that might be another area where you see things begin to take off quite quickly in coming quarters.

Lots of positives, let's talk about some of the potential negatives, whether it's using data to your advantage, some of the data privacy, who actually owns the content at some point in time. And so maybe there's a lot there to kind of unpack, but I'd love to hear how like should there be some agency set up within the government at some point in time to regulate kind of what comes out of ChatGPT or some of these other kind of models.

Yeah.

So I'd say you're absolutely right that these are an important set of challenges and those of us who have placed with these products, we see the challenges like read reference, hallucination. There's questions of data ownership, data privacy in an enterprise context. For these models to really get deployed in mission critical applications where efficacy and accuracy matters, there's a gap that has to be closed. That gap will be closed and it'll be closed in two ways. One is you'll see companies like OpenAI and others find ways to adapt these models to enterprise environments that, that preserve the integrity of the enterprise's data and give better guarantees around performance and not hallucinating. So imagine ChatGPT for Morgan Stanley on top of Morgan Stanley sensitive data in a way where the answers it comes back with are Morgan Stanley specific. We're going to see a lot of that type of adaptation happen with regulatory logging.

And all this stuff is a hassle in your life.

And then the second is tooling that gets built from an explainability fairness, bias perspective that helps make these generations and predictions more tractable and more explainable and that also inspires confidence and allows them to be used more and more in mission critical applications.

One of the things I think that people don't realize is the part of what happens is these large language models are trained on massive amounts of data and it's the mass amounts of data and the general model that gets created there, which doesn't involve any of the things that people are worried about is high value data and so forth actually doesn't even involve any of that. You can then tune them to high value data and then once you tune them, because we don't really have explainability yet, you would then have to keep it contained to whatever that tuning is like. If it's tuned to Morgan Stanley data, you have to have an Authentication for using it. And we have to log everything because this is a 200 or 300 billion parameter model. You can't look under the hood and go, oh, this is where the data is and so forth. But people then say, well, if it was training on, you know, publisher, did it train on Publisher X data? Actually all the base models almost don't do that. And it isn't like, oh, it's, you know, Publication X because it has a really good columnist. It's often massive size of data and it's tuned and the tuning is part of what gets the copilot thing that Sam and I were talking about, which is, well, you can tune it for writing code, you can tune it for writing poetry, you can tune it for doing medical stuff, you can tune it for. There's a bunch of different things you can do that once you have that general model.

Specialized verticals.

Yes, you can do specialized verticals. Sorry, you were to say that.

No, I was just going to add. And I think when people think about tuning, there's like, there's different levels of specificity. So we talk about code. We're all familiar with Copilot. It's having amazing adoption. But if you just take code as a modality and you drill deeper, for example, you look at code quality, code security, you know, infrastructure, instrumentation, from a monitoring perspective, you will see models that will be tuned to those specific use cases and that's how you'll see applications get built that really have a chance to go disrupt these software markets.

Where do you think AI ChatGPT will play into the election next year and where is it going to get its information from?

Well, one of the things that's kind of cutting edge part of what, you know, Bing's announcements have been and so forth is how do you. These are trained on a bunch of data and then they're kind of fixed. So how do you bring in live and current information? That's a work that's very rapidly and iteratively progressed. That's part of what the February Bing announcements are gesturing at. I think there's intersections between that that will be important to how those play out now when it gets the elections. I think that part of the question will be is they say, well, we'll have a. A lot of generated misinformation. It's like, well, we already have a lot of generated misinformation and some of it's from actual people, some of it's from paid people, some of it's from other things. It's not a don't pay attention to it. But I'm actually somewhat optimistic that we might be able to figure out in this kind of these time frames and the actors are saying how do we help? Like for example you said well I have a copilot that helps me ascertain things relative to say, medical knowledge. And I'm looking at well, vaccines, you know, cause Ebola, you know. And I'm like he goes eh, right. You know, and so. But then that's part of where you say well look, it's actually in fact also part of the solution now you have to have people who then say look, I care about something that's trained and tuned on a set of expertise. I'm using it as a copilot in some manner. But if you get that and you have, have that deployed in various ways and it's, you know, makes you know, maybe some product warranties or claims or other things that, you know, have maybe teeth to them, maybe not, whatever. Well like misinformation has been a very common subject around AI for X months. Like well how would you solve it? And I'm just like to sit down for three minutes, say how would I deploy the technology and try to help? And solve doesn't mean solve to zero because by the way there's tons of misinformation on the Internet. There's tons of misinformation within Google search. There's tons of. But it's within the parameters of how we're operating.

Maybe the next question for me is a little around. I talked about what goes wrong and I just like this question. Utopia or dystopia? So maybe Utopia and I'll let you take Dystopia.

So I mean I'll start with Utopia. You're going to hear a recurring theme because it's how important we think this paradigm is, which is the copilot paradigm. I think if you take both what we're all doing in our personal lives, whether booking a trip, you know, thinking through a problem in your personal life and in the enterprise context, what we're doing at work, we are going to have hundreds of tuned co pilots that make us 10 times more effective and 10 times more efficient at everything that we do. And I'll just give you one concrete example. We talked about customer service earlier. We're investors in a company called Cresta. It's a co pilot for the contact center. So if you call into Verizon today and you're talking to an agent in Tempe, Arizona, they have a copilot that's Steering them through that conversation and helping that person sell a new cell phone subscription, earn more money and be more effective at their job. And that's one example. But it is going to proliferate everywhere and I think it's going to be a massive step function in knowledge, work, productivity.

So on the dystopia side, unfortunately, I think a lot of the discussion is a little bit misleading. You have the oh God, it's going to become self aware. And I've watched the movie Traffic, Terminator and so forth. Part of the reason I tell people, go play with chat GBT is you go play with it and you realize, oh well, these are the things it's good at and these are things it's not good at. And for example, right now it can't solve. Like you say you're an expert in investing and you ask a question you're expert at, it will not give you a better answer than the one you.

Have in your head.

So then what do you say? Where are the downsides? We should be tracking well in a kind of transformation with all the co pilots, I think many jobs will actually not be downside. I think that we'll take infinite of those jobs. We'll take infinite engineers, infinite salespeople, infinite marketing people, etc. Customer service I think will be one of the ones that'll be a little bit more challenging. And it's just like for example, once we solve autonomous vehicles, we'll want all the vehicles to be autonomous. We won't have the gridlock that at least I experienced getting here. You might have to, you know, the other kinds of issues that I'll have, climate change, so forth. So that will also be a more of a replacement. But there will also be creation of new jobs. It isn't just like when you get 10x more efficient as a graphic designer, that doesn't mean there's fewer graphic designers. There's a lot of things to do. But the dystopia that I'm gesturing at and saying, this is the transition, not oh my God, I'm worried that it's going to be like Elysium, the movie, because you know, obviously Hollywood films do a lot better with dystopias than they do with utopias. You don't go, oh, there's going to be this small group that lives on the moon and all the rest of the people are whatever, and it's a robot police force. I really think that's like grounds to zero in our grandchildren's lifetimes. And who knows? I don't Know, make projections in that, beyond that in various ways, just because, like, look, the world changes. And I'm generally optimistic, but the transition will be things that we need to be. We need to be intelligent. When we did the transition from agrarian to industrial, it was very difficult on society. Now even truck driving say, oh my God, all the truck drivers jobs gonna go. One, there's huge number of empty listings right now for truck drivers. Two, if all manufacturers on the planet started making autonomous trucks, right now, it's 10 years before you replaced about half of them, right? So, you know, it's not like, it's like, oh my God, tomorrow is gonna be totally different. So again, transitions, I think are really important. The other area that I would go that I think we haven't seen anything yet is cybersecurity, which is, you know, like, people say, well, like, oh, I'm worried about, you know, like the machine becoming super smart. Like, okay, that's like, I'm much more worried about these tools in the hands of malicious humans doing things that malicious humans do. Right? And so the questions about what that that means and what you do there and what happens and are you ready to deal with that is another area that I think is a legitimate area of dystopia that we have to be careful about. So to go like all the way down to one of the reasons why OpenAI maintains a very high ethics standpoint on this stuff. Dall E was ready for four months before it launched. I was playing with Dall E for four months before it actually launched. Why? Because they were like, oh, it could be used for other really terrible things that would have, let's serve it through the API and let's tune it to safety for all that. That's the kind of thing, it's like the humans using it for very bad things is the thing you have to track. Now in many cases, like to your earlier kind of regulatory question, look, I think dialogue and goodness for societies are important, but this will be redefining industries, right? And what we want as societies, as investors, as inventors, as workers, we want to be in the industries of the future, right? We don't want to be Luddites, right? If you say, well, no, no, all looms should be hand done, right? And we should just stay with hand done looms. That position is a very bad position in a very quick amount of time.

Goes back to my. Who's going to regulate?

Yes, but I think that the right answer is not to say, oh, you know, regulators should go drink Mai Tais on the beach and look the other way. Which is sometimes how they hear the don't do anything rash right now. It's like no, no, no. Get engaged in conversation. Like the thing that it's like, what's the outcome you'd like to see? I'll give a parallel. Like say, well I'd like to see very few acts of terrorism and online video. Should I then send. Well, I'm a regulator and so I'm going to create a process like you have to delay broadcasting it by 15 minutes and da da, da. And this is now how we're going to lock in what you have to do. And that's how we're going to do it. You're like, well that's probably not even going to work. And it's like, why is a regulator a good technology design? What I should do is I should go say okay, I realize getting to zero is probably difficult, but let's say the first hundred impressions of saying you have to run through the auditor, that's you turn off. When you got to 100 impressions, you're fine. You get to a thousand, that's a million dollars. You get to 10,000, that's $100 million. You figure out how not to do this. This is what I care about. I want you not to be doing this. You figure out how to do it, we'll give you an economic incentive to do it and then you improve it as you do. Like that's a much, much better way to do this. And that's the kind of thing of like everyone kind of goes okay, what should we be doing in AI? It's like, look, we should be embracing all things. Like for example, what I see line of sight right now is a AI tutor and an AI doctor on every cell phone. If you slow that down, think of the human suffering that you're creating by not having that, right? Like that's super important to get to.

How far away are we from that?

You can build it on today's technology, right? It is doable today. Now you know, there's work to do, interface, bunch of other stuff, but it is doable on today's technology, right? That's the thing. Like if you say, well I'm a government, I want that for all my citizens, please, that would be really good for the well being of my citizens. I want that. That's what you want.

There's probably lots of questions so maybe we can get some mics around.

First question at Greylock, where do you all think most of the value capture is going to happen? Is it going to be at the application layer, the data layer, the closed source, open source model layer, proprietary models, or for that matter, even in the computer. And also, how does one monetize? I mean, you spoke about ethics, you know, biases, whatever might creep in. Right. And also, I'm pretty sure there's data of mine that's being used to create some model somewhere. I wouldn't be surprised. How do I get monetized? And is there, are there unique models that are emerging to help monetize that data that you're seeing?

Maybe I'll take the first part and.

Hand it to Yurid for the second.

So it's a good question. There are three layers and I wish I had like an either or answer, but the answer is like all of the above. So we think of it in three layers. There's the underlying foundation models themselves and our view is there's a select number of players and is sub 10 who have the ability, understanding capital, compute required to actually build these foundation models. We're nowhere close to the end of returns on scaling. And those players will create value. There will be differences, some will focus on one modality versus another, but there will be that layer and that layer will monetize through an API model, you know, like Twilio and others that we're all used to. Then there's a layer on top which we think of as like the infrastructure and orchestration layer. Right. And so for example, like you want to now take one of those things and adapt it to an environment like Morgan Stanley. There's a lot of things you have to go do. There's a data management piece, there's a serving piece, there's a integrating it into my internal application piece. And there's interesting opportunities for new companies there, new tooling companies there, and we're investing in those. And then kind of the top is the application layer. And I think there you're going to see both existing applications that own the end user have some proprietary data asset that has gravity that's a consequence of their workflow. They're going to incorporate these LLMs into their businesses and it might be a turbocharged for kind of existing. And then you're going to see categories. Tome is a good example where the whole ergonomics of how you interact with software shift because of these models and that creates new opportunities for new applications. And some of them will be in existing categories, some of them will be in categories we don't really know how to name because the ergonomics weren't previously Possible and those businesses will monetize like application companies today do. And so I think our view is like very optimistic around all three layers, recognizing each of those layers has different dynamics in terms of what's required for success.

And then the other thing is I think there will be in business models, I think a bunch of the standard business models, I think subscriptions, I think ads, I think a bunch of those things work. I think we will also see some new invention business models that will be unique to technology, just like AdWords is unique ad technology for this stuff. And so I think that's one of the areas we pay attention to now. The data thing is one of the places where a lot of people are just misconstrued. What's the value of this data? And well, the value of the data is to in what context? So like for example, you're in a search and you're typing mesothelioma super valuable, right? I have a possible intent and issue and so forth. Or Sony camera super valuable. I mean Word documents, typing it, not that valuable. And they may even be, you know, kind of signals and so on. And so part of the thing that people understand is there isn't like, you know, the fact that Reid is wearing a blue jacket is not like an intrinsic like that's a two set item piece of data. It's a context which it is. And part of what these systems build is they go, we try to provide you systems of value by which you're engaging data with us. Sometimes by the way, that's I SEO things because I want to be discovered in various ways and then it's in public indices and because that's valuable to me in some way. And then other folks, you know, that trade off has figured out how to use that data in things. And so for example, I had a very funny conversation five or six years ago with someone in Silicon Valley who said what does Google give me for my data? I'm like well research, you know what I mean? So the question is what that context is. So I'm quite certain that all of us data is being used in multiple contexts because we're engaging in various things where our engagement in that is somehow the data is part of what we're being attracted into doing this thing that like for example, posting a web post on and seoing it because I want people to discover it. Well then it's also out there publicly.

And I'd add just one thing, hearing Reed talk about kind of new business models, I think there's new business model invention but another thing we talk about is evolution. So if you take something like ads, right? And just to give a concrete example, like take a software developer who's building a piece of software and goes to a search engine and makes a query and there's ads around different software packages the person could use. Well, in a world with copilot for developers, you still may have that flow happen. It just doesn't look like a query and a set of links. It looks like autocomplete in your ide with a suggestion around a promoted package. Now you still can have that same concept of advertising. It's just going to evolve in the way we consume it.

I have a question on compute usage. So over the period 1960s to 2010, the compute usage doubled every two years. But in 2010 to 2020 it doubled every three to four months. So if you can conjecture 10 years from now, you're sitting on that stage, what do you think that doubling time would be? And are we as a human society constrained by either hardware or data to continue that kind of progress in this AI model?

Let me just throw one in there. Does capital then become the competitive advantage to have all this compute power?

Oh, it certainly is in some contexts, right? It depends on which. So I think there's, on the kind of scale models, I think there's the players who are going to as absolute great scale as they can. And you say, why is it still valuable? It's one of the things you and I were chatting about. We just say, well, if you actually can say I spent a billion dollars and I have a 20% better engineering aid, that's worth it. 20% better doctor, that's worth it. 20 percent better lawyer. That's worth it. It's worth a billion dollars. It's fine, right? In order to do that, you get the Internet and all the rest. And it's a great, great question because I think that we are. Compute is well thought of as like energy, which is we have infinite demand for it at a certain price. And so as long as we can deliver on that price, we will have infinite demand for commute. And so I think the constraint of your acceleration curve is actually a economic constraint, right? Not a natural law constraint. And by the way, of course, part of the cost of providing commuters, you have to build, build data centers and places to do it. Like one of the ones you didn't mention in your variables is power. Like power is a huge variable in all of this because compute is not the only customer for power. As we sit here in this Hotel and so forth. Because part of this AI revolution is a scale compute revolution. That's part of the reason why when I was answering the earlier question saying it's Internet plus mobile plus cloud, what we're doing is saying, oh my gosh, we now figure out how to apply a whole bunch of scale, scale compute to creating amazing things that hadn't existed before. Well, we don't yet see that if you have 20 exaflops versus 10 exaflops. 20 exaflops still better. Right? And so that like the compute demand will be very high. Double clicking on your first layer, you talked about sub 10 foundational systems. You've been interchangeably using chat GPT or GPT OpenAI product. And so go a little deeper onto that first layer. Are there sub 10 large language models or we're talking about some additional natural language processing models? What's in there? Or is this one revolution you're talking about today purely about large language models? It's a great question.

So when I talk about that layer, I think about the distinct companies that can train these foundation models around different data modalities. So it could be language and text with something like GPT, ChatGPT being an interface on top on how you actually interact with that underlying technology. It could be image like the Dall e model from OpenAI or the stable diffusion models that are open source. And from stability it could be audio speech. We're going to see all sorts of these.

Or adept in action.

Adept in action. We're investors in a company called Atomic that's building a foundation model for RNA structure design that can be used for therapeutic design. So we're going to see a proliferation. But these models, you know, connecting to the prior point, they take there's, you know, order of magnitude 100 people in the world that have actually scaled these models to the levels of scale that we're at today. And certainly, you know, it's going to grow from there. So in terms of capability and then capital access to compute and Data, there's likely sub 10 companies that will have offerings at that layer. And so that's what I mean when I say sub 10. And there's OpenAI, there's anthropic, there's adept inflection, Google to name five.

And the other thing is part of the reason why we're so focused on large language models. There is somewhere on the Internet 10 to 14 trillion tokens of data how we can train these things to be really interesting off those tokens of data. Many of these models are trained on about a trillion tokens of data, maybe a trillion and a half for the large ones. And so there's a bunch of headroom there and all kinds of things. And you're not even getting into one of the questions we dealt with earlier, which is how could you fine tune within an enterprise's data or other kinds of things? So anyway, there's a stack of things there and the large language models are the first that are generating this thing. Like oh my gosh, I can see the copilot, I can see how it's working. There will be other scale compute techniques. Large language models I think are now like certainly one of the important tools in the tool chest, maybe in three years, still the most important tool in the tool chest. But maybe it's not. Like for example, if you look back to the very early kickoff with what DeepMind was doing, self play is another scalable compute thing. And so there's various ways that self play comes into these things that also generates interesting things that when you move from, you know, two exaflops to, to 5 to 10 to 20, you get an increasing performance curve. The thing I'm nearly certain will be is that within three years another interesting major scale tool will be oh yeah, we're using that one too. And which one it is. Well, I'm spending time scratching around trying to figure it out because that's my job, but who knows. Hi, I wanted to ask about the copilot thesis and the part that the mobile phone is going to play in it. Like obviously we're seeing the models becoming bigger and bigger, optimizing, but still becoming bigger. I want to ask how much do you think the barriers of the mobile phone to actually have the models locally installed versus API calls is important for the copilot paradigm to actually exist? And how far do you think we are away from that? So I can share a quick thought on that.

And Reid, please add so or maybe two quick thoughts. So one is it's very use case specific. So there are use cases where you can use the mobile device to interact with models on the cloud and the latency is fine. You get things back in the speed you need for the use case and candidly I think a lot are going to play out that way. So that's one thing I'd say we're there today. But the second thing I'd say is it comes back to this concept of smaller models being tuned for very specific use cases. So while you're right that these foundations models that are solving really horizontal Tasks are getting larger and larger. You know, pick your favorite example. But if you take an AI tutor for, you know, fifth grade math, you can actually build a small model tuned to that and that's actually highly performant at that and that could run on your mobile device today. So I think like the combination of those two things, depending on the use case will solve kind of the mobile challenge.

Yeah, it'll be more I think what are the ones that are specific that can't have the latency of the network that then have to be on it. So like for example, you said, well, I want to use my mobile phone as a driver assist that will automatically quickly recognize like there's something and then do something. Don't have time for the latency. So it has to run in the context. If I'm putting my mobile phone thread.

Algebra, I think you're okay.

Yes, but there's algebra. Oh wait, it was 500 milliseconds or even 2 seconds cares and then so much more. I would say that we probably haven't really taken just to give you a sense is we don't really spend a lot of time at the Graylock Partnership going what are the models that can only run on the phone because there are so many interesting super valuable things where the network latency is not a problem. And so you get cloud plus phone. We talk about AI being the next platform shift or next shift of computing. And and I guess if you look over the last few, whether it goes to cloud mobile seems like they've all been won by existing winners and we haven't really had a huge win for new players since the Internet. So why is this any different where the key winners aren't Amazon, Microsoft and Google, who already seem to be really far ahead. Well, if you mean cloud like aws, then yes. But if you've been cloud as a revolution, there's obviously Workday and know Salesforce and a whole stack of companies there. I'd say that the short answer is both are going to win a lot. Not an either or. It's kind of, it's like false dichotomies. You tend to say the well which one's going to win more and so forth. And I think we are going to see great things by the tech companies that do the revolution. Like for example, like you say mobile. Well, it's what resurrected Apple, right? Apple wasn't Apple before the thing. So and so I think that it's just this platform thing is it depends on which. Now if you say well, I really like to do is I'd really like to build a billion dollar computer and be competing with them then. Well, you have a bit of a challenge as a startup. You better have something pretty unique and interesting idea. But that's like for example, also if you had come to me 10 years ago and said, I want to launch a new desktop search engine, you're like, ok, you need new platforms, new technologies, new market opportunities that are not being taken in order to do that. So anyway, I think that it's going to be a huge amount of opportunity across both and growth. Yeah.

And I mean two things I'd add. One is you're going to see dramatic business model disruption that creates the opportunity for new companies. Just thinking about that, we were talking about search and advertising as one example. And then the second is I think one thing that may not be fully kind of priced in yet is how different some of these categories are going to feel because the interface changes.

Right.

And like, you know, we're all now used to interacting with ChatGPT and other tools with natural language. Natural language becomes a very ergonomic and flexible interface to do different workflows. I'd argue it's a much more material shift than the client shift from desktop to mobile and how these categories of application software are consumed. And so what it means to build a PowerPoint or tell a story is going to be very, very different built on top of these technologies. And I think that might create more of a window for disruption.

Actually it's one amplification because I love this comment I read a couple weeks ago, which is what's the biggest programming language on the planet?

English.

Yeah, exactly, exactly. As a function of that on sort of the monetization end, I guess you can think perhaps on the copilot thesis, the upper limit could be what humans or what companies are willing to pay.

Humans to kind of get that work done.

So as we think about moving this.

To more of a technological model or moving it more to machines, are there.

Constraints that you think that now come.

Into play in terms of that could change where we land on that monetization.

Gap of, you know, starting either from free or going up to what we'd pay humans.

Any way to think about how the.

Monetization could land in between there? It's a very complicated question that will change a lot year by year and in different functions. But I was gesturing at principles when I was kind of saying like compare and contrast sales and customer service sales. When you go to a company, you go, well, you have a 10x amplifier on your sales capabilities, like, great. Like, we still like to have our whole salesforce. We'd love to have that, you know, kind of, you know, like, make that better for us. If you say you have a 10x amplifier in your customer service, I go, ooh, we can reduce our clients cost line. Yeah, right. And so, you know, those are places where the kind of the variables will. Will play off in this. And that's reason why there will be some replacement. But I think that the thing that is when you really begin to look at it, you'll find that even on their existing many job cases, when you look at it, you're actually looking at things that. Where the amplification and the use of it is much more interesting than that. Now, that's reason I'm fundamentally kind of optimistic. And fundamentally Elliott on the Dystopia case was saying transition, and you're like, what does the transitions look like? But I do think that, you know, that's part of economy is how we run. So it's like, okay, well what do we pay human for this? What do we pay something else? Do we have to like businesses? You know, what makes businesses really effective is they think, okay, can we provide this quality of product or service and can we do it in a much more competitive way at a cheaper price? That's the iteration. And so figuring out how to do that. We do have time for one more question. There was an interesting paper recently. I think some researchers in Japan, they hooked people up to FMRI machines, they showed them images, and they took the data, they fed it into an LLM, and it spat out the images with pretty decent fidelity. So I'm curious whether you spend any time thinking about novel sources of data signal that maybe were only kind of sort of useful, like data exhaust coming off of all kinds of different things that maybe were not quite commercializable in a relevant way before the advent of large language models, where now all of a sudden, there's all kinds of interesting stuff that you can do. So the short answer is yes, that's among the things that get to like, where are the interesting new companies or business models? Or do you have a data source or an angle and a data source that causes that? Now you can create a really interesting product that couldn't exist before. Maybe it has a different business model. I know about that paper. You make this large language model, this foundational model, and it's 300 billion parameters, and it's trained off the patterns of what we encounter through human language and images and words and mapping them because part of what, like one of the things that people don't fully realize is in order to do the image generators, you had to have the text things first because you had to be able to say, if I would like to have a panda bear in an Astro suit on a rocket and so forth, that has to understand what all those categories are in order to be able to classify them, be able to produce them, et cetera and all that. And so you have that, and that's a massive platform layer that then you say, oh, actually, in fact, I'm just doing a translation, I'm doing a translation of what this neural pattern looks like to that since it's trained on all this human data. One of the things that we've been discovering is the larger and larger the model, the easier it is to tune to what human concerns we want it to have. But also means it tunes to other aspects and kind of human life. And so the short answer to your question is yes. And that's one of the threads that we spent a lot of time looking at. Yeah. And I would.

The one note I would add is if you just generalize that to data, like earlier, there's a question around compute. We talked about compute, but data is an equally important dimension and at all of these companies, that's something that's top of mind. And it's both publicly available data sets, data sets that are proprietary, looking for exhaust from new sources that haven't been considered. And then the last thing I'd note is also using the models themselves to generate more data. And there's a lot of interesting work happening there.

Well, that was extremely interesting and very thought provoking conversation. So Reid and Saam, thank you again for joining us, really appreciate it and thanks for the audience for the great questions.

Thank you.

That concludes this episode of Gray Matter. You can find a transcript of this interview linked in the show notes and you can subscribe to Gray Matter wherever you get your podcasts. I'm Heather Mack, thanks for listening.