AI-generated — may contain transcription errors.

Scott: How do I turn on my other camera?

sam lessin: You just turn on all your cameras. This is and needless to say, guys, as you probably have figured out, there are actually human editors on this. So if we say anything truly stupid or we just ramble for a bit, it's fine. Guess they can you can like ex nay it slash we can start like this and it's completely fine. but just give me one more second to pull notes and then we'll be ready to go.

Scott: Right.

Josh Wolfe (Lux): Yeah.

Scott: Mm. Can't change camera while recording.

sam lessin: That's good. let's see.

Scott: Dang.

Scott: But Rachel, yes, if we end, I would love to ping you next time in D C and catch up. It'd be fun. A lot to catch up on.

Rachel: For sure, yeah. It'd be great, yeah, for s for sure. Now people have a reason to come to D C which is like

Scott: Yeah.

sam lessin: Yeah, for sure. okay.

scott: Has anyone played with the the the the new grokbot extensively yet?

sam lessin: Alright, you gotta let me do my Jessica intro, then we can you can just jump right in and ask that, Scott. Alright, ready? Hello. And welcome to a ridiculous episode of More or Less, where my co-hosts, Dave Morin, Britt Morin, and Jessica did the terrible, terrible judgment error of letting me host without them. they're all actually on vacation, and I am actually also on vacation, but pretending like I'm not on vacation. So I'm thrilled to be here. We have assembled

scott: Okay.

sam lessin: A world class community here. we represent, I think this group represents every flavor and every area of venture capital investing known to man, as well as like every product experience you could have managed. So I'm very, very excited to have a great group here to run through. first we'll start with we have Peter Dang here. Peter actually has a distinction. Peter the one who brought me into Facebook in a lot of ways. He was running profiles right before I was. And so when I showed up on day one, he said, Great, someone else can run this. and taught me everything I need to know and then was very happy to piece off into his next project, which ended up being Instagram and then Uber and then Airtable and then after that big role running product at OpenAI and now is a dirty venture capitalist like the rest of us. So Peter, welcome. it's good to have you. You're GP at Felicitous. Holt, Rachel Holt. I've actually known Holt longer than anyone on this call.

Peter: Thank you.

sam lessin: I first met Rachel on our first day of Bain and Company orientation in New York in 2005, when she we sat down at orientation with about seven of us in Bain, New York. And I love and she picked up as soon as there was a break in our seven-person onboarding. Why not? It's a great story. It's my favorite, it's my favorite story. So she picks, she she is the it was like she basically we have our first break at Bain and Company in a training. This

Rachel: You cannot tell this story, Sam. I know exactly where you're going. I knew that this was gonna come up.

Scott: huh.

Josh Wolfe (Lux): You must tell the story.

sam lessin: Girl who like is with six people she doesn't know, her new coworkers, takes the spider phone in the middle of the conference room, dials Comcast, and starts yelling at them. Just to set the tone of like this is an amazing, I love this. So, but Holt from there has done amazing things. She ran Uber North America. She

Rachel: I mean, first of all I got my cable attached on day one. On day one, yeah, August first, two thousand five, Time Warner was there.

sam lessin: You did get did you get it that day? That was amazing. No one had their cable attached.

sam lessin: Boom, boom. So Rachel is a huge her he took she took those operational shops to then run Uber North America, has started Construct Capital, represents the rare DC VN Venture Capitalist, but a breed that is growing. And we do get to do a lot of business with her now. So Rachel, great to have you on. Scott Belski, I mean, look, Scott's done everything. I knew him originally when he was running Behance and founded Behance, which he sold. he's now a partner. At A24 and thinking more than and and running A24 A Labs founded it. And it's thinking more about kind of the future of Creative than anyone I know. So I'm thrilled to have him here. Scott Stanford, I mean, you started out in an even dirtier job than Venture Capital, running Goldman, or a big piece of Goldman for a long time on the TMT side. You can give me the you can correct me on the details, but you somehow did that forever, and then have had a l an illustrious venture career. Now is the co-founding and partner of of Acme. and then Josh Wolf, who is does do you even need an introduction at this point? He was doing of course. So Josh, the co-founder and managing partner at Lux, which has done basically did the contrarian thing, which has now become consensus around defense and deep tech and hard tech, and is kind of one of the men of the moment for things calling things like Androil, et cetera, early and backing up the truck when that wasn't cool, which is now very cool, which means

Josh Wolfe (Lux): I need I need I need the longest intro no.

sam lessin: Hard to seed invest in anymore because these numbers have gotten so big. But so we have an amazing gang here. and I I have my topic list, which obviously I had my bot right of what's going on to run through about what's going on with NVIDIA, this crazy balance sheet stuff. Peter, we're gonna have to talk a little bit of Airtable, given the fact you ran product there for years and I was an investor and things like that. But Scott, you started out when we were BSing before before hitting record. We wanna start with Grokbot.

scott: Well I yeah, I mean, I think that the agents have been a topic that we've all been thinking about and talking about for quite some time. But we're like just in this moment, it seems, where they're becoming like really useful on a consumer level. And I don't know about a lot of you, but I've been playing with some of these. They're always a little scary, these startup agents, because they ask you to like log in via Google and give it access to everything. And you get those like crazy Google warnings of this is not a authorized, secure, blah, blah, blah, but you do it anyways. But I have to say, like they're kind of blowing my mind. and seeing some of the demos that people have shared in the last 24 hours of their Grok bot, having and I think it's a combination of obviously the agenda capabilities, but also the memory and this notion of like having the the bot have a con computer instance of its own and even seeing their computer set up. It's truly this notion of someone with a screen, and then you get this like consumer click of anything you can do on a screen. they can do on a screen is like a really fascinating turning point, it seems.

Scott: It it's it's a shift from help me do this to just do it.

sam lessin: So yeah.

sam lessin: Yeah. I mean, I don't know about you guys, but like I I'm like I'd like to put out that like I'm so down the rabbit hole on this that none of this stuff impresses me. Like I I have a persistent Claude instance running in DigitalOcean. It I actually did I had it do its own analysis and math right now. I forget the I'm not talking about the AI. My bot is running about three grand a month of infrastructure. database app servers, et cetera. Whenever I want something, it builds a new app for me. It like runs its own databases. It sets everything like, and I actually interact all with it through email. So I like basically start email threads and I'll like ask it 50 things and then just it dispatches them, build an app, analyze my health, send me this cron job, whatever. So like I've given my bot so much infrastructure that I'm not that impressed with it as a computer. But I'm kind of curious if any of you guys, I mean Scott, you've built a c some crazy stuff with AI.

scott: Well you're you're not normal, Sam. So it's like I think that but it's I think it's a question, yeah. I'm curious what others think, but it it does feel like there's like a crossing the Rubicon moment for consumer accessibility and just the ease of onboarding and giving it authority for certain websites and just it's something seems to be clicking to me.

Peter: Yeah.

sam lessin: Yeah.

Scott: But it's it's a rapid evolution in an old paradigm. Like this concept of websites is going to disappear very quickly before we know it. And this concept of like help me do a bunch of stuff versus just do it, predict I need it, get it done. And yeah, I love the Comcast example, Rachel, of like you having to call Comcast. Forget that. Like there's no humans involved anymore. It it's I simply I don't even need to tell it that my cable's out. It knows my cable's out, it will reach out to the agent at Comcast, it will sort it out. And so I I mean my no, no, but I mean, where were we last week, right? So w we look at this with such a a microscopic lens. This is gonna happen before we know it. And then we're all gonna be sitting around looking at each other saying, What are we supposed to do?

scott: Not yet, right?

Josh Wolfe (Lux): When

sam lessin: Maybe, but Comcast doesn't but Scott, Comcast doesn't give a fuck. Like they're not gonna like solve that. They you this it's like a regulated they're

Scott: no, it's it's our agent, our the the reason why they're gonna give a fuck is when they get bombarded by a bunch of automated agents that are hitting them twenty four seven saying fix this and they have to respond. And so then they rev up their agents and then it's agent to agent and we fish or we ride on Boston whalers or we do whatever.

Josh Wolfe (Lux): So to me that question, because when you said it it's inevitable, directional arrow of progress, every one of us is gonna have many, many bots doing many, many things from the trivial to the substantive. Which bots? Because then it comes down to one key factor, which is trust. And we're screwing around with lots of different ones. And at some point, I do believe that the distinction between them will be who do you trust or entrust with your data? Who are you giving your APIs? Who are you giving your

sam lessin: Yeah.

Josh Wolfe (Lux): Inbox two, who are you giving your calendar, who are you giving your chat?

Scott: Clearly Grok.

sam lessin: Well and well and who and whose answers do you do you trust? Like is the upshot, right? It's like whose matrix multiplying are you willing to rely upon in what scenarios? But Peter, you worked at OpenAI, give us the scoop.

Peter: Yeah, mean, it's, yeah.

Peter: I mean, look, I think people's perception of AI is just going to change. We know it's continuing to change. the idea of trust is actually getting, people are just starting to trust it a little bit more. In the beginning, everyone was so like kind of skeptical of like, it's going to steal all our jobs. Yes, it's taking some jobs. But the fear is going to be there until people really experience it. And I said this when I was at OpenAI, one of the best things that OpenAI did for AI was just make it just free and usable by anyone to be like, yeah, you could try it. You know? Yeah.

sam lessin: Well, Peter, can I push you on that? That is not that is the maybe your and my narrative, but the general, like the the only thing that pulls worse in consumer polling in terms of trust than like it than I'm sorry, the the I sorry, I totally screwed that up so someone can edit it. But like in r there was a study I saw recently that said that AI is so hated, it's actually hated more than Jeffrey Epstein by the general public. Right? It's like it is absolute, it is the worst brand like you could possibly imagine. No one trusts it.

Peter: I'm not disputing that. What I'm saying is that that barometer is going to shift over time. I was just on a call with a bunch of creatives, and maybe Scott, I'd love to hear your thoughts on this because you're in the thick of it, who, you know, they're writers who just started saying like how awesome it's been for them to build portfolios for their friends with Claude Co that they were unable to do before. That's like not a conversation I would have had with them six months ago. So I'm not saying where the trust is right now, Sam. I'm saying that that shifts as people get more comfortable with it and it's just a matter of time.

Josh Wolfe (Lux): Do you do you trust all models do you trust all models equally right now? Or are there some brands where you're like, I'm not giving it my data?

sam lessin: I don't know man, but

Peter: No, I actually have an irrational distinction between what I use Cloud for and what I use ChatGPD for. And that's, yeah.

sam lessin: Okay, so let me push so so let me push you then.

Rachel: But is that on trust or on performance, Peter? Because I think you're basing it on something different than a lot of other people.

Peter: It's on just like, it's actually based on what has my history, to be honest. It's like, yeah, like it has all my work contexts on cloud. So I'm just gonna, it's easier. It's like back in the day when we got iPhones, we had too many apps on our app screen. We started just choosing icons to like, to just compartmentalize things, right?

sam lessin: Okay, so a few can I can I push then for a second, which is like look, you gotta say in terms of where this actually plays out, one, this idea that AI companies be able to lock in your data makes no sense to me, right? If that makes sense. Like the the model companies. It's i it makes no sense. Like you you're you're gonna like self-sovereign your identity in some way, shape, or form. I do it by owning all my own infrastructure and then I just point at different models for different types of tasks. It's like

Josh Wolfe (Lux): Great. Agreed.

sam lessin: etc. But this idea that like with a consumer app you're gonna like lock people in because of their data, I just don't believe it. It's like it's one of the only things the Europeans did well with GDPR. GDPR is a piece of shit. But it

Josh Wolfe (Lux): Well well, wait, wait, wait, wait, but but but some significant portion of the average population, which again today is probably just on Chat GPT, is not anywhere near sophisticated, sort of like what Scott said about you. You are special, you are unique, and you are an advanced user. And the vast majority of people do get habituated and will just use what they've always been used.

sam lessin: Sure, and that's why lots of people still use Yahoo dot com for search.

Peter: Yeah, so it's not about I never said that there's lock in Sam. I didn't say there's lock it. just said that there is like there's irrational preference and then there's there.

scott: Can we can we take a step back though and ask like what's the ideal customer experience that would be most likely to be successful? That is like we're from a tech and see you know trust lens, it's one thing, but from a consumer, they like get the new version of their their iOS software and they get a magic text message that's like, hey, I'm the Apple agent. I've got all access to everything on your phone. Don't worry, it will never be in the cloud, it will never be exploited by a third party. What do you want me to do for you now?

Scott: Done.

scott: And you just like go through and give it connectors and access as you need certain things. It saves your logins as you need it. And then like, why wouldn't they just win this?

sam lessin: I don't know. I don't know. You know, well, I think Apple might win it, but it I think the interesting thing is I don't think you want to be the one who takes on that liability. Like, if anything, I think the goal is to like radiate out risk to other people and be like, Well, we're not like to keep yourself safe and be the one that's aligned. Because like the reality is you're like, Peter, you're saying that you think the ban is gonna go up. The reality is the hacks are only gonna continue to grow and scare people more and more and more, right? Not less and less and less, right? And everyone's gonna have them. And so I just wonder if The the real thing is gonna be how do you if you really wanna win in like the new economy, in some ways like you have in this AI thing, you have to somehow figure out how to hoard the data and the trust and build the trust, but then when things go badly, always have a third party to point at to blame.

Scott: I don't know. I'm I'm I'm with Scott on this. He's got the name advantage, number one, but number two, I mean, Gemini just hit a a billion users. it's pretty obvious. It's friction. Like you get rid of friction and most consumers are gonna gravitate in that direction. People are lazy.

Peter: I would question that metric. think they're probably counting the people who are just typing the search box and maybe clicking expand. There's many ways to game the metric. I actually think that distribution is a good part of it, but I do think consumers taste and their bar goes up in terms of what they want to use. They have to really like using. I mean, there's many examples of this in the past of why certain apps win in the app store, why Instagram took off and a bunch of the other ones didn't.

Scott: If

Peter: I think consumers have a discerning eye, but it's...

sam lessin: Well c but consumer but could we sorry, just to be consumers are also really cheap, right? Like consumers want free, right? Like and they're lazy and they're lock in. I mean it's it's kinda one of things like is it really better experience or is it like cheap, accessible, free, and habituated for consumers?

Peter: Yeah. Yeah.

Scott: And lazy.

Peter: Okay, so let me ask you this, like series available, when's the last time you trusted with a question you asked? It's crap, right?

sam lessin: Well that thing doesn't w that's not a thing. That's not even the same class, yeah.

Scott: Well series crap. Series total crap. That's why they hired Google to build a real AI agent, which will be here yesterday.

sam lessin: I don't know.

Josh Wolfe (Lux): Yeah, well I'm gonna say that I'll

Rachel: I think there's a question, it's like a go for it, Josh.

Josh Wolfe (Lux): I was just gonna I want to go back to Scott's thing because it is a sort of a profound declaration that Apple wins. And you know, and and how you define that, but they have not spent the CapEx that everybody else has, which is an advantage. They have waited. You could argue that what they did in search, which is effectively, you know, get paid to let Google you know be on search bar and Safari and the ecosystem, they could do same thing i in AI now. There are two facets of my phone that are probably the most used that today. and Sam, you can enlighten me if there's something that you found for MCP plugins or others that across your agents are working that are the highest frequency utility and to me the look the least surface area exposed to AI of productivity, which are my WhatsApp chats and my text messages. It can ingest my Gmail, it can ingest my calendar, it can do search, it can query all of our internal docs, it can use my Google Drive.

Scott: Location location's pretty key too. There no one's exploiting it the way it will be exploited, but location's key.

sam lessin: I I got my location feed. It's sick.

Josh Wolfe (Lux): So So so add those three things though, because each of those are today sandboxed Meta with WhatsApp, Apple with messages, although Cloud can sort of access if you're using it through your your laptop. And then you can use which a lot of people don't trust, but I forget the name of the service. It's a YC company right now that I was an early adopter to. something pipe. screen pipe, screen pipe, which which basically does like constant DVR recording of your screen and if your WhatsApp

Scott: It's insane.

Peter: Click blue message.

scott: the technique.

Josh Wolfe (Lux): is open on your laptop, which is probably ten percent of my utility, ninety percent on my phone, then it'll record your WhatsApp messages. But but those two things, to me, to Scott's point, feel like Apple could win on those.

Scott: Well, so add two more payments, so Apple Pay. So they're seeing where you're spending your money. And then the second thing is location. I don't want to underestimate that. I I created an app called Dippity, which I wanted to be like, you know, stat maps for old people. So you you you can't see where people are, but you know if they're near you or you know if they're in town. So I built this thing and it was it's pretty good. But as I dug in, Apple and Google or Android have both locks.

sam lessin: If they force them open.

Scott: Wi-Fi location. You cannot access it as a developer. And that's why find my is so freaking accurate. But me, the independent developer, I can't get near it. I'm looking at pinging towers. And and it's just stuff like that that I think Apple keeps a lock on. I don't want them to win. Let's be clear.

sam lessin: So look, I I look to be clear, I think the the funny from my there there are certain things that Elon has said that I actually really grudgingly completely agree with. And like one thing I think he said several times, I I'm not sure if it's his or someone else's, but I think it's true, is that the most hilarious outcome is usually what's gonna happen. And Apple winning is the most hilarious outcome, right? So I think that's like a very good chip in the in the in the factor of like Apple wins by doing nothing. And actually It's not even because they didn't, they tried to do nothing. From everything I've heard, Apple internally is extremely confused organizationally on AI. There's a lot of reasons they're having trouble moving forward on it. So in some ways, like organizational dysfunction creates the buffer for them to be highly successful by doing nothing, which is hilarious, right? Like and well, until it's just like you just stick it in. So like that's the Apple thing. Here's the thing I want to ask about is like, are we all just arguing about what has APIs and what doesn't?

scott: Until it's clear what to do, until it's just like clear.

sam lessin: Like ironically, the reason that iMessage and WhatsApp are are have some defensibility or lock into them is they don't have APIs. Like forget MCP. If you don't have an AP it's not about MCP is stupid. Like you can just use API. I don't really understand why MCP is a big deal when everything has an API and you can just rip through it with with Fable, right? So like there's just there are things you can access programmatically which have no barriers and therefore go into the hive mind. And then there's this interesting exception to it, which is the mech is like Apple as a platform order, which can enforce API rules, right? And they can have effectively APIs to things that no one else can have APIs to, right? So is the real story just like it turns out that openness and APIs was a is bad, right? And that what you really want is let's close the stack with valuable data you can have uniquely.

Scott: LinkedIn.

Rachel: I mean, this is literally what's going on right now in like the industrial and the physical AI world, right? Like this data is not readily accessible. You can't commoditize it nearly as quickly. And so it gives, you know, it it gives startups like a chance to actually gain distribution in like a more methodical way before the, you know, the mega platforms can swallow it up. Like this is exactly what we're seeing play out. And I think like digital and physical are two totally separate things. It all has to do with the availability of the data for the mega platforms. Like it is it is really closed. In some cases, it's not even like accessible by anyone. It's like on a PDF somewhere, right? It's like literally the opposite. And

scott: Will that continue more so or do you think that will no.

Rachel: Well, I think you it it it creates choice around who gets the data, right? And so like it gives you know, in the continuum and the like the race between that exists between startups and mega platforms, like in some ways you think it actually gives startups a chance, right? Because they can gain distribu if they can gain distribution faster than like a mega platform can innovate, you have or a hyperscaler could commoditize it, like Those factors actually favor a startup, maybe more so than than what you're seeing in the digital.

Peter: It definitely evens the playing field. There's just a lot of data that's just not even digitized, it's not even in a PDF somewhere. mean, one of my favorite thing, yeah, exactly. It's in someone's head. A good example is like, we have a machinist population that's retiring and a lot of the knowledge is actually in their head. And I think that companies that can go and figure out what does it sound like when a tool bit is gonna break and what is the path that you take.

Rachel: The i the the factories weren't connected.

sam lessin: And that becomes their defensibility.

Rachel: Yeah. It does.

Peter: I mean, there's a bits of data that I think, in some ways, Sam, I think you and I saw this at Facebook. It's like, what did Facebook do? It just digitized the information that was in our head. Who's your friend? What did you do last week? Yeah.

sam lessin: Yeah. Well that we we yeah, a hundred percent. I the I used to we used to talk about I like you talk about the composer in Facebook era, it was a data pump, right? It was an efficient, free data pump to bring data online that couldn't exist anywhere else. And the reality is, if you go way back in the Facebook history, the real story was on day zero, internet's scary, so no one wants to post on it. You're never gonna put a real photo on it, da da da. Facebook created the trust and security for people to take a bunch of real world data.

Peter: Yeah, exactly.

sam lessin: Photos, relationships, messages, et cetera, and bring it online, right, for the first time. And and created an enormous amount of value doing that, right? And so now we're in this weird multiverse where it's like the internet has all been scraped to hell, right? Like good luck, New York Times with their lawsuit, right? But like that has been like copied a thousand times and distilled a thousand times. You have personal private digital information, what's in Gmail, what's in da-da-da-da. Where people have some say in theory, as opposed to it just all being like dumpable effectively. and so that's where we were talking about the trust, things like that. There's then like the, you know, the whole class of companies that are out there trying to, in some ways, build new data sets, the handshakes, the scale AIs, where you just pay a ton of money and you turn financial capital in some form into like unique data. And then there's all the offline stuff. And so I guess the question for me is like assuming that multiplying big numbers is a commodity and will be. Right. Like is the story you just like really you go back to day one, which is like if your data isn't unique, you're not gonna be valuable long term.

Peter: I think there's also a class of data that you didn't talk about, which is something that Periodic Labs is doing, which is like science data and material science data that doesn't even exist yet. You have to actually do what lab experiment to get that data. So absolutely, like finding those bits of data has always been, I think the name of the game, it's going all the way back to Facebook. And you and I used to talk about this in those small little rooms about how we're just trying to extract a bunch of data and how obsessed you were about buying the DMV. Yeah.

Josh Wolfe (Lux): Why I

sam lessin: Not just trying obsessed. I did no, that is true. Peter, we used to have meetings for what we should buy, and my pitch was always that the Facebook should buy the D V.

Peter: the DMV.

scott: Ha ha ha.

sam lessin: Okay, can I ask, can I pivot the conversation a little bit? Unless people want to go further on this, is there's an there's another attack I'm curious people's take on, which is open router. Like talk a little bit about enterprise and things like that. If we're saying that your data is so valuable, right, and but at the same time enterprise really cares. my goodness, a guest visitor, Jessica Lesson?

scott: Hell yeah.

Scott: Wow.

Jessica Lessin: How's going guys? How's Sam doing? I I am so jealous. Continue on. I'm just here. I'm just, you know, in a taxi listening.

sam lessin: we're so deep in it.

Josh Wolfe (Lux): J Jess, just to get you up to speed, we've got two Sam's and we've got two Scott's, but there's uppercase and lowercase Scott. So that's the layout.

Jessica Lessin: Guys, I I am so happy to see all of you. I just came back from the big city of Manhattan, so I'm I'm on route. But dive in, dive in. I'll just I'll add a little bit here and there.

sam lessin: Yeah.

scott: Yeah, you do.

sam lessin: Are you wait, are you in New I wanna w for dinner plan, are you in Massachusetts or in New York?

Jessica Lessin: I'm in Massachusetts. I'm in the back. I was in midtown at three thirty. It's it's really a miracle I'm here.

sam lessin: you are. great. I'll see you soon.

scott: Sam, I would have thought your agent was better than this. your agent is not able to like

Josh Wolfe (Lux): Yeah.

sam lessin: I I don't have I don't have access to Jessica's I don't listen, la you wanna I knows every footfall I've done in the last twenty years, but I don't know where my wife is. That's kinda the way it works.

Scott: That's what you get for three thousand dollars a month.

scott: I know.

Rachel: He she won't share her by my friends and this is the core problem, right? So

Jessica Lessin: You know. Ha det allora. I love it. I love it.

scott: Yeah, yeah.

sam lessin: She sh

Scott: Sounds like she's got a better agent than yours.

scott: Lockdown data. Lockdown data. So routing. Let's talk routing.

sam lessin: He can Alright, so I wanna go back to open router. Open router. Open router. Routing. Everyone wants cheaper stuff. AI the the mate multiplication commodity. There's no reason to pay a PhD to d to fold your laundry. But you also don't want to give open router all your data. Or do you? Or do you not care? Or like how does that play out?

Josh Wolfe (Lux): Sam tier tier.

Josh Wolfe (Lux): I I do think that the bigger question here, which is gonna also buy for Kate, it's not gonna be one takes all, is open verse closed. And I am saying this obviously with a bias, as I've said in the past, that where you stand on the issue depends on where you sit in the cap table and where large investors and hugging face, which made a lot of news over the past few weeks. But I think that the future is for the more sophisticated users and serve for the enterprise, the It's gonna be your longitudinal proprietary repository silo of data that is yours. That is exactly as we were just talking about, not ingested by the machines. And if you have that and can run open source models on that time series of data that is exclusively yours, whether you're a pharma company, a finance company, insurance company, a retail company, a venture firm, increasingly you are realizing that to give to get the benefit, you're giving a benefit. And you're gonna say, I'm not gonna give that data, especially if I'm not gonna give that data and watch. As one of the closed labs competes me away. And you saw that first with Figma, which was sort of scary. And I think the push now from Anthropic and OpenAI to go into pharma and pharma companies saying, whoa, whoa, do we want to do this? Do we want to give them our data and then potentially open up competitors? and I think there's a big move right now for them to try to ingest as much information from as many non-internet sources that they can. And I think that there's gonna be a movement underway. whether it is industrial, manufacturing, defense, pharma, biotech, and all of our personal data that we're gonna say, you know what? The models are good enough, they are performative enough, and I'm gonna retain the sovereignty of that in the same way that Sam does and and I'm not giving it over. So I do think that that's where the value's gonna accrue.

sam lessin: So where are we in terms of people waking up versus like the you know the whatever the s the the door closing on the like are we are are do you think that these guys are gonna get away with getting enough data before people wake up that they're gonna be competitive, or do you think that the the game's too obvious?

Josh Wolfe (Lux): I I I think that this this past month, since the open AI hugging face moment, which I think, you know, there's probably three or four factors. That's one, companies waking up to the token economics and realizing like how much are people spending and like what's the productivity gains we're getting. the rise of the Chinese models and then a combination of like fear mongering and regulatory capture or attempt thereof against the contrast of people that have been in the White House and been the sort of AI SARS that are warning about this and saying, No, no, no, if you use these cutting it to Chinese models, but you use it on sovereign hardware and instantiations that are effectively air gapped, you have nothing to worry about. So where are we? I I think that still going back to the first point about habituation, the vast majority of people are just going to continue to use closed labs and trust it more and they're gonna try to ingest more and more and more. and you have whole teams, particularly at Anthropic but increasingly at OpenAI, that are trying to make big pushes into every enterprise. I mean, Scott, I'm not sure if you can reveal like at A twenty four, but like a lot of proprietary data and information on everything from like every pitch that you've never you know, greenlit to the ones that you have that never got made to you know, that's in that's a trove of valuable information that I'm not sure

scott: anyone wants like share their company data with a company that but I I do I I do know a number of startups who are kind of using Claude individual accounts and you know using their ramp card type stuff and then now have to for whatever reason because of scale go to the enterprise plan and they realize that it's like 10x more expensive and then they're realizing that they're using a lot of tokens for things that are pretty mundane and then someone on the engineering team is like I think we can do this with this like open model that we can, as Josh said, there's no risk. It's like local. We can mess with the weights and do whatever we need. And so there are now, and then there the next question is okay, what router should we use? And it seems to me like there's a lot of router options out there, and it's going to be probably pretty commoditized pretty quickly. And and and so you're going to all have routers that direct all sorts of spend to look more local, more efficient models. Like this is going to become the default. My question is actually like what more opportunities do routers have to not become commoditized? Like I thought you were gonna go in that direction. Like routers can start to ha carry logic around which models you should use.

sam lessin: Well they they're all gonna do that. They're all gonna do that, right? Like and they're all already doing that. I think the question is I think was what Peter started with, which is like at some point, like you do have to expose something to them, or you just build your own, but like who do you trust, right?

Peter: I think a lot of companies are building their own. It's not rocket science to build it. Josh, I'm on the same side as you, a board member at Arena. We see every week the models change in who's the leader. Not only that, Arena has a great insight into exactly for each use case, accounting or whatever, what is the actual leaderboard for it. Quite frankly, you don't need to drive a Ferrari to do the task of a certain smaller job.

sam lessin: Amen.

Peter: So you're going to see a lot more of this, you know, I believe in a multimodal world and I believe there's going to be a lot of customization and tuning and yet companies like Apply Compute that's doing this, many others as well. So I think it's going to be a pretty tough road for something like Overrouter. I just, feel like either roll your own or you got to find another way to make a model your own.

sam lessin: This is a great bridge. Can I bridge this? Let's talk about the crazy shenanigans NVIDIA's pulling with financing these things. Because the the question you gotta ask yourself, right, is like, okay, multiplying big numbers, kind of a global commodity. You're not gonna have PhDs folding your laundry, right? You're gonna route smartly over time. Like n what percentage of things do you actually need an edge model for? Like it just it seems like you have this you have some like machines that can compute numbers, you have people who can use numbers being computed. And then a lot of stuff in the middle, right? And it seems like OpenAI, I'm sorry, NVIDIA with this new five hundred billion Wall Street crazy alliance with everyone is propping up putting a lot of computers in places. Like discuss.

Scott: Do you guys do you remember Lucent? I know I'm dating myself. Do you remember when? So Lucent, yeah, sorry, the old guy. And hey, for the record, Sam, I'm now in my 13th year independent venture capital. I was in Goldman for 13 years. So at least at least, well, no, no. And then I was at a startup in the dot bomb, trying to beat Google, look smart for four years. So I'm there you go. Why just look when you can look smart?

sam lessin: Give us some history lessons, Scott.

Josh Wolfe (Lux): Yes.

sam lessin: so you're fifty fifty.

Josh Wolfe (Lux): I used Fluxmart. That was that was a that was a great yeah. Yeah.

Scott: So okay, Lucent put yeah, that that explained the the bust right there. 99 captured. Look, Lucent put what eight billion, nine billion, I can't remember what the number was, into fiber. And that well, yeah, exactly. Your next series, No, they put like eight billion, which was a big number back in 2000, 2001, and and they put it on their balance sheet.

sam lessin: Did you come up with that was that your contribution, Scott?

sam lessin: So like a series A.

Peter: you

Scott: Right. So they basically did vendor financing. They said, okay, everybody's gonna need fiber. and they were right. What what they were wrong, which is exactly what you were just talking about with open router and some of the other developments, is the whole concept of wave division multiplexing. Nobody baked that in. And all of a sudden you get a hundred at a hundred X improvement on Fiverr. And so your your demand and your pricing fall through the floor. So Let's just assume for a second it doesn't fall through the floor and we've got a half a trillion dollar financing facility thanks to the six amigos that lined up to lend their balance sheet. None of this goes back to NVIDIA. This is why Jensen's brilliant. He basically set this up as SPVs that will basically borrow against the take or pay agreements of the off takers, of of the neo cloud guys that. are saying we're all gonna need three thousand dollars a month for our personal agents. And they are signing take or pay agreements. So independent of whether or not, Sam, you keep your three thousand dollar agent alive, they still have to pay for the demand. So for a couple years, the the

sam lessin: To be clear, the three thousand dollars is just for databases and app servers, the agents cost something else, but yes, point taken.

Scott: Sorry, I I don't want to cause any marital issues when you talk about shopping and budgets. so the the take or pay agreements gives some security. Loose Lucent, NVIDIA is basically saying vaguely we're good for 25% collateral, but the real onus falls on BlackRock, Blackstone, Goldman Asset Management, all the guys that lined up. And those are like pension funds, right? Those are those are assets they're managing for the public.

Peter: No.

sam lessin: So okay, so is this even worse, Scott? Are people gonna be like, not only is AI like using all the moder and have the wrong answers and all that stuff, but it also is gonna make most people poorer in the process of like NVIDIA stockholders being getting well off?

Scott: And so

Scott: I mean, it's still a drop in in the bucket for overall assets for these guys, but you have to assume a double collapse, right? You have to see price collapse and you have to see demand collapse, which is basically what we saw in the mortgage crisis, right? If you see both of those collapse, then you have a systemic issue. Otherwise, it's idiosyncratic and it just will compress margins if it's price, and then they have a

Josh Wolfe (Lux): Well th th this

Scott: The way NVIDIA set it up was they said, look, we're gonna make these kind of generic, and so anyone can use them. So if I have a default from one off taker, I can just move it to the other off taker. And so there's some good protections in there, but it's fascinating. And well, I didn't I don't necessarily think Apple means on pr on device or edge.

sam lessin: But not if Apple wins. Not if they're doing it on device.

sam lessin: I know, I know, I know, I know. I'm just I'm just messing with you.

Josh Wolfe (Lux): Sc Scott Scott Scott I think nailed this because the big difference whether lesson learned from Lucent, Lucent actually from their balance sheet lent the money to their customers so that the customers could buy Lucent. Nvidia is taking no balance sheet risk on this. They're basically just convening the lenders and and the lenders see that there's opportunity in the spread. But to your point, yes, the people that are being stuck with the risk, I'm pointing, but it's like my mother, your parents, it's the pensioners.

Scott: Brilliant.

Josh Wolfe (Lux): Because that's the capital that's in those large funds that are aggregated. And then you have like duration mismatch between like the hard assets and the financing. So like the chip, depending if you look at Meta or Amazon, you know, we can debate whether it's a three-year or five-year depreciation, but it's, you call it three. the debt amortizes over 10 years. The power constructs and contracts are like 10, 15, 20 years. And so by any measure, the critic would be right to say that the liability outlives the asset. And anytime you have a mismatch, forget about bubbles. That's the problem. Yeah.

Scott: That's a problem. Yeah. It's like

Rachel: Yeah, I mean they're assuming, right, these are gonna act like any asset that's gonna spit off cash for a extremely long duration. And like this isn't a toll road. You know, this isn't something that like inherently is gonna have value 30 years from now. And I think that's to Josh's point, like the fundamental risk and question is in that, you know, how much money can you spit off versus the duration that these are useful for?

Scott: Six years.

sam lessin: Well, although just to push on this, and this is a place I don't actually know the answer, but I'm sure someone on this call knows better than I do. It's like in terms of it being a rail like, you know, how long the duration is and the value, how much of this is financing power and shells and things like that versus like the racked GPUs and connectivity? Cause like I would argue that like the shell and the power and other things like that actually does have a longer duration. This is just pure chips.

Scott: No, this is chips. Yeah, and and look at the H one hundred, right? Like w the the H one hundred used to be eight bucks or something like that per hour, like two years ago, and now you can get it for two bucks. That's in like

sam lessin: Well it was one seventy, it's back up to two something or whatever. But yeah, like yeah, yeah, yeah.

Scott: That's at like two years. So I I've Rachel, yeah, Rachel and Josh nailed it right on the head. You've you've got a mismatch of duration. It's a duration mismatch. And so you have to assume you can stretch the longevity of the chips longer than what we've seen so far in the life cycle. So

Rachel: And they're correlated. There's risk that's correlated here too.

sam lessin: Right, which by the way, I do I yeah. And and the interesting thing, I wonder how much this is I mean, like, look, it turns out that in traditional data centers, at big hyperscalers, the chips, the computers have actually stayed valuable longer than expected, right? Is the upshot. Like people are still using, like whether it's spinning disks or whatever, that that's the the a people model it one way, and I think everyone's over time been impressed that actually There's a more of a tail on it than people realize in a lot of ways. How much of the story of like actually this is better than we thought is being carried over into multiplying big numbers in this new era? And is that part of the story that like Blackstone and Apollo is swallowing?

scott: Hmm.

sam lessin: Or is that above our pay grade, which it might be 'cause I'm talking in half truths.

Scott: I mean they're swallowing something,

sam lessin: Or is it just they have so much money and nowhere to deploy it and they need asymmetry?

Josh Wolfe (Lux): Yeah.

Scott: I think it's a macro bet they're making and and that's it. I I think, you know, we're probably overthinking it. Fitch is doing the rating work right now and and they have an open consultation, which basically means they're trying to figure out which end is up because no one's ever securitized compute. it is sorry, it is fascinating to think about securitizing compute. That's basically what's going on. And and Jensen's a Genius, right? He pitched Wall Street on this. Wall Street did not go to Jensen and say, hey, we have an idea. He went to Wall Street. So the guy's genius. And to securitize compute is pretty cool. because no one's no one's done that. I mean, you know, we've done we've securitized other things. However, I have to do a shout out to my last piece about Americans getting equity in the AI thing. If it's securitized, it makes it even easier for a dividend for all Americans. But

sam lessin: It's super cool.

Scott: That's for those of you who read my newsletter, all one of you on this call, Sam.

scott: Ha ha

sam lessin: But I my bot's opening it. I'm kidding.

Josh Wolfe (Lux): American American participation in all this I think is a great virtue. I think the the Trump accounts, I think what Michael Dell has done, I think what what Brad that Altimer I think is awesome because you have generate you have a generation of young people that either are apathetic or feel left out, are driven to socialism and communism or graduating with a higher unemployment rate than like the average American, like all of that. Like I just I think it's good. It'll take time and like the political entrepreneurs that are trying to co op these people.

sam lessin: Amen.

Josh Wolfe (Lux): You know, it's like a race, but but to Scott's point on like the collateral and this like compute not having been collateralized before, we don't know the depreciation. And maybe that new models, new algorithms are able to take older compute and find great utility out of it. The GPU boom that benefited first from crypto mining that then sort of fell off and then regained again from AI, you know, was sort of saved. but these words are dangerous words, right? Just like this time is different are dangerous words, collateral debt. collateralized debt obligations. Like that was, you know, for let's say 20 years ago roughly. So for the average, you know, 27 year old today, they were seven years old when this shit was happening, in housing, which was an asset class that we assumed could only go up. And then it was arguably, you know, if you read like the academic literature of like John Jinkoppoulos, it was a collateral problem. Like it wasn't just a bubble, it wasn't dead. It was just like the proverbial merchant of Venice pound of flesh. It was the collateral And the collateral became worth less and the debt became worthless. and so that that's the risk.

scott: And that is that is the analogy there, like that that that moment from the film where this woman in the yeah, in hot in Florida and she's like, I've got seven houses, and you know, is that is Jensen her?

sam lessin: Margot Robbie in the hot tub?

Scott: Ha ha ha.

Josh Wolfe (Lux): No. because he's not as good looking, but

sam lessin: Yeah, the jacket doesn't do it for him. What you know, I always say, Rachel, I'm not sure if you agree this, but I always make the point that I actually think

Josh Wolfe (Lux): that's so anticlimactic.

Rachel: I know, wait. I I missed that. Yeah, I have no idea 'cause you Yeah, exactly. If it's Sam, it's always false.

Scott: Yeah, Rachel, what do you think about that? Yeah, true or false? True or false, Rachel.

Peter: Ha

sam lessin: No okay. It's been wildly successful. There you go, you gotta start somewhere. It's been wildly successful. I actually credit it to the housing crisis. Because basically, if you look at the like seven or eight people we started being in New York with right at being New York with in two thousand five, we were all basically coming up to to private equity jobs exactly as the economy was falling apart. And so everyone who was gonna go into private equity, because that's what you did. basically got kicked out of private equity because there were no jobs, right? And everyone started companies. And a huge percentage of them have been massively successful. And so, you know, it it's it's very v you know, that as much as we were young, it's very much in my mind what happened there. And I think some of the weird unintended consequences of it. But

Rachel: Why we have like the best three year being New York class ever.

sam lessin: It's what I I I frequently say and I think it actually is in s demonstrably true. No, no, I'd also say like I'm like the least successful person from that and I've done okay. like so it's been fine.

Rachel: The best network.

Scott: Hey Sam, I wrote my my social studies thesis on affordable mortgage lending in nineteen ninety three. So I actually am the one to blame for the financial the housing crisis, in case you're looking. Yeah.

Peter: you

sam lessin: I knew it was your fault. I knew it. I knew it.

scott: By the way, what what happened to the the the whole report of these big AI labs like giving equity to the government? Is that that just like disappeared? Is that not happening?

sam lessin: That was the most cynical bullshit I've ever heard from my perspective, even though I do actually like Trump accounts and the idea of the the the index. But yeah, what did happen to it other than my personal take?

Josh Wolfe (Lux): Well, I think that there was

scott: I don't know if like the AI lab equity is the right thing to distribute because for obvious reasons, but I do think that there's some like the collateralized compute actually might be a better or some sort of like instrument that exposes you to the cost of you know the the the the rising appreciation of energy in compute and I don't know like how do you get exposure and distribute it like Norway did to like oil proceeds to like address what Josh was saying.

sam lessin: I know you're important.

Scott: This was the op ed I just did with the information and then the little interview that Sam Sam, what'd you accuse me of? Being mercantilist or something?

sam lessin: Did he manage did he like your You're a mercantilist. You're a mercantilist. You're not a capitalist.

Scott: More cantalist. and so then I came out with a a newsletter post of like the reaction to my idea was way more interesting than the idea itself of all the of all the ideological battles. And but the title of the post was I don't have property in New Zealand. Cause that's where all these conversations end. Everyone throws their ideology around and they go, What's your escape plan? And I don't have an escape plan. So I'm I'm with you, Scott.

sam lessin: That's usually how works.

Scott: Like there's some merit there's merit here. We have to th rethink the social security system. We have to rethink

sam lessin: So can I I I I have I wanna I let me a let me push on this for a second, then I want to talk I really do want to talk about one or two other topics before we wrap on the day. I hear you, and I totally think it would be great to have Americans more dealt in in a lot of ways. I think in obviously I believe deeply in capitalism and would love people to be more pro-capitalist and feel part of the story. Here's the problem. Historically appeasement never works, right? It just doesn't. Right. Like the it's like whenever you appease, you're like, okay, and and the reality of capitalism, right, is that the story is supposed to be, hey, if you work hard and get lucky, you do really, really well. It's not a social security net story, right? And like is the American story, I would say. Now, we had a lot of benefits. First, we had a country which unfortunately all the people who lived here originally died from plagues and then wars. We had a huge amount of space, we had n natural resources, we had Western expansion, we had the internet, we had all this terra nova, which meant that Historically, if you're an IQ one hundred person and worked hard, you'd do fine, right? If you had a shovel, then it was like, just go to school and you'll do fine because there was so much demand for like intellectual jobs. Those stories have run out on the capitalism side and the American side. It's not clear what the next story is, but the story of, hey, here's a thousand bucks, hopefully it compounds in the market, I don't think solves the underlying problem of like how do you create opportunity for people and like the American story of capitalism free enterprise.

Scott: Two two two distinct challenges. One is a existential meaning, meaningness, meaningfulness question. And then the other one is how we fund this? How do we put food on the table and police on the streets and you know, kids in school and all of that? So they're they're linked, but they're very different. And I agree with you, we have not answered the post labor existential question. We just haven't. We don't need more poets, I don't think. I mean, I love

sam lessin: that's been my essay. My essay isn't

scott: I I think this is a question of, you know, parents can can can have money saved for their children and screw them up, or they can have money saved for their children and not screw them up. So it's you know, I've seen both instances happen. And I don't think that there's a rule against one or the other. It's more about how it's done. So that's why I thought it's

Scott: That's a great point.

sam lessin: Well that's so it's a culture. So it's like a culture and values and all that stuff is like what matters.

scott: I wonder what the instrument should be. I mean, what should the instrument be of these savings accounts, which I think I'm you know, I'm also a fan of that are that are being created for kids that are, you know, now being born? Like what why shouldn't they get some exposure to the well? Some of them will become like the deadbeat kid that got, you know, a silver spoon and screwed it up. But I feel like a lot of them will also not graduate with college debt or debt for whatever education ends up looking like in the future. And Have a little startup capital to take a risk in their lives and all the other stuff. So we're not going to solve it for everyone, but isn't that net net better?

sam lessin: I'm I'm very pro the truck account. I think it's good and smart. I'm pro it. I I'm not convinced that the picture of how it compounds is so certain, especially as we're talking about resets and all this other stuff that you know we're trying to figure out right now. But like I'm certainly pro it. I just except with the exception of when people start lobbying for regulatory capture by giving a five hundred bucks of their stock to someone, right? Like that seems insane to me. And we gotta figure out how to like not have that happen. Okay, I wanna pivot. I wanna pivot one or two more times before I lose you guys, and I really appreciate you you joining for this fun conversation. OpenAI, exec Exodus. Brad Lightcap, out, Kevin Wheel, who we used to work with, out. Peter, you've been out for like years, so like you were just you're a forebearer on this. Like w why and like what what's going on?

Peter: Yeah. Yeah.

Peter: My take on this is it's actually a feature, not a bug. It's a feature of having insanely great talent. think Sam is N of one at recruiting really ambitious, talented people. And in a time when you can build anything, you should expect the turn. These are people who want to go and kind of build something really, really insanely great. And that's not our talent problem. That's just kind of a natural output of what he's optimized for. mean, you're seeing this. And also the other thing is that it's not just having the models. You have to point the models with a sense of obsession at a problem. you know, like something like Periodic Labs, Liam left OpenAI. We were the first Czech investors into that. They're doing material science. It requires a different level of obsession on building a wet lab that just doesn't exist at a bigger company. Right. So you have more of the companies out there like Applied Compute, Core Automation, other...

Scott: Okay.

Peter: other NeoLabs. There's a point of view that you want to go chase, and if you're ambitious, you're going to go out and chase that. So I think that's largely what I see from this. You have to point the models, and it's a great time to build, and so that's what everyone is really doing.

scott: Peter, let me you a question. Is it because of their there's like this sense that these generalized models are pat are are like the old news and specialized models are the future and it's like gonna be specialized models plus all these open local?

Peter: I think it's great question. I think there's a lot of stuff and Rachel, you and I live this in Uber. The last mile problem is a real problem. You can't just drop GPS on phones and be like, okay, all right, let's give rides to people, right? You know, one of my analogies I use is like, you can't vibe code a home health nurse to show up in your living room to take care of you post-surgery. There's a company I'm working with called Adaptive based in New York. They're doing exactly this. They're like an AI sort of healthcare provider delivering healthcare and that delivery Rachel, you and I know this from Uber days is not easy. There's a lot of stuff that goes on to make sure that that value can be delivered. And that requires an obsession that goes beyond, AGI is going to solve everything. So I think that's what it is. It's like when you're obsessed over a problem like that, you're going to go make a company and make it.

Josh Wolfe (Lux): I I've got a quick cynical take, which is Peter, how long were you open AI?

Peter: a year and a half or so.

Josh Wolfe (Lux): Okay. and Kevin I think was there maybe four years, maybe a little bit less, three years.

Peter: I think he was there for a lot less than that.

sam lessin: Less, less.

Josh Wolfe (Lux): Two, two and but my my my cynical take was gonna be you go, you're there at a valuation, you get stock, you vest, the stock appreciates, maybe appreciates very rapidly, and then you're looking at the incremental upside from where you are and what you're doing in a ever growing thing and having to deal with more people and more politics and more knife fights and all that kind of stuff, or you can take the cachet of having been where you've been and you have a robust market that will finance you. Kevin is gonna go, he's gonna start something, he's gonna get funded. we're probably all gonna be involved, you know, with whatever he funds. And and and I I think that that's the simple calculus is you believe in yourself, you your financing risk is pretty low, your talent risk is pretty low, and your upside is gonna be way higher. And so the incremental return on the incremental expenditure of your time is worth leaving. So I don't think it says anything and I've been critical of every one of the labs at some point for whatever, but I don't think it says anything about open AI.

sam lessin: Josh, but d can you build a company that way? Like with people coming in, stamping the car, getting cheap ca and getting infinite free capital for their next startup. Like does that work?

Josh Wolfe (Lux): I mean, so far with open AI, valuation has gone from like, you know, sub hundred million to trillion and plus. And you've had six diaspora companies that have all achieved, you know, tens of billions of dollars that have spun off of it. And I think everybody looks at that. And by the way, the number one motivator that I've always found is somebody looks and sees a peer that they worked with and being like, Are you fucking kidding me? That person just raised how much at what? And then that becomes the catalyst, you know, moment where they're like, I'm going to do that. So

scott: By the way, that's a great point. That's a great point. That's the PCs call. The PCs keep giving money to certain people. I'm like, what is going on?

Peter: I mean, Josh, think.

sam lessin: Well, a hundred percent. And a percent and the talk talk is usually the the talk talk is usually and that guy's an idiot.

Peter: I think it's

Peter: Yeah. Look, think, I think all your point of view could be true. My point of view could be true. I don't think they're directly contradicting. I'm not saying that you think they are, but you're right. Like it absolutely, when you are able to take the risk to go build something bigger, that's kind of my point. And what I'm saying is that these people are ambitious people who are like, okay. and implied in that is like, okay, I can stick around and you know, the way I'll put it is the physics of a big company, know, Sam and I, we've had the big companies there's the physics are just different, right? And the people who are. wire to go, know, again, Sam is Sam, Sam Altman has assembled a set of really great researchers who have these big ideas. You're going to want the freedom to do that elsewhere too. Right. So again, it's just, it's just a natural consequence.

sam lessin: Well, except for so so okay, so the two directions about Burgess and jump ball. One direction is sure, I agree with that. But the early open AI picture was we're building AGI, winner take all. You're either on our train or st have fun staying poor. Like it was like a one thing. That is kind of evolved into a lot of people who are really smart being like my god, this is the moment where everything matters. Cost of capital is zero. I gotta take my shot. And by the way, I've already punched my card four times. If it does turn out to work, I've I'm protected. I've got plenty of stock. Right? Like

Josh Wolfe (Lux): How how how how different was it in the Facebook days? 'Cause you had sort of the same dynamic.

sam lessin: Well, you had gener I I think first of all, it's kind of funny because by historical stand by historical standards then it was a incredibly meteoric rise. Facebook's valuation grew unbelievably slowly compared to the labs, right? And there were obviously generations. Like I would argue I was Peter, you were kind of generation one and a half, maybe. I was generation two at Facebook. There's now been generations three and generation four, but there's not twenty generations, right? There was like a s a set of people who worked for four years, did great work, and then passed the torch.

Peter: very slowly.

sam lessin: And that's happened like a few times. And actually, the core leadership team at Facebook has been unbelievably stable now for like a decade, right? Like, which is a very different dynamic.

Peter: Very stable. Very stable. And Josh, the other thing to look at, just not to be too cynical, is that Facebook had great network effects. So if you wanted to build anything in social, that's where you stayed. I got to say that if you take a look at AI, again, being on the board of Rena, it just changes every week. It's like, well, who's up there this week at this point? And I would say AI is a lot more akin to something like GPS on a phone or a database where it is a technology.

Josh Wolfe (Lux): For sure, for sure, for sure.

Peter: But again, you can't just like vibe code an Uber. You have to deal with all the real world problems. can't just vibe code at home.

Josh Wolfe (Lux): way your your network effect your network effect is actually a profound point because I think that I think that Meta has actually done an extraordinary job well well documented in their acquisition strategy and I think part of the pitch that Mark has made at least in two of our companies is the scale that we can give you to deploy your product whether it's a brain machine interface like control labs or you know like is enormous you're gonna reach billions of users and I don't know that

sam lessin: Yes.

Josh Wolfe (Lux): OpenAI has done that. So that actually is a good point that I think the the

Peter: Yeah, I learned how to do acquisitions from my time at Facebook. That was, that's exactly the goal. Yeah.

sam lessin: So okay, so I I don't wanna keep hard with Peter. I wanna ask Peter one more and then I wanna pivot to a few of the people we could close. Peter, and I I have not he's a friend, but I haven't spoken to him in about this in months. Is the air table how we story I wanna work on AI? Or what happened?

Peter: Yeah. Yeah.

Peter: I haven't spoken to him in a few months either, but let me tell you a few things. You're talking about air table, the exit, bending spoons, et cetera. I think there are two truths here and both truths are true. Number one, I think founders are built differently. mean, you know, founders are going to continue to build towards AI like you, like you alluded to how he pivoted to hyper agent long ago. think Scott built Scott, know him well, right? Like he made that pivot and he's been pivoting. And the thing is that that acquisition, I think this is public knowledge. was had HyperAgent carved out. So that was really the legacy business that was acquired. So, you know, great founders who have founded all these companies like Howie are going to continue to build because that's what that's in their DNA and then builders going to build. The second thing is there's a real shift and both of these things are true. Right. And we're seeing that at Felices too. So in our last couple of funds, over 20 % of our capital has been deployed to global resilience, energy, manufacturing, defense tech, all stuff that, that you all are investing in as well. And science, mentioned we were the first check in periodic labs because we believe that there's actually more real world problems to be solved. So I think both are true in terms of what happened with Airtable and how he's just, from what I understand, he's just continuing to build. He's a fantastic builder.

sam lessin: Okay. I agree. I wanna go lightning round and call it a Because this has been awesome. And it's really this is fun by the way. Honey, we should do this again sometime. Rachel, I'll start with you. DC regulation, all the heavy stuff you're involved in, what are you thinking the most about right now that we have not discussed?

Scott: Yeah.

Rachel: I mien

Rachel: I don't know about a DC regulation piece, but like life, I'll just go there. we happen to live in DC. I mean look, I think no, I'm you know, I I do live here and sadly that's becoming no, I mean look, I think I mean Peter brought you know just brought it up at the end. We're spending, you know, the they're putting 20% plus of their fund into, you know, these areas that

sam lessin: Or just life, or just life.

sam lessin: I know, I just think of you as my like D C person now, you know? Like

Rachel: Josh, you've been investing it even longer than we have, but like we've been focused on and in 2020 when we started construct, right? Like a whole bunch now that was mostly LPs that we were talking about at the time, but people were like, You're crazy to go. The same people who said it, by the way, I was crazy in 2011 to join Uber because taxis are you're never you're never gonna get a venture outcome. One year, one year, yeah, exactly. Saladressing.

sam lessin: Weren't you at Clorox before that? What what you were working on charcoal? Yeah. I remember that.

Rachel: But no, but but like you're crazy to go join a taxi company because it's you know it's slow moving, it's heavily regulated, it's high capex. And like those people just like didn't spend two brain cells thinking about what Uber was actually building. And I feel like it's the same thing that people were saying to us in 2020 when you know we went out and started construct. And it was, you know, there's actually first of all, there's a lot of software companies. Secondly, there's a lot of physical world companies that don't need to look like really slow moving high capex companies. And the only way, you know, the our whole thesis is, you know, the only way we're gonna get productivity is, you know, is is is putting software into these spaces. And like w we're, you know, if you looked at the signals even like four or five years ago, like the crisis was mounting. And while we You know, we appreciate the markups that come from everyone piling into these spaces. or also, you know, now the prices are going up like you know, to a crazy amount, obviously, on on some of these early stage stuff, because a lot of people who like would never give these spaces kind of a second look are now devoting half of their, I don't know, billion and a half dollar funds to them all over the place. So anyway, it's an interesting time. People actually come through DC. It's great for our large conference room, which now gets use. and and good for the existing funds and the portfolio companies.

sam lessin: Fair enough. Scott, we didn't really get into proof of craft, providence, watermarks. You've been thinking about this stuff forever. Anthropic just is now watermarking text as of today. Like there's a lot of stuff going on. I mean, I I assume you are still deeply on this train. Can you give us a give us a quick update on that?

scott: Yeah, I think it's The C2 C2PA, which is kind of like the op it's the it's the open protocol that kind of everyone signed on to for these content credentials to be added to assets as they're generated or edited, by the way, for that matter, has been widely spread. The problem is on the consumer experience side. These companies like Instagram and, you know, and products that are surfacing media. You know, it the the first for them it's binary. It's like if it has a C GPA in it that says any AI was used, they report it as being made with AI. And then consumers are like, What the hell? I only used AI to like remove a blemish from my face. Like, why are you labeling me as a slot maker? And and so there's we're in this like moment where there's a disconnect between the underlying efforts, which I think are net net good to like help people be able to discern the provenance if they want to. and the consumer interfaces that don't know how to surface this or merchandise this to the end user. So I feel like that's where we're at right now. I don't think it's going to go away, but I also think we're all going to be gradually inoculated. And I think honestly like probably fake fake stuff is really helpful for society right now because it inoculates us and makes us realize we can't trust what we see anymore.

sam lessin: Doll fake. Josh, Wolf, got anything you're excited about we haven't hit on or you're freaking out about?

Josh Wolfe (Lux): Freaking about the rise of socialism and communism in our country.

scott: Yeah.

sam lessin: Yeah, I've noticed on Twitter you and I have that in common.

Josh Wolfe (Lux): Yeah. I I've I've been like a center left Democrat my entire life and my party has just like run far left and I'm yeah, like

sam lessin: Well it's 'cause no one has good like the marketing sucks on the center left. There's no good marketing.

Josh Wolfe (Lux): It does, but but I grew up caring about the guy that lost the Ovarian lottery in the Rawsian sense and you know, needs a leg up and social welfare and but I believe in the American dream and

sam lessin: Yeah, but the veil of ignorance is kinda bullshit. Let's not go Rawls. Can you be a Rawls guy?

Josh Wolfe (Lux): Yeah, yeah. No, I I I I believe that that's the right model for designing a society. You know, you wake up tomorrow and you don't know if you're gonna be born in Brooklyn or Bangladesh, black or white, male or female, able bodied or handy catch.

sam lessin: Yeah yeah yeah I agree with that part, but I'm more of a nosic guy myself.

Josh Wolfe (Lux): you know, I I I I think that Europe is fucked. and there are bright lights of amazing entrepreneurs that are doing the very thing that Europe has historically not done, which is risk taking. The the floor for Europe is higher than the US, so people do not fall through that. But the ceiling is never been that high. And I think that there are companies that are breaking through that. But I think Europe is screwed mostly for demographics. in influx largely because of political positions of people that are unwilling to assimilate. it is destroying society. I believe that there are bad foreign actors, whether Russia, China, Iran, North Korea or whatever, that are helping to foment that. And I do believe that the Sahel in Africa and the Maghreb, which is, you know, Mali, Sudan, Niger, Burkina Faso, you are one terror event away projected into Europe, that that entire region becomes the West next Afghanistan, all of which culminates in my conclusion that Europe needs very strong defense. it is very much a tower of babel of uncoordinated systems, but there needs to be an enormous expenditure, which they have all been under pressure of Trump and Rubio, you know, being sort of chastised at Davos and Munich and security conference. But Europe's gotta spend. We're seeing some of our companies benefit from it. but but Europe is at real risk, not just from Russia, but from violent extremists that are infiltrating the continent. And if you care about the West, you gotta defend it.

sam lessin: So you have to we have go defend Europe again?

Josh Wolfe (Lux): Yes. Yes. I I I I flew over. We were in France for part of the summer and I flew over. It was actually really amazing. Just like you're you're in a plane, you're flying over the beaches of Normandy and you're looking at Ohio and you know and it's just like eighty years ago, it's like it's crazy, you know? Americans no, it's crazy. It's two generations, three gener like we stormed the beaches of Normandy's against the Nazi like it's insane, you know? And we're we're we're not fighting Nazis, but it's like

sam lessin: It's not that long.

Josh Wolfe (Lux): It's it's wild and it's important and I think a lot of people it's like too weak to be felt until it's too strong to be broken and a lot of people are really not appreciating how serious this stuff is.

sam lessin: Fair enough. Well on that uplifting note, anyone else have anything else that jump ball they want to throw out before we call it?

Josh Wolfe (Lux): We're all gonna die.

scott: First step is talking about

sam lessin: Well, but it that but we've been s but the problem the problem is talking about it is not yes, but we're all talking to each other about it.

Josh Wolfe (Lux): Ha ha ha.

Josh Wolfe (Lux): No, the second step is is funding. Is you you you you you fund these entrepreneurs that are taking it on, you're getting patriotic, not egoistic.

scott: Second step is doing something about it. Yeah.

sam lessin: I'm with you. I look I'm with you. I'm with you. I have my own takes on where the leverage and I can deploy on this stuff is. I do think I'm with you ideologically. I do think it's one of the t problems of a Twitter or fun podcast with your friends who are all venture capitalists in different places is we can all fervently agree with each other and feel really good about it. guys, this was guys, this was really fun. I appreciate you all joining. I hope you had fun doing it. this is great. So thank you all. I hope you all and really enjoy the rest of your August and I will see you around on the internet and in physical space.

scott: Amen.

Peter: Yeah.

Peter: Super fun.

Scott: Thank you for pulling us together.

Rachel: Thanks for having us, Sam.

scott: Awesome. Yeah.

Josh Wolfe (Lux): Make sure you're make sure your bot gets the recipe for Jess for dinner. She's gonna be there like

Peter: Thank you,

Rachel: All right.

sam lessin: I she she's here. She actually just walked in. If she's still around, maybe she's coming out are you coming out to end the show? Here she is. She's pure wedding over. A song and dance. There she is. There she is. And lock I texted you a question based on what you said that I didn't understand. So if you could get back to me, that would great. Well, yeah. Yeah, that's you got it she got a text from Jess. Okay. I love that. I'm gonna

Rachel: I was gonna say she checked in with us, wanted to make sure it was actually happening.

scott: Take a clear.

Josh Wolfe (Lux): Me

Josh Wolfe (Lux): What should I eat for dinner?

scott: All right, later everyone.

sam lessin: Yeah.

Josh Wolfe (Lux): Tater docked. No, I don't know.

sam lessin: I thought we had a chef, but there's also pizza. Bye everyone. The chef made pizza.

Peter: Bye.

Rachel: Bye.

Scott: Yeah.

Scott: It's a rapid evolution in an old paradigm. Like, this concept of websites is gonna disappear very quickly before we know it.

Guest: NVIDIA is taking no balance sheet risk on this. The people that are being stuck with the risk, I'm pointing, but it's, like, my mother, your parents, it's the pensioners. The liability outlives the asset.

Rachel Holt: This isn't a toll road. You know, how much money can you spit off versus the duration that these are useful for.

Scott: No one's ever securitized compute. We have not answered the post-labor existential question. We just haven't.

Peter: You can't vibe code a home health nurse to show up in your living room to take care of you post-surgery.

Guest: Your financing risk is pretty low, your talent risk is pretty low, and your upside is gonna be way higher. Are you fucking kidding me? That person just raised how much and what?

Peter: I would say AI is a lot more akin to something like, you know, GPS on a phone or a database where it is a technology.

Guest: Fake stuff is really helpful for society right now because it inoculates us and makes us realize we can't trust what we see anymore.

Guest: More or less. Is it no or is it yes? We'll debate the test that's best. When we get more or less. Dave and Britt, plus Sam and Jess. Put it all right to the test. More or less.

Sam Lessin: Hello, and welcome to a ridiculous episode of More or Less, where my co-hosts, Dave Morin, Brit Morin, and Jessica did the terrible, terrible judgment error of letting me host without them. They're all actually on vacation, and I am actually also on vacation but pretending like I'm not on vacation. So I'm thrilled to be here. We have assembled a world-class community here. I think this group represents every flavor and every area of venture capital investing known to man, as well as, like, every product experience you could imagine. So I'm very, very excited to have a great group here to run through. First, we'll start with, we have Peter Dang here. Um, Peter actually has a distinction. Peter's the one who brought me into Facebook in a lot of ways. He was running profiles right before I was, and so when I showed up on day one, he said, "Great, someone else can run this," and taught me everything I needed to know, and then was very happy to peace off into his next project, which ended up being Instagram and then Uber and then Airtable and then big role running product at OpenAI and now is a dirty venture capitalist like the rest of us. So Peter, welcome. Uh, it's good to have you.

Peter: Thank you.

Sam Lessin: Your GP at, uh, Felicity. Holt, Rachel Holt. The, w- I've actually known Holt longer than anyone on this call. I first met Rachel on our first day of Bain & Company orientation in New York in 2005 when she... We sat down at orientation about seven of us in Bain, New York. As soon as there was a break in our seven-person onboarding-

Rachel Holt: You cannot tell this story, Sam.

Sam Lessin: Why not? It's a great story.

Rachel Holt: I know exactly where you're going. I knew that this was gonna come up.

Sam Lessin: It's my favorite story. She basically... We have our first break at Bain & Company in a training. This girl who, like, is with six people she doesn't know, her new coworkers, takes the spider phone in the middle of the conference room, dials Comcast and starts yelling at them. Just to set the tone. I was like, "This is an amazing... I love this." So, but Holt from there has done amazing things. She ran Uber North America.

Rachel Holt: I mean, first of all, I got my cable attached on day one.

Sam Lessin: Did you get it that day? That was amazing. No one had their cable attached.

Rachel Holt: Yeah, August 1st, 2005. Time Warner was there.

Sam Lessin: Boom. Boom. So she took those operational chops to then run Uber North America, has started Construct Capital, represents the rare DC venture capitalist, but a breed that is growing, and get to do a lot of business with her now. So Rachel, great to have you on. Scott Belsky, I mean, look, Scott's done everything. I knew him originally when he was running Behance and founded Behance, which he sold. He's now a partner at A24 and, and running A24, A Labs, founded it, and is thinking more about kinda the future of creative than anyone I know. So I'm thrilled to have him here. Scott Stanford, I mean, you started out in an even dirtier job than venture capital, running Goldman, or a big piece of Goldman for a long time on the TMT side. You can correct me on the details, but you somehow did that forever, and then have had a, an illustrious venture career, now is the co-founding, uh, partner of, of Acme. And then Josh Wolfe, who... Does, do you even need an introduction at this point?

Peter: I need the longest intro... No.

Sam Lessin: Of course. So Josh, co-founder and managing partner at Lux, which has done, basically did the contrarian thing, which has now become consensus around defense and deep tech and hard tech and is kind of one of the men of the moment for things, calling things like Anduril, et cetera, early and backing up the truck when that wasn't cool, which is now very cool, which means hard to seed invest in anymore 'cause these numbers have gotten so big. So we have an amazing gang here, and I, I have my topic list, which obviously I had my bot write, of what's going on to run through about what's going on with NVIDIA, this crazy balance sheet stuff. S- Peter, we're gonna have to talk a little bit of Airtable-

Peter: Sure, yeah

Sam Lessin: ... um, given the fact you ran product there for years and I was an investor and things like that. But Scott, you started out, when we were BS'ing before hitting record, we wanna start with Grok Bot?

Guest: Well, I, I, I mean, I think that, uh, the, uh, agents have been a topic that we've all been thinking about and talking about for quite some time, but we're, like, just in this moment it seems where they're becoming, like, really useful on a consumer level. And I don't know about a lot of you, but I've been playing with some of these. They're always a little scary, these startup agents, because they ask you to, like, log in via Google and give it access to everything, and you get those, like, crazy Google warnings of this is not a authorized, secure blah, blah, blah, but you do it anyways. But I have to say, like, they're kinda blowing my mind. And seeing some of the demos that people have shared in the last 24 hours of their Grok Bot... And I think it's a combination of obviously the agent capabilities but also the memory and this notion of, like, having the bot have a computer instance of its own and even seeing their computer set up. It's truly this notion of someone with a screen, and then you get this, like, consumer click of, oh, anything you can do on a screen They can do on a screen is like a really fascinating turning point, it seems.

Scott: It, it's, it's a shift from help me do this to just do it.

Sam Lessin: I'd like to point out that, like, I'm so down the rabbit hole on this that none of this stuff impresses me. I set, have a persistent Claude instance running in DigitalOcean. I had it do its own analysis in math right now. I'm not talking about the AI. My bot is running about three grand a month of infrastructure, database, app servers, et cetera. Whenever I want something, it builds a new app for me. It, like, runs its own databases. It sets everything. Like, and I actually interact all with it through email. So I, like, basically start email threads, and I'll, like, ask it 50 things, and then just, it dispatches them, build an app, analyze my health, send me this cron job, whatever. So like, I've given my bot so much infrastructure that I'm not that impressed with it has a computer, but I'm kind of curious if any of you guys... I mean, Scott, you've built a, some crazy stuff with AI.

Guest: You're not normal, Sam. So it's like-

Scott: No

Guest: ... I think that yeah, I'm curious what others think. But it, it does feel like there's, like, a crossing the Rubicon moment for consumer accessibility, and just the ease of onboarding and giving it authority for certain websites and just, it's... Something seems to be clicking, to me.

Scott: But it's, it's a rapid evolution in an old paradigm. This concept of websites is gonna disappear very quickly before we know it, and this concept of, like, "Help me do a bunch of stuff" versus "Just do it. Predict I need it, get it done." And, uh, yeah, I love the Comcast example, Rachel, of, like, you having to call Comcast. Forget that. Like, there's no humans involved anymore. It, it's, I simply... I don't even need to tell it that my cable's out. It knows my cable's out. It will reach out to the agent at Comcast. It will sort it out.

Guest: Not yet, right?

Scott: No, no. But I mean, where were we last week, right? So w- we look at this with such a, a microscopic lens. This is gonna happen before we know it, and then we're all gonna be sitting around looking at each other saying, "What are we supposed to do?"

Sam Lessin: But Scott, Comcast doesn't give a fuck. Like, they're not gonna, like, solve that.

Scott: The, the reason why they're gonna give a fuck is when they get bombarded by a bunch of automated agents that are hitting them 24/7 saying, "Fix this," and they have to respond. And so then they rev up their agents, and then it's agent to agent, and we phish, or we ride on Boston Whalers, or we do whatever.

Guest: So to me, the question, 'cause when you said it's inevitable, directional arrow of progress, every one of us is gonna have many, many bots doing many, many things from the trivial to the substantive, which bots? Because then it comes down to one key factor, which is trust. And we're screwing around with lots of different ones, and at some point I do believe that the distinction between them will be who do you trust or entrust with your data? Who are you giving your APIs? Who are you giving your inbox to? Who are you giving your calendar? Who are you-

Scott: I mean, clearly Groq.

Sam Lessin: And whose answers do you trust? Like, is the upshot, right? It's like, whose matrix multiplying are you willing to rely upon in what scenarios? But Peter, you worked at OpenAI. Give us the scoop.

Peter: I think people's perception of AI is just gonna ch- it can, we know, we know it's continuing to change, and the, the idea of trust is actually getting... People are just starting to trust it a little bit more. Uh, in the beginning everyone was so, like, kind of skeptical. Like, "Oh, it's gonna steal all our jobs." Yes, it's taking some jobs, but the fear is gonna be there until people really experience it. And I said this when I was at OpenAI, one of the best things that OpenAI did for AI was just make it just free and usable by anyone to be like, "Oh yeah, you could try it."

Sam Lessin: Well, Peter, can I push you on that? That is the maybe you're in my narrative, but there was a study I saw recently that said that AI is so hated, it's actually hated more than Jeffrey Epstein by the general public, right? It's like, it is absolute... It is the worst brand, like you could possibly imagine. No one trusts it.

Peter: I'm not disputing that. What I'm saying is that that barometer's gonna shift over time. I was just on the, a call with a bunch of creatives, and maybe Scott, I'd love to hear your thoughts on this because you, you're, you're in the thick of it, who, you know, they're, they're writers who just started saying, like, how awesome it's been for them to build portfolios for their, you know, friends with Claude Code that they were unable to do before. That's, like, not a conversation I would have had with them six months ago. I'm not saying where the trust is right now, Sam. I'm saying that that shifts as people get more comfortable with it, and it's just a matter of time.

Guest: Do you trust all models equally right now? Or are there some brands where you're like, "I'm not giving it my data"?

Peter: No, I have an irrational distinction between what I use Claude for and what I use ChatGPT for.

Rachel Holt: But is that on trust or on performance, Peter? 'Cause I think you're basing it on something different than a lot of other people.

Peter: It's actually based on what has my history, to be honest. It's like, oh yeah, like it has all my work context on Claude, so I'm just gonna... It's easier. It's like back in the day when we got iPhones and we had too many apps on our app screen, we started just choosing icons to, like, to just compartmentalize things, right?

Sam Lessin: You gotta say, in terms of where this actually plays out, one, this idea that AI companies are gonna be able to lock in your data makes no sense to me. You're gonna, like, self-sovereign your identity in some way, shape, or form. I do it by owning all my own infrastructure, and then I just point at different models for different types of tasks. It's like, et cetera. But this idea that, like, with a consumer app, you're gonna, like, lock people in because of their data, I just don't believe it. It, it's one of the only things the Europeans did well with GDPR. GDPR's a piece of shit, but-

Guest: Well, wait, wait, wait. But, but, but some significant portion of the average population, which again today is probably just on ChatGPT, is not anywhere near sophisticated, sort of like what Scott said about you. You are special, you are unique, and you are advanced user, and the vast majority of people do get habituated and will just use what they've always been used.

Sam Lessin: Sure, and that's why lots of people still use yahoo.com for search .

Peter: Yeah. So it's not about... I, I never said that there's lock-in, Sam. I didn't say there was lock-in. I just said that there is, like, there's irrational preference.

Guest: Can you take us to go back though and ask, like, what's the ideal customer experience that would be most likely to be successful? That is like we're... From a tech and sa- trust lens, it's one thing. But from a consumer, they, like, get the newer version of their iOS software, and they get a magic text message that's like, "Hey- I'm the Apple agent. I've got all access to everything on your phone. Don't worry, it will never be in the cloud. It will never be exploited by a third party. What do you want me to do for you now? And you just, like, go through and give it connectors and access as you need certain things. It saves your login as you need it. Why wouldn't they just win this?

Sam Lessin: Well, I think Apple might win it, but it-- I think the interesting thing is I don't think you wanna be the one who takes on that liability. If anything, I think the goal is to, like, radiate out risk to other people and be like, "Well, we're not..." Like, to keep yourself safe and be the one that's aligned. 'Cause, like, the reality is you're like, Peter, you're saying that you think the ban is gonna go up. The reality is the hacks are only gonna continue to grow and scare people more and more and more, right? Not less and less and less. Everyone's gonna have them. And so I just wonder if the real thing is gonna be if you really wanna win in, like, the new economy, in, in this AI thing, you have to somehow figure out how to hoard the data and the trust and build the trust. But then when things go badly, always have a third party to point at to blame.

Scott: I don't-- I'm, I'm with Scott on this. He's got the name advantage, number one. But number two, I mean, Gemini just hit a, a billion users. It's pretty obvious it's friction. You get rid of friction, and most consumers are gonna gravitate in that direction. People are lazy.

Peter: I would question that metric. I think they're probably counting, like, the people who are just typing in the search box and, like, maybe ex-clicking expand. I-- There's many ways to game the metric, but I, I think distribution is a good part of it. But I do think consumers' taste and their bar goes up in terms of what they want to use, they have to really like using. I mean, there's many examples of this in the past, why certain apps win in the App Store and why others, you know, why Instagram took off and a bunch of the other ones didn't. Consumers have a discerning eye.

Sam Lessin: Wait, sorry, just to be-- Consumers are also really cheap. Consumers want free, and they're lazy, and they're lock-in. I mean, it's, it's kind of one of those things, like, is it really better experience or is it cheap, accessible, free, and habituated?

Peter: Okay, so let me ask you this, like, Siri is available. When's the last time you trusted it with a question?

Sam Lessin: Well, Siri's crap. That's not a thing. That's not even the same class.

Scott: That's why they hired Google to build a real AI agent, which will be here yesterday.

Guest: I was just gonna... I wanna go back to Scott's thing because it was a sort of a profound declaration that Apple wins, and how you define that, but they have not spent the CapEx that everybody else has, which is an advantage. They have waited. You could argue that what they did in search, which is effectively get paid to let Google, you know, be on the search bar in Safari and in ecosystem, they could do the same thing in AI now. There are two facets of my phone that are probably the most used today, and Sam, you can enlighten me if there's something that you found for MCP plugins or others that cross your agents are working, that are the highest frequency utility and to me, the l- the least surface area exposed to AI, uh, productivity, which are my WhatsApp chats and my text messages. It can ingest my Gmail, it can ingest my calendar, it can do search, it can query all of our internal docs, it can use my Google Drive.

Scott: Location's pretty key too. No one's exploiting it the way it will be exploited, but location's key.

Guest: So add those three things, though, because each of those are today's sandbox. Meta with WhatsApp, Apple with Messages, although Claude can sort of access if you're using it through your laptop, and then you can use, which a lot of people don't trust, ScreenPipe, which, which basically does, like, constant DVR recording of your screen. And if your WhatsApp is open on your laptop, which is probably ten percent of my utility, nine percent on my phone, then it'll record your WhatsApp messages. But, but those two things, to me, to Scott's point, feel like Apple could win on those.

Scott: Add two more. Payments, so Apple Pay, so they're seeing where you're spending your money. And then the second thing is location, and I don't wanna underestimate that. I created an app called Dipity, which I wanted to be like Snap Maps for old people. So you can't see where people are, but you know if they're near you or you know if they're in town. So I built this thing, and it was... it's pretty good. But as I dug in, Apple and Google or Android have both locked Wi-Fi location. You cannot access it as a developer, and that's why Find My is so freaking accurate. But me, the independent developer, I can't get near it. I'm looking at pinging towers and, and it's just stuff like that that I think Apple keeps a lock on. I don't want them to win. Let's be clear.

Sam Lessin: Look, to be clear, I think there are certain things that Elon has said that I actually really grudgingly completely agree with. And, like, one thing I think he said several times, I'm not sure if it's his or someone else's, but I think it's true, is that the most hilarious outcome is usually what's gonna happen. And Apple winning is the most hilarious outcome, right? So I think that's, like, a very good chip in the factor of, like, Apple wins by doing nothing. And actually, it's not even 'cause they didn't, they tried to do nothing. From everything I've heard, Apple internally is extremely confused organizationally on AI. There's a lot of reasons they're having trouble moving forward on it. So in some ways, like, organizational dysfunction creates the buffer for them to be highly successful by doing nothing, which is hilarious. That's the Apple thing. Here's the thing I wanna ask about is, like, are we all just arguing about what has APIs and what doesn't? Like, ironically, the reason that iMessage and WhatsApp are-- have some defensibility or lock into them is they don't have APIs. Like, forget MCP. MCP is stupid. Like, you can just use API. I don't really understand why MCP is a big deal when everything has an API, you can just rip through it with Fable, right? So, like, there are things you can access programmatically which have no barriers and therefore go into the hive mind. And then there's this interesting exception to it, is, like, Apple as a platform owner, which can enforce API rules, right? And they can have effectively APIs to things that no one else can have APIs to, right? So is the real story just it turns out that openness and API is bad, right? And that what you really want is, let's close the stack with valuable data you can have uniquely.

Rachel Holt: I mean, this is literally what's going on right now in, like, the industrial and the physical AI world. This data is not readily accessible. You can't commoditize it nearly as quickly. And so it gives startups, like, a chance to actually gain distribution in, like, a more methodical way before the mega platforms can swallow it up. Like, this is exactly what we're seeing play out, and I think, like, digital and physical are two totally separate things. It all has to do with the availability of the data for the mega platforms. It is really closed. In some cases, it's not even, like, accessible by anyone. It's, like, on a PDF somewhere.

Guest: Right? And it's, like, literally the opposite.

Sam Lessin: Will that continue more so, or you think that will-

Guest: Well, I think you-- it creates choice around who gets the data in the continuum and the, like, the race between-- that exists between startups and mega platforms. Like, in some ways you think it actually gives startups a chance, right? Because if they can gain distribution faster than, like, a mega platform can innovate or a hyperscaler can commoditize it, those factors actually favor a startup maybe more so than, than what you're seeing in the digital world.

Peter: It, it definitely evens the playing field. There's just a lot of data that's like it's not even in a PDF somewhere.

Guest: No, the factories weren't connected.

Sam Lessin: And that becomes their defensibility.

Guest: Yeah.

Peter: A good example is, like, we have a machinist population that's retiring, and a lot of the knowledge is actually in their head, and I think that, you know, companies that can go and figure out what does it sound like when a tool bit's gonna break and how do you-- what is the path that you take? I mean, those are, those are bits of data that I think, you know, in some ways, Sam, I think you and I, uh, saw this at Facebook. It's like, what did Facebook do? It just digitized information that was in our head. Who's your friend?

Sam Lessin: Yeah. No, a hundred percent. I mean, I used to... We used to talk about, I mean, like, you talk about the composer and Facebook era, it was a data pump, right? It was an efficient free data pump to bring data online that couldn't exist anywhere else. And the reality is, if you go way back in the Facebook history, the real story was on day zero, internet's scary, so no one wants to post on it. You're never gonna put a real photo on it, da, da, da. Facebook created the trust and security for people to take a bunch of real world data, photos, relationships, messages, et cetera, and bring it online, right, for the first time, and created an enormous amount of value doing that, right? And so now we're in this weird multiverse where it's like the internet has all been scraped to hell, right? Like, good luck, New York Times, with their lawsuit, right? But, like, that has been, like, copied a thousand times and distilled a thousand times. You have personal private digital information, what's in Gmail, what's in da, da, da, where people have some say in theory as opposed to it just all being, like, dumpable effectively. And so that's where we were talking about the trust, things like that. There's then, like, the, the whole class of companies that are out there trying to, in some ways build new data sets, the Handshakes, the Scale AIs, where you just pay a ton of money and you turn financial capital in some form into, like, unique data. And then there's all the offline stuff. And so I guess the question for me is, like, assuming that multiplying big numbers is a commodity and will be, right? Like, is the story you just, like, really... You go back to day one, which is, like, if your data isn't unique, you're not gonna be valuable long term.

Peter: I think there's also a class of data that you didn't talk about which is just something that Periodic Labs is doing, which is, like, science data and material science data that doesn't even exist yet. You have to actually do wet lab experiment to get that data. So absolutely. Like, finding those bits of data has always been, I think, the name of the game. It's go- it's going all the way back to Facebook, and you and I used to talk about this in those small little rooms about how we're just trying to extract a bunch of data, and how you obsessed you were about buying the DMV.

Sam Lessin: Not just trying obsessed. I did-

Peter: Yeah.

Sam Lessin: No, that is true. Peter, we used to have meetings for what we should buy, and my pitch was always that the Facebook should buy the DMV.

Peter: The DMV.

Sam Lessin: Okay. Can I ask, can I pivot the conversation a little bit, unless people wanna go further on it? There's another tack I'm curious people's take on, which is open router. Like, talk a little bit about enterprise and things like that. If we're saying that your data is so valuable, right? And but at the same time, enterprise really cares... Oh my goodness, a guest visitor, Jess Alessi.

Guest: Oh, hell yeah.

Peter: Wow.

Guest: How's it going, guys? How's Sam doing?

Sam Lessin: We're so deep in it.

Guest: I, I am so jealous. Continue on. I'm just here. I'm just, you know, in a taxi listening.

Guest: J-Jess, just to get you up to speed, we've got two Sams, and we've got two Scotts, but there's uppercase and lowercase Scott, so...

Guest: I got... Guys, I, I am so happy-

Guest: There he is

Guest: ... to see all of you. I just came back from the big city of Manhattan, so I'm, I'm en route. But dive in. Dive in. I'll just, I'll add a little bit later on.

Sam Lessin: Are you-- Wait, are you in New-- I just wanna... For dinner plan, are you in Massachusetts or in New York?

Guest: I'm in Massachusetts. I made it back.

Peter: Oh, you are? Oh, great. I'll see you soon.

Guest: I was in Midtown at three thirty. It's, uh, it's really a miracle I'm here.

Sam Lessin: Open router. Routing. Everyone wants cheaper stuff. AI, the, the m- matrix multiplication commodity. There's no reason to pay a PhD to fold your laundry. But you also don't wanna give Open Router all your data. Or do you? Or do you not care? Like, how does that play out?

Guest: I, I do think that the bigger question here, which is gonna also bifurcate, it's not gonna be one takes all, is open versus closed. And, um, I am saying this obviously with a bias, as I've said in the past, that where you stand on the issue depends on where you sit in the cap table, and we're large investors in Hugging Face, which made a lot of news over the past few weeks. But I think that the future is for the more sophisticated users who serve for the enterprise, it's gonna be your longitudinal proprietary repository silo of data that is yours, that is, exactly as we were just talking about, uh, not ingested by the machines. And if you have that and can run open source models on that time series of data that is exclusively yours, whether you're a pharma company, a finance company, insurance company, a retail company, a venture firm, increasingly you are realizing that to get the benefit, you're giving a benefit, and you're gonna say, "I'm not gonna give that data," especially if I'm not gonna give that data and watch as one of the closed labs competes me away. And you saw that first with Figma, which was sort of scary, and I think the push now from Anthropic and OpenAI to go into pharma and pharma companies saying, "Whoa, whoa, do we wanna do this? Do we wanna give them our data and then potentially open up competitors?" And I think there's a big move right now for them to try to ingest as much information from as many non-internet sources that they can, and I think that there's gonna be a movement underway, whether it is industrial, manufacturing, defense, pharma, biotech, and all of our personal data, that we're gonna say, "You know what? The models are good enough, they are performative enough, and I'm gonna retain the sovereignty of that in the same way that Sam does, and I'm not giving it over." So I do think that that's where the value's gonna accrue.

Sam Lessin: So where are we in terms of people waking up versus, like, door closing on the... Like, are-- do you think that these guys are gonna get away with getting enough data before people wake up? That they're gonna be competitive, or do you think that the, the game's too obvious?

Guest: I think that this past month since the OpenAI hugging face moment, which I think, you know, there's probably three or four factors. That's one. Companies waking up to the token economics and realizing, like, how much are people spending and, like, what's the productivity gains we're getting? The rise of the Chinese models, and then a combination of, like, fearmongering and regulatory capture or attempt thereof against the contrast of people that have been in the White House and been the sort of AI czars that are worrying about this and saying, "No, no, no, if you use these cutting-edge Chinese models but you use it on sovereign hardware and in instantiations that are effectively air-gapped, you have n- nothing to worry about." So where are we? I think that still, going back to the first point about habituation, the vast majority of people are just gonna continue to use closed labs and trust it more, and they're gonna try to ingest more and more and more. And you have whole teams, uh, particularly at Anthropic, but increasingly at OpenAI, that are trying to make big pushes into every enterprise. I mean, Scott, I'm not sure if you can reveal, like, at A24, but, like, a lot of proprietary data and information on everything from, like, every pitch that you've never greenlit to the ones that you have that never got made to... You know, that's a trove of valuable information.

Guest: I don't think anyone wants to, like, share their company data with a company. Yeah, I, I do know a number of startups who were kind of using Claude individual accounts and, you know, using their ramp card type stuff, and then now have to, for whatever reason 'cause of scale, go to the enterprise plan, and they realize that it's, like, 10X more expensive. And then they're realizing that they're using a lot of tokens for things that are pretty mundane, and then someone on the engineering team is like, "I think we can do this with this, like, open model, but we can... As Josh said, there's no risk. It's, like, local. We can mess with the weights and do whatever we need." And so they're now, the next question is, okay, what router should we use? And it seems to me like there's a lot of router options out there, and it's gonna be probably pretty commoditized pretty quickly. And so you're gonna all have routers that direct all sorts of spend to lo- more local, more efficient models. Like, this is gonna become the default. My question is actually, like, what more opportunities do routers have to not become commoditized? Like, I thought you were gonna go in that direction. Like, routers can start to ca- carry logic around which models you should use. Why-

Sam Lessin: Well, they, they're all gonna do that. They're all gonna do that, right? Like, and they're already doing that. I think the question is, I think was what Peter started with, which is, like, at some point, like, you do have to expose something to them or you just build your own. But, like, who do you trust, right?

Peter: A lot of companies are building their own. It's not rocket science to build it. And, you know, Josh, I'm on the same side as you. I'm board member at Arena. We see every week the models change and who's the, who's the leader. And not only that, Arena has a great insight into exactly for each use case, accounting or whatever, what is the actual leaderboard for it. And quite frankly, you don't need to drive a Ferrari to do the, the task of a certain smaller job. You're gonna see a lot more of this. Like, I, I believe in a multi-model world, and I believe there's gonna be a lot of customization and tuning, and you got companies like Applied Compute that's, they're doing this and many others as well. So I think it's gonna be a, a pretty tough road for something like a router. I just, I feel like either roll your own or you gotta find another way to make a model your own.

Sam Lessin: Okay, this is a great bridge. Can I bridge this? Let's talk about the crazy shenanigans NVIDIA's pulling with financing these things 'cause the, the question you gotta ask yourself, right, is like, okay, multiplying big numbers, kind of a global commodity. You're not gonna have PhDs folding your laundry. You're gonna route smartly over time. Like, what percentage of things do you actually need an edge model for? It seems like you have this machines that can compute numbers, you have people who can use numbers being computed, and then a lot of stuff in the middle, right? And it seems like OpenAI... I'm sorry, NVIDIA, with this new 500 billion Wall Street crazy alliance with everyone, is propping up, putting a lot of computers in places. Like, discuss.

Scott: Do you guys... Do you remember Lucent? I know I'm dating myself.

Sam Lessin: Give us some history lessons, Scott.

Scott: So Luce- Yeah, sorry, old guy. And, and for the record, Sam, I'm now in my 13th year independent venture capital. I was in Goldman for 13 years, so at least-

Sam Lessin: Oh, so you're 50/50.

Scott: Well, no, no. And then I was at a startup in the dot bomb trying to beat Google, Looksmart, for four years.

Sam Lessin: I used Looksmart.

Scott: There you go. Why just look when you can Looksmart? Um, so okay. Lucent put-

Sam Lessin: Did you come up with that? Was that your contribution, Scott?

Scott: Yeah. That, that explained the, uh, the bust right there. 99 captured. Lucent put what? 8 billion, 9 bil- I can't remember what the number was, into fiber.

Sam Lessin: So like a Series A.

Scott: Yeah, exactly. Your next Series A. No, they put, like, 8 billion, which was a big number back in 2001, and they put it on their balance sheet, right? So they basically did vendor financing. They said, "Okay, everybody's gonna need fiber." And they were right. What they were wrong, which is exactly what you were just talking about with Open Router and some of the other developments, is the whole concept of wave division multiplexing. Nobody baked that in, and all of a sudden you get 100X improvement on fiber. And so your demand and your pricing fall through the floor. Let's just assume for a second it doesn't fall through the floor, and we've got a half a trillion dollar financing facility thanks to the six amigos that lined up to lend their balance sheet. None of this goes back to NVIDIA. This is why Jensen's brilliant. He basically set this up as SPVs that will basically borrow against the take or pay agreements of the off-takers, of the Neo cloud guys that are saying, "We're all gonna need $3,000 a month for our personal agents." And they are signing take or pay agreements. So independent of whether or not, Sam, you keep your $3,000 agent alive, they still have to pay for the demand. So for a couple years-

Sam Lessin: To be clear, the $3,000 is just for databases and app servers. The agents cost something else. But yes, point taken.

Scott: Sorry. I, I don't wanna cause any marital issues when you talk about shopping and budgets. The take or pay agreements give some security. NVIDIA is basically saying vaguely we're good for 25% collateral, but the real onus falls on BlackRock, Blackstone, Goldman Asset Management- All the guys that lined up, and those are like pension funds, right? Those are assets they're managing for the public.

Sam Lessin: So, okay, so is this even worse, Scott? Are people gonna be like, "Not only is AI, like, using all the water and have the wrong answers and all that stuff, but it also is gonna make most people poor in the process of, like, Nvidia stockholders being, getting well off"?

Scott: I mean, it's still a drop in, in the bucket for overall assets for these guys, but you have to assume a double collapse, right? You have to see price collapse, and you have to see demand collapse, which is basically what we saw in the mortgage crisis, right? If you see both of those collapse, then you have a systemic issue. Otherwise, it's idiosyncratic, and it just will compress margins if it's price. The way Nvidia set it up was they said, "Look, we're gonna make these kinda generic, and so anyone can use them. So if I have a default from one offtaker, I can just move it to the other offtaker." And so there's some good protections in there, but it's fascinating and-

Sam Lessin: But not if Apple wins. Not if they're doing it on device

Scott: I don't necessarily think Apple means on, on device.

Sam Lessin: I know, I know, I know, I know. I'm just-

Scott: Or edge

Sam Lessin: ... I'm just messing with you.

Guest: Scott I think nailed this because the big difference whether lesson learned from Lucent. Lucent actually from their balance sheet lent the money to their customers so that the customers could buy Lucent. Nvidia is taking no balance sheet risk on this. They're basically just convening the lenders, and, and the lenders see that there's opportunity in the spread. But to your point, yes, the people that are being stuck with the risk, I'm pointing, but it's like my mother, your parents, it's the pensioners because that's the capital that's in those large funds, um, that are aggregated. And then you have, like, duration mismatch between, like, the hard assets and the financing. So, like, the chip, depending if you look at Meta or Amazon, debate whether it's a three-year or five-year depreciation, but it's, you know, called three. The debt amortizes over 10 years. The power constructs and contracts are, like, 10, 15, 20 years. And so by any measure, the critic would be right to say that the liability outlives the asset. And any time you have a mismatch, forget about bubbles-

Scott: That's a problem

Guest: ... that's the problem. Yeah.

Rachel Holt: Yeah, I mean, they're assuming, right, these are gonna act like any asset that's gonna spit off cash for a extremely long duration, and, like, this isn't a toll road. You know? This isn't something that, like, inherently is gonna have value 30 years from now. And I think that's, to Josh's point, like, the fundamental risk and question is in that, you know, how much money can you spit off versus the duration that these are useful for.

Sam Lessin: Although just to push on this, and this is a place I don't actually know the answer, but I'm sure someone on this call knows better than I do. It's like in terms of it being a real, like, you know, how long the duration is and the value, how much of this is financing power and shells and things like that versus, like, the racked GPUs and connectivity? 'Cause, like, I would argue that, like, the shell and the power and other things like that actually does have a longer duration.

Scott: No, this is chips.

Sam Lessin: This is just pure chips.

Scott: Yeah, and, and look at the H100, right? Like, the H100 used to be eight bucks or something like that per hour, like two years ago, and now you can get it for two bucks.

Sam Lessin: Well, it was 170. It's back up to two something or whatever, but yeah. Like-

Scott: That's in like two years.

Sam Lessin: Yeah. Yeah.

Scott: Rachel, yeah, Rachel and Josh nailed it right on the head. You've, you've got a mismatch of duration. It's a duration mismatch, and so you have to assume you can stretch the longevity of the chips longer than what we've seen so far in the life cycle.

Rachel Holt: And they're correlated. There's risk that's correlated here too.

Sam Lessin: And the interesting thing, I wonder how much this is... I mean, like, look, it turns out that in traditional data centers and big hyperscalers, the computers have actually stayed valuable longer than expected is the upshot. Like, people are still using, whether it's spinning disks or whatever. That, that's... The, the a- people model it one way, and I think everyone's over time been impressed that actually there's a more of a tail on it than people realize in a lot of ways. How much of the story of, like, "Actually this is better than we thought" is being carried over into multiplying big numbers in this new era, and is that part of the story that, like, Blackstone and Apollo is swallowing? Or is that above our pay grade? Which it might be, 'cause I'm talking in half-truths. Or is it just they have so much money and nowhere to deploy it, and they need asymmetry?

Scott: I think it's a macro bet they're making, and that's it. I think, you know, w- we're probably overthinking it. Fitch is doing the rating work right now, and they have an open consultation, which basically means they're trying to figure out which end is up because no one's ever securitized compute. It is fascinating to think about securitizing compute. That's basically what's going on. And Jensen's a genius, right? He pitched Wall Street on this. Wall Street did not go to Jensen and say, "Hey, we have an idea." He went to Wall Street. So the guy's genius. And to securitize compute is pretty cool 'cause, no, it's not that. I mean, we've securitized other things. However, I have to do a shout-out to my last piece about Americans getting equity in the AI thing. If it's securitized, it makes it even easier for a dividend for all Americans, but that's for those of you who read my newsletter, all one of you on this call, Sam.

Sam Lessin: Uh, my bot's opening it. I'm kidding.

Guest: Am- American participation in all this I think is a great virtue. I think the, the Trump accounts, I think what Michael Dell has done, I think what, what, um, uh, Brad at Altima, I think it's awesome because, uh, you have a generation, you have a generation of young people that either are apathetic or feel left out, are driven to socialism and communism, are graduating with a higher unemployment rate than, like, the average American. Like, all of that, like I just, I think it's good. It'll take time, and, like, the political entrepreneurs that are trying to co-opt these people, you know, it's like a race. But Scott's point on, like, the collateral and this, like, compute not having been collateralized before, we don't know the depreciation. It may be that new models, new algorithms are able to take older compute and find great utility out of it. The GPU boom that benefited first from crypto mining that then sort of fell off and then regained again from AI, you know, sort of saved. But these words are dangerous words, right? Just like this time is different are dangerous words. Collateral, debt, collateralized debt obligations. Like, that was, you know, for let's say, uh, 20 years ago roughly, so for the average, you know, 27-year-old today, they were 7 years old when this shit was happening in housing, which was an asset class that we assumed could only go up. And then it was arguably, you know, if you read like the academic literature of like John Geankoplis, it was a collateral problem. Like it wasn't just the bubble, it wasn't debt, it was just like the proverbial Merchant of Venice pound of flesh. It was the collateral, and the collateral became worth less and the debt became worthless.

Guest: The analogy there, like that moment from the film where-

Sam Lessin: Margot Robbie in the hot tub?

Guest: The, yeah, in hot in Florida, and she's like, "I've got seven houses," and you know, is that, is Jensen her?

Guest: No, because he's not as good looking.

Sam Lessin: Yeah, the jacket doesn't do it for him. What, uh you know, I always say, Rachel, I'm not sure if you agree with this, but I always make the point, it's been wildly successful. I actually credit it to the housing crisis. Because basically if you look at the like seven or eight people we started Bain New York with, right, at Bain New York with in 2005, we were all basically coming up to private equity jobs exactly as the economy was falling apart. And so everyone who was gonna go into private equity, 'cause that's what you did, basically got kicked out of private equity 'cause there were no jobs, right? And everyone started companies. And a huge percentage of them have been massively successful. And so you know, it, it's, it's very, you know, that, as much as we were young, it's very much in my mind what happened there, and I think some of the weird unintended consequences of it. But-

Rachel Holt: Why we have like the best three-year Bain New York class ever, right?

Sam Lessin: It's why, I mean, I, I frequently say it, I think it actually is in-

Rachel Holt: The best network

Sam Lessin: ... demonstrably true. No, no, I'd also say like I'm like the least successful person from that, and I've done okay. Um, like so it's been fine.

Scott: Hey, Sam, I wrote my social studies thesis on affordable mortgage lending in 1993, so I actually am the one to blame for the financial, the housing crisis, in case you're looking.

Sam Lessin: I knew it was your fault.

Scott: Yeah.

Sam Lessin: I knew it.

Guest: What happened to the whole report of these big AI labs like giving equity to the government? Is that, that just like disappeared. Does that not happen anymore? Is it not happening?

Sam Lessin: That was the most cynical bullshit I've ever heard from my perspective, even though I do actually like Trump accounts and the idea of the, the, the index. But yeah, what did happen to it, other than my personal take?

Guest: I just don't know if like the AI lab equity is the right thing to distribute bec- for obvious reasons, but I do think that there's some, like the collateralized compute actually might be a better, or some sort of like instrument that exposes you to the cost of the rising appreciation of energy and compute and... I don't know, like how do you get exposure and distribute it like Norway did to like oil proceeds to like address what Josh was saying?

Scott: This was the op-ed I just did with the information, and then the little interview that Sam... Sam, what'd you accuse me of? Being a mercantilist or something?

Sam Lessin: That you're a mercantilist. You're a mercantilist. You're not a capitalist.

Scott: And so then I came out with a, a newsletter post of like the reaction to my idea was way more interesting than the idea itself of all the-

Sam Lessin: That's usually how it works

Scott: ... of all the ideological battles and, but the title of the post was, "I Don't Have Property in New Zealand," 'cause that's where all these conversations end. Everyone throws their ideology around, and they go, "What's your escape plan?" And I, I don't have an escape plan. I'm with you, Scott. There's merit here. We have to th- rethink the social security system. We have to rethink-

Sam Lessin: Let me push on this for a second. Then I wanna talk, I really do wanna talk about one or two other topics before we wrap on the day. I hear you, and I totally think it'd be great to have Americans more dealt in in a lot of ways. I think, and obviously I believe deeply in capitalism and would love people to be more pro-capitalist and feel part of the story. Here's the problem. Historically, appeasement never works right? It just doesn't. Whenever you appease, you're like, okay, a- and the reality of capitalism, right, is that the story is supposed to be, "Hey, if you work hard and get lucky, you do really, really well." It's not a social security net story, right? And like i- is the American story, I would say. Now, we had a lot of benefits. First, we had a country which unfortunately all the people who lived here originally died from plagues and then wars. We had a huge amount of space, we had n- natural resources, we had Western expansion, we had the internet, we had all this terra nova, which meant that historically if you're an IQ 100 person and worked hard, you'd do fine. If you had a shovel, then it was like, "Oh, just go to school and you'll do fine," because there was so much demand for like intellectual jobs. Those stories have run out on the capitalism side and the American side. It's not clear what the next story is, but the story of, "Hey, here's 1,000 bucks. Hopefully it compounds in the market," I don't think solves the underlying problem of like how do you create opportunity for people and like the American story of capitalism free enterprise.

Scott: Two distinct challenges. One is a existential meaningfulness question.

Sam Lessin: Ah, that's been my essay.

Scott: And then the other one is how we fund this. How do we put food on the table and police on the streets and kids in school and all of that? So they're linked, but they're very different. And I agree with you, we have not answered the post-labor existential question. We just haven't. N- we don't need more poets, I don't think.

Guest: I think there's just a question of, you know, parents can have money saved for their children and screw them up, or they can have money saved for their children and not screw them up. You know, I've seen both instances happen, and I don't think that there's a rule against one or the other. It's more about how it's done.

Sam Lessin: So it's like a culture and values and-

Guest: Mm-hmm

Sam Lessin: ... all that stuff is like what matters.

Guest: I wonder what the instrument should be. I mean, what should the instrument be of these savings accounts, which I think of, you know, I'm also a fan of, that are being created for kids that are, you know, now being born? Like what, why shouldn't they get some exposure to the... Some of them will become like the deadbeat kid that got, you know, a silver spoon and screwed it up. But I feel like a lot of them will also not graduate with college debt or debt for whatever education ends up looking like in the future and have a little startup capital to take a risk in their lives and all the other stuff. So we're not gonna solve it for everyone, but isn't that net better?

Sam Lessin: I'm very pro the Trump account. I think it's good and smart. I'm pro it. I, I'm not convinced that the picture of how it compounds is so certain, especially as we're talking about resets and all this other stuff that, you know, we're trying to figure out right now, but like I'm certainly pro it. I just, except, with the exception of when people start lobbying for regulatory capture by giving a 500 bucks of their stock to someone, right? Like that seems insane to me, and we gotta figure out how to like not have that happen. I wanna pivot one or two more times before I lose you guys, and I really appreciate you, you joining for this fun conversation. OpenAI exec exodus. Brad Lightcap out. Kevin Weil, who we used to work with, out. Peter, you've been out for, like, years, so, like, you were just-- you're a forebearer on this. Like, why and, like, where-- what's going on?

Peter: My take on this is it's actually a feature, not a bug. It's a feature of having insanely great talent. I think Sam is N of one at recruiting, like, really ambitious, talented people. And in a time when you can build anything, you should expect the churn. These are people who, who wanna go and kinda build something really, really insanely great, and that's not a talent problem. That's just kind of a natural output of what he's optimized for. I mean, you're seeing this everywhere. A-and also the other thing is that it's not just having the models. You have to point the models with sense of obsession at a problem. And, you know, like something like Periodic Labs, uh, Liam left, uh, OpenAI. We were the first check investors into that. They're doing material science. It requires a different level of obsession on building a wet lab that just doesn't exist at a bigger companies, right? So the companies out there, like Applied Compute, Core Animation, other Neo labs, there's just a, there's a point of view that you wanna go chase, and if you're ambitious, you're gonna go out and chase that. That's, uh, largely what I see from this. You have to point the models, and it's a great time to build, and so that's what everyone is really doing.

Sam Lessin: Is it because of there's, like, this sense that these generalized models are the old news and specialized models are the future, and it's, like, gonna be specialized models plus all these open local-

Peter: I think it's a great question. I, I think there's a lot of stuff, and Rachel, you, you and I lived this in Uber. The last mile problem is a real problem. You can't just drop GPS on phones and be like, "Okay, all right, let's give rides to people," right? One of my analogies I use is, like, you can't vibe code a home health nurse to show up at, i-in your living room to take care of you post-surgery. There's a company I'm working with called Adaptive based in New York. They're doing exactly this. They're like an AI sort of healthcare provider delivering healthcare, and that delivery, Rachel, you and I know this from Uber days, is not easy. There's a lot of stuff that goes on to make sure that that value can be delivered, and that requires an obsession that goes beyond, "Oh, AGI's gonna solve everything." So I think that's what it is, is, like, when you're obsessed over a problem like that, you're gonna go make a company and make it great.

Guest: I, I've got a quick cynical take which is, uh, Peter, how long were you at OpenAI?

Peter: Year and a half or so.

Guest: Okay. And Kevin, I think, was there maybe four years, maybe a little bit less, uh, three years?

Peter: No, I think he was there for less than that, but yeah.

Guest: Two? But my cynical take was gonna be you go, you're there at evaluation, you get stock, you vest. The stock appreciates, maybe appreciates very rapidly, and then you're looking at the incremental upside from where you are and what you're doing in an ever-growing thing and having to deal with more people and more politics and more knife fights and all that kind of stuff. Or you can take the cachet of having been where you've been and you have a robust market that will finance you. Kevin is gonna go, he's gonna start something. He's gonna get funded. We're probably all gonna be involved, you know, with whatever he funds. And I, I think that that's the simple calculus is, um, you believe in yourself, you, your financing risk is pretty low, your talent risk is pretty low, and your upside is gonna be way higher. And so the incremental return on the incremental expenditure of your time is worth leaving. So I don't think it says anything, and I've been critical of every one of the labs at some point for whatever, but I don't think it says anything about OpenAI, uh-

Sam Lessin: So but Josh, but d- can you build a company that way? Like, with people coming in, stamping the card, getting infinite free capital for their next startup. Like, does that work?

Guest: I mean, so far with OpenAI, valuation has gone from, like, you know, seven hundred million to trillion and plus, and you've had six diaspora companies that have all achieved, you know, tens of billions of dollars that have spun off of it, and I think everybody looks at that. And by the way, the number one motivator that I've always found is somebody looks and sees a peer that they worked with and being like, "Are you fucking kidding me? That person just raised how much and what?" And then that becomes the catalyst, you know, moment where they're like, "I'm going to do that."

Sam Lessin: Well, a hundred percent. The, the talk after is usually, "And that guy's an idiot."

Peter: Look, I think, I think all, uh, b- your point of view could be true. My point of view c- could be true. I don't think they're directly contradicting. I'm not saying that you think they are. But you're right. Like, it absolutely, when you ha- are able to take the risk to go build something bigger, that's kind of my point. And what I'm saying is that these people are ambitious people who are like, "Okay," and, and implied in that is like, "Okay, I can stick around." And the way I'll put it is the physics of a big company, you know, Sam and I, we, we've been to big companies. There, the physics are just different. And the people who are wired to go, or Sam Altman has a-assembled a set of really great researchers who have these big ideas. You're gonna want the freedom to do that e-elsewhere too, right? So again, it's just, it's just a natural consequence.

Sam Lessin: There are two directions I wanna divert us and jump off. One direction is, sure, I agree with that. But the early OpenAI pitch was, "We're building AGI, winner take all. You're either on our train or st- have fun staying poor." Like, it was, like, a one thing. That has kind of evolved into a lot of people who are really smart being like, "Oh my God, this is the moment where everything matters. Cost of capital is zero. I gotta take my shot. And by the way, I've already punched my card four times. If it does turn out to work, I'm protected. I've got plenty of stock."

Guest: How different was it in the Facebook days? 'Cause you had sort of the same dynamic.

Sam Lessin: I think, first of all, it's kind of funny 'cause by historical standards then it was an incredibly meteoric rise. Facebook's valuation grew unbelievably slowly compared to the labs, right? And there were obviously generations. Like, I would argue I was... Peter, you were kinda generation one and a half maybe. I was generation two at Facebook. There's now been generation three and generation four. But there's not 20 generations, right? There was, like, a s- a set of people who worked for 40 years, did great work, and then passed the torch, and that's happened, like, a few times. And actually, the core leadership team at Facebook has been unbelievably stable now for, like, a decade, right? Like, which is a very different dynamic.

Peter: And, and Josh, the, the other thing to, to look at just to, you know, not to be too cynical is that, like, Facebook had great network effects. So if you wanted to build anything in social, that's where you stayed. I gotta say that if you take a look at A-AI, like, you know, again, being on the board of Arena, it just changes every week. It's like, "Well, who's up there this week at this point," right? And I would say AI is a lot more akin to something like, you know, GPS on a phone or a database where it is a technology. But again, you, you can't just, like, vibe code an Uber. You have to deal with all the real world problems. You can't just vibe code a home health-

Guest: Your network effect is actually a profound point because I think that Meta has actually done an extraordinary job in their acquisition strategy, and I think part of the pitch that Mark has made, at least in two of our companies, is the scale that we can give you

Guest: to deploy your product, whether it's a brain machine interface like Control Labs or, you know, like is enormous. You're gonna reach billions of users, and I don't know that OpenAI has done that. So that, that actually is a good point

Peter: Yeah, I, I learned how to do acquisitions from my time at Facebook. That was d- that's exactly the story.

Sam Lessin: I don't wanna keep harping here. I wanna ask Peter one more thing, and then I wanna pivot to a few other people before we close. Peter, and I, I have not... He's a friend, but I haven't spoken to him in, about this in months. Is the Airtable Howie story, "I wanna work on AI," or what happened?

Peter: I haven't spoken to him in a few months either, but let me tell you a few things. You talk about Airtable, the exit, Bending Spoons, et cetera. I think there are two truths here, and both truths are true. Number one, I think founders are built differently. I mean, you know, founders are gonna continue to build towards AI, like you, like you alluded to. Howie pivoted to Hyperagent long ago. I think Scott built... Uh, Scott, you know him well, right? Like, he made that pivot, and he's been pivoting, and the thing is that acquisition, I think this is public knowledge, was had Hyperagent carved out, so that was really the, the, the legacy business that was acquired. So great founders who have founded all these companies, like Howie, are gonna continue to build 'cause that's what, that's in their DNA, and then f- a builder's gonna build. The second thing is there's a real shift, and both of these things are true, right? And we, we're seeing that at Felicis too. So in our last couple funds, over 20% of our capital has been deployed to global resilience, energy, manufacturing, defense tech, all stuff that you all are, are investing in as well. On, on science, I mentioned we were the first check in Periodic Labs because we believe that there's actually more real-world problems to be solved. So I think both are true in terms of what happened with Airtable and Howie's just, from what I understand, is just a good build. He's a fantastic builder.

Sam Lessin: Okay. I agree. I wanna go lightning round and call it a day 'cause this has been awesome, and it's really... This was fun, by the way. I think we should do this again sometime. Rachel, I'll start with you. DC regulation, all the heavy stuff you're involved in, what are you thinking the most about right now that we have not discussed?

Rachel Holt: I don't know about a DC regulation piece, but, like-

Sam Lessin: Or just life. Or just life

Rachel Holt: ... life. I'll just go there. We happen to live in DC. Uh, I mean, look, I think-

Sam Lessin: I know. I just think of you as my, like, DC person now.

Rachel Holt: No, I-

Sam Lessin: You know? Like

Rachel Holt: ... I know. I, I do live here, and sadly that's becoming... No, I mean, look, I think- ... uh, I mean, Peter bro- you know, just brought it up at the end. We're spending... You know, they're, they're putting 20%-plus of their fund into these areas that, uh, Josh, you've been investing in even longer than we have. But, like, we've been focused on, and in 2020 when we started Construct, a whole bunch... Now, that was mostly LPs that we were talking about at the time. But people were like, "You're crazy to go..." Uh, the same people who said, by the way, I was crazy in 2011 to join Uber because taxis are... You're ne- you're never gonna get a venture outcome.

Sam Lessin: Weren't you at Clorox before that? What, what were you... You were working on charcoal.

Rachel Holt: One year. One year, yeah. Exactly. Salad dressing. But, like, you're crazy to go join a taxi company 'cause it's, you know, it's slow-moving, it's heavily regulated, it's high CapEx. And, like, those people just, like, didn't spend two brain cells thinking about what Uber was actually building, and I feel like it's the same thing that people were saying to us in 2020 when, you know, we went out and started Construct. First of all, there's a lot of software companies. Secondly, there's a lot of physical world companies that don't need to look like really slow-moving high CapEx companies. And, you know, the only way we're gonna get productivity is putting software into these spaces. And, like, if you looked at the signals even, like, four or five years ago, like, the crisis was mounting. You know, we appreciate the markups that come from everyone piling into these spaces. We're also... Now the prices are going up, like, you know, to a crazy amount obviously on, on some of these early stage stuff because a lot of people who, like, would never give these spaces kind of a second look are now devoting half of their, I don't know, billion and a half dollar funds to them all over the place. So anyway, it's an interesting time. People actually come through DC. It's great for our large conference room, which now gets use, and good for the existing funds and the portfolio.

Sam Lessin: Fair enough. Scott, we didn't really get into proof of craft, provenance, watermarks. You've been thinking about this stuff forever. Anthropic just is now watermarking text as of today. Like, there's a lot of stuff going on. I mean, I, I assume you are still deeply on this train. Can you give us a quick update on that?

Guest: The C- C2PA, it's the open protocol that kind of everyone's signed on to for these content credentials to be added to assets as they're generated or edited, by the way, for that matter. It has been widely spread. The problem is on the consumer experience side. These companies like Instagram, you know, and products that are surfacing media, for them it's binary. It's like if it has a CGPA in it that says any AI was used, they report it as being made with AI. And then consumers are like, "What the hell? I only used AI to, like, remove a blemish from my face. Like, why are you labeling me as a slot maker?" And, and so there's... We're in this, like, moment where there's a disconnect between the underlying efforts, which I think are net-net good to, like, help people be able to discern the provenance if they want to, and the consumer interfaces that don't know how to surface this or merchandise this to the end user. So I feel like that's where we're at right now. I don't think it's gonna go away, but I also think we're all gonna be gradually inoculated, and I think honestly, like, fake stuff is really helpful for society right now because it inoculates us and makes us realize we can't trust what we see anymore.

Sam Lessin: It's all fake. Josh Wolfe, got anything you're excited about we haven't hit on or you're freaking out about?

Guest: Freaking about, uh, the rise of socialism and communism in our country.

Sam Lessin: Yeah, I've noticed on Twitter you and I have that in common.

Guest: Yeah. I've, I've been, like, a center-left Democrat my entire life, and my party has just, like, run far left and, um, I'm... Yeah.

Sam Lessin: Well, it's 'cause no one has good... Like, the marketing sucks on the center left. There's no good marketing.

Guest: But I grew up caring about the guy that lost the ovarian lottery in the Rawlsian sense and, you know, needs a leg up and social welfare and-

Guest: But I believe in the American dream, and, um-

Sam Lessin: Yeah, but the veil of ignorance is kinda bullshit. Let's not go Rawls. You a Rawls guy?

Guest: Yeah, yeah. No, I, I, I believe that that's the right model for designing a society. You know, you wake up tomorrow, and you don't know if you're gonna be born in Brooklyn or Bangladesh, Black or white, male or female, able-bodied-

Sam Lessin: Yeah, yeah, yeah. I agree with that part, but I'm more of a Nozick guy myself.

Guest: Um, you know, I, I, I, I think that, um, Europe is fucked, and there are bright lights of amazing entrepreneurs that are doing the very thing that Europe has historically not done, which is risk-taking. The floor for Europe is higher than the US, so people do not fall through that, but the ceiling has never been that high, and I think that there are companies that are breaking through that. But I think Europe is screwed mostly for demographics and influx, largely because of political positions of people that are unwilling to assimilate. It is destroying society. I believe that there are bad foreign actors, whether Russia, China, Iran, North Korea, or whatever, that are helping to foment that. And I do believe that the Sahel in Africa and the Maghreb, which is, you know, Mali, Sudan, Niger, uh, Burkina Faso, you are one terror event away projected into Europe that that entire region becomes the West's next Afghanistan, all of which culminates in my conclusion that Europe needs very strong defense. It is very much a Tower of Babel of uncoordinated systems, but there needs to be an enormous expenditure, which they have all been under pressure of Trump and Rubio being sort of chastised at, uh, Davos and Munich Security Conference. But, um, Europe's gotta spend. We're seeing some of our companies benefit from it. But Europe is at real risk, not just from Russia, but from violent extremists that are infiltrating the continent. And if you care about the West, you gotta defend it.

Sam Lessin: So you're saying we have to go defend Europe again?

Guest: Yes. Yes. I, I, I, I flew over. We were in France, uh, for part of the summer, and I flew over. It was actually really amazing, just, like, you're, you're in a plane, you're flying over the beaches of Normandy, and you're looking at Ohio and Juno, and it's just, like, 80 years ago. It's like, it's crazy, you know? Americans-

Sam Lessin: It's not that long.

Guest: No, it's crazy. It's two generations, three genera- Like, we stormed the beaches of Normandy against the Nazis. Like, it's insane, you know? And, uh, we're, we're, we're not fighting Nazis. Uh, but it's like, I don't know. It's, it's wild, and it's important, and I think a lot of people, it's, like, too weak to be felt until it's too strong to be broken. And a lot of people are really not appreciating how serious this stuff is.

Sam Lessin: Fair enough. Well, on that uplifting note, anyone else have anything else, a jump ball they wanna throw out before we call it?

Guest: We're all gonna die.

Sam Lessin: The first step is talking about it. Second step is doing something about it.

Guest: No, the second step is, is funding. You fund these entrepreneurs that are taking it on. You're getting patriotic.

Sam Lessin: I'm with you. I have my own takes on where the leverage and I can deploy on this stuff is. I do think I'm with you ideologically. I do think it's one of the tr- problems of a Twitter or fun podcast with your friends who are all venture capitalists in different places is we can all fervently agree with each other and feel really good about it. Yeah. You guys, this was really fun. I appreciate you all joining. I hope you had fun doing it. Super fun. This was great. Thank you for hosting. So thank you all.

Guest: Thanks for having us.

Sam Lessin: I hope you all really enjoy the rest of your August, and I will see you around on the internet and in physical space.

Guest: Awesome.

Sam Lessin: Yeah.

Guest: Make sure your, make sure your bot gets the recipe for Jess for dinner. She's gonna be there, like-

Sam Lessin: She, she's here. She actually just walked in if she's still around. Maybe, is she coming out? Are you coming out to end the show? Here she is.

Guest: I was gonna say, she checked in with us.

Guest: If you hear a wedding alarm-

Guest: Wanted to make sure it was actually happening.

Sam Lessin: There she is. There she is. And la-

Guest: Josh, I texted you a question based on what you said that I didn't understand. So if you could get back to me, that'd be great.

Guest: Me? Me?

Sam Lessin: Well, yeah. Yeah. She got a text from Jess.

Guest: Okay.

Guest: What should I eat for dinner?

Sam Lessin: I love that. I'm gonna...

Guest: Tater Tots.

Sam Lessin: Yeah.

Guest: No, I don't...

Guest: Bye.

Sam Lessin: All right. We had a chef, but there's also pizza.

Guest: Bye.

Sam Lessin: Bye, everyone.

Guest: If you enjoyed this show, please leave us a virtual high five by rating it and reviewing it on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts. Find more information about each episode in the show notes, and follow us on social media by searching for @moreorless, @davemorin, @lesson, @jlesson. And as for me, I'm @brit. See you guys next time