Silicon Valley Is Great at Tech and Terrible at Business

0:00 Did you know Meta and Microsoft, two of Anthropic's largest enterprise clients, have basically hald their use of Claude recently? >> Claude must be so lonely. >> And the reason is they're using their own models instead. >> One of the labs, I'll let everyone guess which one, believes that if you don't bring about the machine god, then the machine god, once it is created, will go back in time and put you into a simulation of hell. Like, this is a real belief.

0:29 is more or less. Is it no or is it yes? We'll debate the text that's best when we get more or less. Dave and plus put it all right to the test. More or less. Welcome back to another episode of More or Less. This is Brit Mren subbing in as moderator today for our dear friend Jessica Lesson while she attends to a trivial thing called a team offsite.

1:02 Just kidding. We love all the people at the information and we're glad they're offsiting about some strategy. >> There's one person we don't love, but we're not going to tell you who it is. >> Oh yeah, you don't want to piss off a journalist. Sam, >> you got to keep people on their toes. >> Okay. Yeah, there's one person. You know who you are. >> We might know who they are, but there's one. There's always one. So, dear listeners/viewer, you're getting three VCs and no journalists today.

1:27 >> It's going to be unhinged. >> And I apologize in advance for the that's about to go down. So, let's start with the important stuff. >> We're going to be in a truly fact-free zone. No babysitting today. >> We don't have facts. >> So, since Jessica's not here, let's pick some topics. I want to talk to Nathan Fielder. >> I don't know what that means, but okay. >> I want to hear about your peptide journey. I went to the LA premiere last night. Oh wow, fancy man.

1:52 >> Yeah, I got to give the guy a high five. I want peptides. I want Fielder. And I want Dexus scans. And then if there has to be some tech news, we can talk about that as well. >> Okay. Cuz I have four big tech topics. A lot's actually going on, but but we can start with Dexus scans. >> Well, why don't you tell us what the four big tech topics is and then we'll just do them.

2:09 >> Okay. Four big. We can't speed through them that people want to ingest and understand. I'll tell you the topics. I'll give it a macro overview. We're going to talk about why people continue to not be loyal to Claude. We're going to talk about and how and what that means for their IPO. >> Friends, don't let friends use Claude. Okay, you can use Claude all you want, but if Claude goes away, who cares? >> Okay, we're going to get to your comments shortly. Thank you, gentlemen.

2:32 We're also going to talk about even more personal agents that have come out of stealth with hundreds of millions of dollars in funding. >> We don't need to. >> We're going to talk about the personal agent protocol. We're going to talk about Tik Tok getting into commerce agents. We're going to talk about why founders are 17 years old raising so much money. We're going to talk about local computing, computing at the edge as some call it. And then lastly, we're going to talk about how some people are now starting to buy Pratada shares from early stage funds.

3:03 >> Did you prep us with an agent? >> Well, of course I did, Dave. >> Oh god, no. Oh, they're starting to buy What does that even mean? They're buying pro rata shares. >> I know you're going to enjoy this segment. >> No slap grenades for agendas. One of the biggest fights that Dave and I get into as a married couple, for those who don't know, is my use of AI, which he calls all use of AI, AI slop. And it's not because I iterate with these things.

3:30 >> I'm actually starting to love that I'm in several meetings a week, not including you, where people are like, "No more slop. You need to bring human created stuff to this meeting, and we're not doing slop anymore." >> I love slop. Let's start with I want to hear first Brit about your Dexus scan today. What's going on after the call? >> Well, first of all, last week on the pod we had Matt Mazio join who has a company called Noho Labs and is all about peptide protocols.

3:54 >> Brit's now on all the peptides. >> Are you not on then Dave? >> I signed up and I'm going to do the onboarding this week or next week. I'm getting it into my schedule. >> Did you use our discount code? >> Yes, I did. >> More or less VIP is your 50% off discount, everybody. >> Actually, I learned that Mazio had already given me a discount when I signed up. So, that was nice of him.

4:12 Okay. Well, I have four peptides I'm about to start. I'm a little nervous about it. Well, I've got the Wolverine. >> Oh, I love the Wolverine. >> By the way, all the peptides need way better branding. >> Are you going to grow a beard? Nah. >> We're going to see. It's going to be a 12week trial. Things got might get weird. >> I hope not. >> I've got one that's a bunch of numbers. >> Are you doing Chlo >> Chlo with a K?

4:35 >> Well, no. You got to explain what these are. What is Chlo? It's the blue one. I have this like constant hip and back pain, but it's not something that I can actually solve through surgery or anything. And so the theory is that this Wolverine blend is helpful for muscle relief and pain. And so I'm going to try to inject it close to my hip. >> Wait, is that also the blue one or is that a different one? No. Clo is the blue one.

5:01 >> Yeah, it's a blend of the Wolverine. So it's a special type. >> No. Well, there's Wolverine and the CL, but are you doing separate Wolverine or you're just doing >> No. Wolverine and Clo. Yeah. Tessamoralin. Tess. I'm gonna get ripped. >> Tessamoralin. >> So, I'm gonna get ripped is the second one. That's all I know. And I'm gonna lose some visceral fat maybe. And I'm going to have NAD, but I'm going to use that sparingly after conferring with some some experts.

5:26 >> Yeah, the NAD is not my favorite. >> I'm only going to use it if I'm feeling sick. >> Yeah. NAD should only be used when you're getting sick. >> I think that's right. That's sweet. Well, and then you're probably doing BP like 1295 or some Yeah. >> Yeah. There's one with a random number in it that I also was like, "Sure." >> Why is it that everyone that I talk to, no matter which doctor they're working with, this is the five of these that they go recommended. So, like, what value is there in the service?

5:50 >> It's the stack. >> Well, the doctors have to consult with you to pro because it's a trial experimental. It's a it's a compliance thing. But yeah, this is going to be great, Brad. I'm very excited for you to try them. I'm a few months into the journey. I'm pro. I think it's been great for training and marathon and I feel better. Well, Ann, so what I'm doing today and why we have to be timely with this pod boys is because I have a DEXA scan in 1 hour because I need to get all my stats prep stats and then I'll measure again post peptides and I've been doing my DEXA scans for many years and I have all of my data now. I actually use chat GBT health. I find it's the best health agent because it connects >> Oh, Ashley Yuki.

6:28 >> What? >> What? Ashley Yuki is on it. She's great. But why why use I just dump it all in myself. Why do you need the Chachi BT thing? Well, chatbt health has a bunch of connectors into like my chart and a bunch of hospital systems and medical systems. >> Yeah, but whatever. You just dump it all in. >> No, but I like having the connectors and >> Don't you want to own your own data? >> No, because friends don't let friends use claude.

6:49 >> This is chatbt. Dave, >> I can use whatever. I actually just dumped in is I dumped in a full sequencing.com like my full genome, which is pretty cool. >> You got to get on sequencing.com. >> Isn't that kind of scary? Okay, let's come back to that when we talk about some of these personal agents because going on device and private is now in vogue. Who would have thought earlier this year when we were talking about that?

7:12 >> What are you talking about? That started in January with open clause. This is nothing new. >> So, okay. So, we got Brit's health journey. I'm very excited. I'm like on the same journey and I also I did Dexus game when I got back to California and I'm going to do one November like 5th which the only compounding factor is I am running a marathon November 1st which has other issues but it's an interesting journey. >> Yeah, I love it. I'm like all about the quantified self. So I'm down to experiment with anything and see how it goes.

7:37 >> All right, topic two guys. >> I love how you're the moderator now. Sam, >> I went to this movie premiere last night. >> No one can see that that's listening. Explain what it says. It's the you can see everything movie with that Nathan Fielder made with Elizabeth Holmes. >> I cannot say anything other than this is potentially the greatest movie ever made. It's very long. It is painful to watch at moments. >> Why can't you say anything?

8:02 >> Because it's like it's a whole thing. You don't want me to say anything, but I cannot recommend this movie highly enough. It is high art. I was saying to someone, it literally to me justifies the threehour movie format as opposed to TV shows or like epic series versus like shorts because it had to be the length it was. And I have to just say like I've always loved I love comedians. I think comedy is like the highest art in the world in terms of like the human experience and truth and meaning. I like love comedians. Nathan Fielder like I don't know if he's going to win an Oscar, but he deserves one. Oh, >> this movie is so good. So, I know there's a lot of hype, but let me add on to that and extend it. I thought it was incredible piece of media. Why can't you talk about it? Did you have to sign an NDA?

8:49 >> No, it's not NDA. It's just like it's this movie deserves to be seen without the spoilers and it just it deserves it is the honest answer. Dave, >> cool. >> When does this roll out to the rest of us? Do we know? >> I think it comes out in like 10 days or something. Anyway, I just cannot speak highly enough of it and him and like I literally a friend person I know invited me the day before as like an extra ticket to the premiere in LA and I literally like hopped on a plane and went down and watched it and flew back up this morning and it was totally worth the trip.

9:20 >> Nice. I love that. Okay, Sam, are you now guiding this episode? Would you would >> No, I'm done. I have no other points on the week. >> Okay, great. Well, then I want to switch to some tech news that we can talk about, weigh in on. Did you know Meta and Microsoft, two of Anthropic's largest enterprise clients, have basically h haveved their use of Claude recently. >> Claude must be so lonely. >> And the reason is they're using their own models instead. Like Muse is using its like Muse model and Microsoft's using OpenAI and GitHub Copilot. The end state of all things is narcissism where you just talk to yourself. Muse talks to muse.

10:02 >> Is this going to continue happening? Do we think this adds any weight? >> I think it also overstate its welcome. >> What did >> Yeah. Why are you so negative on Claude all of a sudden? Well, you said you were hating on Claude. >> I just think that the way that it talks to you is so biased and kind of there's like a certain flavor to Claude slop that you can actually sense who on Twitter is talking to Claude too much now.

10:26 I do think that's true. >> I think that that's just bad. It's actually the way that it so fundamentally was tuned to make people get into a parasocial relationship with it, I think is actually the thing that nobody's talking about. You know, you can like think of these things as dangerous weapons or that they're going to kill us all or whatever fear doommongering thing is going on in any given week. But I think actually the thing that Claude did that has burned its has sort of caused what's happening right now is that it was overtuned for parasocial relationship. People got deeply into relationship with it in a way that was pretty unhealthy. And I think people are waking up to it now both individually and collectively as organizations. And I think that's like a big part of what's going on.

11:12 >> So it's interesting. I use a lot of cloud but I don't actually talk to Claude that way. You just use it as inference. >> Yeah, it's just it's just infrastructure for me and it's just excellent at what it does. I mean I published my diagram of like how all of my infrastructure works and like I'm loving it. I as I said on the last pod like my thing is though I think the the warning sign you mind is what we've been saying all along which is this stuff is great.

11:35 Multiplying big matrices is amazing. You can do great stuff with it but it's very very hard to build important businesses for the long term when if it went away tomorrow there's a perfect replacement and always will be. It's fine. And so like I'm not switching off of Cloud. I think Claude's completely fine and does what I need, but it's a oneline change if I want to and like it doesn't matter. Silicon Valley to be silly about it for a second. I just think Silicon Valley is very bad at business. I think Silicon Valley is very good at technology. And the irony is is technology has been so dominant in certain areas that like being good at technology has swamped all sense of business and strategy for a long time. So it's like you don't need to be good at business, you just do technology. But this is a little different. Well, Sam, I think that you've rightly pointed out that Silicon Valley's great at one kind of business, which is create enormous capital bubble to make number go up until the like the one emerges that can outstrip all of the business losses and become infinity.

12:35 That's like the kind of easy way to explain the Silicon Valley business. Although it's a recent phenomenon, right? Because like even as far as 101 15 years ago, the businesses that Silicon Valley generated were like unbelievably capitalist. You know what I mean? Like it wasn't about like moving heavy dollars. Like that's kind of a recent Bitcoin style. There's a series of things. I I honestly think a lot of like what's gone on has been people learning the wrong lessons from like Bitcoin and religion effectively in this era, right? where if you think about like this dynamic it's like you had the era of Google and Facebook which everyone got wrong from like a storytelling perspective largely because like well how big could they possibly even I got this wrong like I was like what is the thing that is more than a hundred billion dollar that's unfathomably large in historical context and so people got oh my god things can be huge then bitcoin which again I was early on you know debtor too and like then ripple and salana and like all this stuff taught people the lesson of like effectively the average which I love this line is this true which is the average of zero infinity is infinity which is like hyper asymmetry because of scale and so people got really comfortable with that GME >> and those were even weirder by the way cuz they had no dilution involved in some cases >> well they were they were even better because there were no delusion they were totally capital light like we talk about the whole like setup of them but you had this kind of like set of compounding narratives I I actually gave a talk about this last week and about how like you know this is like if you really want to understand what's going on with the labs and these stories and the capital allocation It's basically the West's really lack of comfort with the concept of zero and infinity. It doesn't fit in like a western framework, which is why we went around killing all the Pythagoreans in the times of the Greeks.

14:15 It's why the church fought zero so hard. There's like this long history of the western tradition hating on zero and infinity because it like breaks all logic. >> Sam, can you unpack that a little bit? I'm very educated on this, but I actually think most people are not. What do you mean by that? Okay. So, there's a great book if you haven't read it. There's this great book about the number zero and it's called the biography of a dangerous idea.

14:35 >> It's a great book. >> It's an awesome book and it's an important one to internalize. I think it talk about what Silicon Valley's business model is and and kind of where we are. The basic point of the book is that if you go all the way back to like the time of the Greeks and before the idea of true zero, right, was like highly fought. And so like the first historical vignette is the Pythagoreans who wanted to kind of introduce the idea of true nothingness. This is like kind of like creation x nilho from the bible.

15:02 How do you like create something from nothing? These are very deep philosophical problems and the west hated it because it's really hard to reason about and it's really hard to talk about consistent growth and like the things we're good at. Same thing with like infinity. Like infinity is a weird concept. Because of that the zero came from India. The west fought it like hell. The only reason it exists in western culture is that it's just too useful. Like you need zero. Like the traders made it happen. The church fought it like holy hell to like keep it out of western culture. But like fast forward if you think about it like our markets and the way we build businesses.

15:33 We think about competition and compounding and all of things. They're great at like most normal modeling in numbers that we're familiar with what we can argue about and multiply. D the second you introduce a zero or an infinity anywhere in the chain you break everything because the tail which is infinity to zero dominates the whole story right and what's really interesting I think is in some ways if you look at the story that we've seen the last many years in Silicon Valley the tendency towards infinity which you saw the geometric returns on Google and Facebook and oh my god the world is infinitely large and we can dominate the whole thing started this things like bitcoin which is like oh my god if everyone believes then belief is self-perpetual ating and it can be infinitely valuable, right? And it can only be zero. Then you see like GameStop and like this whole series of things. I think there is this in fact in some ways like the tales infected the entire narrative of Silicon Valley, right? And so that's why we get to today's >> which tales the tales like meeting the extreme like zero or infinity like the tales of numbers.

16:32 >> Oh yes yes yes. >> And you're saying we've been on an infinity curve for a while now. Well, we've been putting infinity into our Excel models. And when you put infinity into the Excel model, I think if I play back, Sam, what you're suggesting is that just like, you know, this notion that zero, we assigned a number to nothing. We've now like assigned infinity, it's become part of our models in Silicon Valley in the LA really in the last 5 years and it's like a it's a new phenomena.

17:00 >> Well, I like the Bitcoin example. I think crypto too. Crypto definitely is part is a huge unlock for the thinking. >> But crypto was pretty isolated from the entire game for a while. Agreed. But then it kind of the crypto thinking plus a lot of this stuff came into the AI story. Now you have these two interesting infinities and zeros. Infinity and 01 was like the Sam Alman original pitch which is we will own everything. Whoever gets the AGI first owns infinity. There are no more businesses except for us and it can only go to zero, right? Okay. And so you're like, okay, well, the average of infinity and zero is still infinity. So I guess we should invest in this. Who cares what the price is? Who cares how much capital it needs? Like that was an infection one. Now we have the inverse going on, which I think is to be taken very seriously, which is it might be that it's actually not. And like that was a silly narrative when you think about it, but people believed it enough.

17:50 People still believe it, Sam. >> People are still trying to buy. >> I mean, just to interject one point and then get to your next point. I was at a dinner last week and there were a bunch of researchers there from both of the labs. People from the actual labs. >> I think I know what dinner this was because I was invited to it. I couldn't go. But keep going. >> Yeah, it was a good dinner. You should have been there. One of the guys from one of the labs, I won't say which one just to be nice, framed the beginning of the dinner around this question of who at this table believes there will be only one company left.

18:19 >> Who raised their hand? What percent of the dinner table raised their hand? >> Only the people at the labs. Oh, both at both labs. >> Yeah. Okay. So, yes, there are it's religious, but it's the thing. Here's the flip side. Now, we have the opposite thing in the story, too, which is actually the labs are creating infinite liability and the best case is that they're zero, right? The infinite liability, which is kind of the flip side, which is this is like P Doom people, >> negative infinity, >> which is like it's negative infinity, right? It's the same story. And the problem is like you're like, and this is the thing I actually have a real issue.

18:54 I've been trying to think a lot about what the right answer is. Although we know the right answer will never happen, which is like, look, if you're like, oh no, it's not that it's going to create infinite value and only be one good company. Actually, we're all going to die, right? Infinite negative. >> That's like actual PDM. >> Then you should spend an infinite amount of resources to stop the end of the world, which is also a useless statement in terms of like how you address the world. And so I actually think the real story of now is we have narrative of wait, which one is that? I'm trying to think. the P Doom story. It's like if there's a 10% or even a 1% chance of cataclysmic everyone dies, then in theory you should spend all of the money in the world to stop it. But what's interesting, what's coming up for me in this is more again on the religious side of this infinity argument, which is that they're both religious. It's Christ and Antichrist. There's another one though, Sam, which is there one of the labs, I'll let everyone guess which one, believes that if you don't bring about the machine god, then the machine god once it is created will go back in time and put you into a simulation of hell.

20:00 Like this is a real belief. But this is my point is like this is the religious zero in infinity infestation of technology. Yes. And like the thing is it's causing a bunch of really bad decision-m because here's the thing. I'll give you an example. Like let's talk about practical land. Forget Cataclysm, but like there's a real chance that we will have lab leaks as we've already had with Hubbing P. You have to call them lab leaks. It's like Wuhan lab leaks, right?

20:26 That's what it is. >> But the Russia one this week, too. Mhm. They'll keep going. >> The Russia one's fake news. It turns out >> the lab leaks are going to cause trillions of dollars of practical damage. like Katrina size catastrophes. They're going to shut down Bank of America and freeze up the US whatever and lots of people are going to die or hack everybody's secrets, >> all these things, right? And like that's actually as wild as it is fairly practical. The funny thing about that is we could do something about that. Like it's not that hard to think about what the US liability framework could be.

20:54 Like think about post08 where you had to have cash reserves whether it's you have to carry a lot of insurance cuz if you're doing lab research and things might get out but you don't have the balance sheet to pay for it you can't dump that risk on the US government or the US people like you should pay for that right so because we're so obsessed with the tales of zero and infinity and the religion we're missing actually good decisions in business about how to operate in a new technological landscape and like I think that's the thing is like to me is like religion is super fascinating >> or are we just human and trying to figure it out. I mean, there's like this interesting thing which is that you could look at it and be like, well, we're not taking this seriously enough, but we are being kind of human about not releasing these models, giving the top companies in the world the best models to secure themselves from these Katrina scale disasters. Sam, we are kind of being human about this. I don't know.

21:45 It's just a weird world cuz like things are so consolidated in terms of the decision-m on this and like the people who are are like hyper religious, right? Like >> yeah, >> it's very imamike. >> Doesn't this bring us to why I mean the news this week that I mentioned in the agenda is like more of these personal agents, but many of them private running on device running open models. >> I love you Brit, but none of these work and nobody's using them yet. personal agent protocol that Meta, Walmart, Instinct, Shopify, Sierra, Stripe, everyone else is trying to publish a standard for how AI bots transact and buy things for you so that you are protected.

22:27 >> No, they're not. That standard is not about protecting anyone. That standard's about making commerce happen. >> Yeah. But like having agent identity and like primitives for commerce, I think make a lot of sense, right? Because there's a audit trail for instance of what the agent is doing every step of the way. I don't know. I mean, like, look, this feels like to me we're back to like more fundamental openlo style things like conversations, right? And here's the thing. The internet was born without an identity layer so that anything could interact with anything, right? That was both its biggest weakness. It's why we have spam and risk and d it's also why it grew.

22:58 >> We're never going to get out of talking about identity, are we, Sam? >> Well, it is all an ident. We've been talking about the same problem for like 20 years, right? And like it's the same problem, just more scale. And here's the interesting like I still don't get why anyone gives about MCP. Like why does anyone care about this? Like it's totally irrelevant. It's like >> explain explain why you think it's irrelevant >> because everything it's just an API.

23:19 These things can read APIs and documentation. It doesn't matter whether it's MCP or not. It's just ways for machines to talk to machines, right? And so for me it's when people like oh we have new standards. It's actually a very simple thing. It's like either your machine or service you have has an API in some way, shape or form which Dave was pointing out a year ago. Even if you don't publish it, I can reverse engineer it now. Right? either it is accessible to other machines in which case it will be used by other machines or it is not in which case you can have something that is protected and so the entire thing like these standards and protocols is all What is interesting though is like I don't know about you guys, I've already noticed for news for instance, Facebook's thing that certain workflows that used to work have already been turned off, right? Because service like >> like what what's an example?

24:04 >> Well, like I tried to book a hotel last minute last night and like it got in like a doom loop on like Hilton's website, which I am sure, right, is just because people are already ratcheting up their anti-bot stuff cuz like we need to have a relationship with the customer. Like we actually don't want a bunch of bots running around booking this stuff now that it's practical, now that it's a real concern. >> I think that's what this standard is about, Sam, >> is to say that like we want to leave it open effectively.

24:28 >> No, I think it's a Yeah, it's a standard designed to get people to talk about this problem, right? Because I think you're right. People are people are throwing captions everywhere. Like I saw a bunch this morning on things that >> But people don't the thing I don't understand about like the the standard doesn't change anything. Like it's either simple. It's like as a business you're either going to be able to say, "Well, I literally cannot stop people from crawling me or taking my That's what happened with like the textual internet and why we even have AI is they basically just like Carmen San Diego stole the entire internet to bootstrap these things, right?" Like, >> so what you're saying, Sam, is your Muse agent tried to book a room for you on the Hilton site, which it used to probably do easily, and now it's getting in these loops because Hilton has done something to like block the agents.

25:10 >> Yeah, it's Cloudflare. I mean it's basically businesses understanding correctly that if you get intermediated and to become yet another way to clear a transaction at the back end the market becomes too efficient you'll have no relationship with the customers and your business is worthless >> which is what Amazon did too right to muse as well >> which is correct if you have like enough power to do it >> do you think Hilton has enough power to do that >> I don't know I mean this is >> for sure they do I mean I I'll tell you that I stopped when I eventually after the third retry stopped muse and just went and booked it myself cuz I was like clearly not going to like it was just spinning.

25:42 >> Well, a lot of the ticketing sites have started doing this too. I've noticed cuz I'm I buy a lot of tickets for things. >> This is like the downfall of open graph. This is the same story over and over and over, right? Which is like the technology is not a when you talk about all these agents spinning up. We don't have a technical problem. We can send bots on the internet to do lots of commerce for us now, right? Like everyone can do it. There's no I mean and Muse might be slightly better, but like it's all the same stuff. The question is now business and politics.

26:08 You either have a set of people are like, "Look, I need volume at all cost and by the way, maybe I'm a hired gun CEO, so I only care about the next three years anyway, and then I'm gonna retire, so it. Let's like let everyone rip on our website and book whatever they want." Or you have people who are like stewards of capital and companies. You're like, like I kind of need that relationship to the customer myself, right? Like I can't be part of a perfect market.

26:30 >> I mean, imagine you're the CEO of Hilton. What is your take right now? Is it just is it like open to the agents all day long? like how do you how should we think about this? It's a really interesting question and what's also interesting question is what they're allowed to do with other hotelers because in one sense if they're all together like no we're just not going to serve as bots because like we need the human relationship that's like collusion if that makes sense right I don't know how you think about that but on the other hand it's a really interesting collective action problem because and I think that's the interesting question is like will the tech companies and bots exploit collective action for volume in the short term but long- term term business efficiency and no value or will the industries figure out a way to work together and push back >> but isn't this part of why maybe this personal agent protocol is coming together so I've been seeing Dave and Josh Elman even was talking about how this is very similar to Facebook connect you know the I and like that if you have authorized your identity behind these agents you are a real person like you should be able to get into walmart.com and buy whatever you want through your agent >> oh I I thought you were going to say something different, which I have a half-written to say on actually the the fact that Facebook connect ultimately didn't work and open graph failed and had to be shut down is the same story where like there's not enough value trade for the brands.

27:52 >> Why wasn't there enough value trade for Facebook connect? You brought your whole social graph along with you. The brands didn't find that valuable. >> Like all the thing that was valuable in early in Dave's era of platform was feed distribution. It was a trade. I give you distribution, you let me do this and give me data. And that was a trade that worked. Every iteration of after that leading up to when I ultimately had to be silly one who had to say fine, we're done. And really shut things down a lot on platform. The problem was you couldn't find a value exchange that worked. It was supposed to be like, oh, you give us data. There was never a clear business case for why Facebook connect worked. But for retailers and agents, commerce agents, wouldn't it be that the data is already starting to show, at least in the reports I'm reading, that like people are spending more, transaction time is faster, like all the e-commerce metrics you would want to see, time to check out, all of that stuff is through the roof.

28:46 >> Short-term, but like at what cost? I mean, the problem is like short-term, true, potentially. I literally I booked a car service on used the other day. It worked fabulously. So, like for sure it'll drive volume shortterm. The question is that in the mid to long term, if we actually live in a world of agentic efficiency, no one's gonna make any money. >> Well, it just means that your brand has to be stronger than ever before.

29:07 >> No, you do. You have no brand anymore because like it doesn't matter. >> Well, if you have a commoditized product line, >> well, like hotels, like the Hilton for instance, this is my my example last night. I went to Muse and said, "Hey, I need a hotel room. I have an early morning flight out of LAX. I don't care." >> Well, you don't care, but a lot of people are hotel snobs. I don't know. Are they This is the question, right? Like for me, that's becomes a pretty efficient market. Now, it didn't work, right? And I had to then do it myself and then I all of a sudden did care.

29:33 >> You should have used Instinct. Instinct is the travel agent as people are calling it right now. >> Maybe. I don't know. It's just an interesting It's an interesting era. >> Yeah. I don't know. It's like I agree with you that the business incentives are the hard part, but like we're describing this as a fight between Hilton and an AI company over who controls the customer relationship. Where's the agency in that? There's a difference between another platform inserting itself into my life and me sending a tool that works for me. And they're not necessarily the same thing.

30:02 If an agent answers to somebody else's advertising or commercial incentives, your critique, I guess, applies, but it connects to this thing that I was thinking about at Eselin this weekend. Human potential isn't just about making people more productive. It's about expanding their ability to act on their own intentions. And I don't know, like I kind of think about building agents versus building agency, which was like a thing that came up this morning in an openclaw meeting that I was in. It's like, are we building a new intermediary that controls everything?

30:34 Are we giving the individual more power? And I think this connects to what you're talking about, Brit, or I guess trying to bring up around some of these more personal agents that are starting to pop out. And to the earlier conversation, we're talking about these extraordinary valuations. We believe AI has this extraordinary potential, but like whose potential are we actually underwriting? I don't know. Like I think that if the user has extraordinary power and has extraordinary agency and they're able to send one of these things on their behalf to do what they want to do, they should be able to do that. But the companies and all their incentives getting all mixed up in all of this, I guess, are like this giant messy smoke screen around what people are actually trying to figure out in all of this. I like the human agency thing. Like I love this stuff as a user. I use it so aggressively. I'm like really happy with what I have my infrastructure for lots of self-actualization. So I'm in I just think it's like it feels very much to me like the old argument about who owns my data. I remember that like 10 15 years ago though everyone was obsessed with well I own my data so it should be on an API and if I want to take it with me I should take it with me and then all was like no privacy and security like you should not be able to do that it's not your data and then Europe decided we're going to have this ridiculous download your data thing which download your data is fine download your data from Europe is like the saving grace because if you didn't have that if I make eur a lot of fun with Europe and the right to be forgotten is the dumbest law ever written But the download your data forcing, right, which is like you get to get caught like is pretty critical. Like you'd be very sad. We'd be all be very sad and locked into things we don't like if it weren't for that, right? Back door out. And so it's just an interesting I feel like now we're in the inverted thing where like on one hand if I'm a company like the whole value is the relationship with the customer. It is the brand. It is the trust. It is the loyalty. The second you become a commodity in a market you're worthless, right? That is the zero case.

32:28 >> I know. But I just don't think that companies have a choice. Well, but I think this is the thing that's playing out is that these companies have very different incentives than truly giving you maximum personal agency. That's correct. Their goal is money. What does agency mean? It means you have the ability to act on your own intentions to make something, understand something, to change your life. I don't know. I think Sam, to your point, like I think people feel that these agents are extremely powerful and that they can maybe do things with their potential that they couldn't do before, but often the incentives of these companies because they're offering them for rent and based on their own incentives, I mean, very simply, it's like getting in the way of your ability just to book a hotel, let alone like, I don't know, try to do something far more powerful with your life, right? And so like maybe that's the that's the breaking line.

33:22 >> This is the ultimate wild part of capitalism, right? Is that it's such a good system because it encourages people to push the edge and invent new things and meet the market. But the whole point is that it tends towards zero profit which is why you have to keep doing new stuff to meet the market and kind of the edge, right? Like that's the fundamental. It's so cool, right? But that is like the structure of it. And the problem is is if you collapse relationships to zero and everything becomes commoditized, then Hilton has to figure out how to innovate or they can't afford to or they can't afford to provide their service.

33:54 >> This happened in the social media era. Do you remember when all the brands were holding back from like starting an Instagram profile or a YouTube page or something because they like were like no people will come to our site directly. We will not have an intermediary through our social media account. >> But I mean that was a pretty messy era Britt. That was pretty messy. I mean, entire swaths of the media ecosystem went out of business.

34:15 >> I know. I was there. >> If Instagram never existed, the companies that were in the lead would actually be better off than they are today, >> right? They were all intermediate. Like everyone was in media companies, retailers, everyone got disintermediated. >> But the reason was really interesting, I think, which is not because they didn't see the game that was being played and the reality of it. It was a collective action problem, which is that they couldn't gang up and together like hold the line of being like no. And in the case of Instagram, if they did, a thousand new people would come in and fill the hole. That's a little harder in the hotel business.

34:46 >> They also couldn't say no to going where the audience was. And the audience was on social media and Google was like putting sponsored ads in front of all their search results. So people weren't coming to the website anyways. So now agents are going to go to the websites and they can't refuse the demand of these people that want to buy stuff from them. I don't think. >> No, Brit, you're just wrong. They already are. Some of them are I'm not I'm giving you that I think Hilton and Amazon but like my bet is that over time majority of these brands fold and let agents buy stuff from them, book stuff from them. Like McKenzie is saying this is a 3 to5 trillion dollar market by 2030.

35:25 >> They've never been wrong. >> Well, TRBTs already has like 50 million shopping queries per day. Like shopping and booking is a huge part of what's happening. >> But on the other hand, it's like here's the thing. It's just an interesting business problem because like what chat GBT originally what chat GBBT originally said is oh put yourself inside us you know what I mean like we'll effectively literally front end you or give us specific APIs that will let us do this to make it easier companies have been trying this forever right which is like hey we should be partners give us the APIs we'll sit on top >> even Tik Tok's doing that now Tik Tok you can like oneclick buy stuff now in Tik Tok >> the only change now is that like you don't only you don't need custom APIs and not only that but you can't tell what's a human and that's a bot. So like basically if you want a human to be able to shop at the thing, it's very hard to stop the bots. Although it is doable or like there's going to be a cat and mouse game around that for the next 10 years.

36:15 >> But I think that's what this protocol again is trying to do is to say that there is a human and an agent. This is a real thing, not a spoof. >> I I don't know. I I think it's going to be really We'll see how it all plays out and who wins in it. I think what the brands would probably hold out for what they'd rather have is to be able to like just do business development deals, right? They're like Amazon. It's not that Amazon isn't going to do anything in commerce with these things. What they're going to do is they're going to ransom it. Be like, "Okay, who's going to be like we we'll we have a lot of things that are hard to get at other places. So, like let's do some business here. Let's come up with like the deal.

36:48 >> All the agents are going to do it. All the harnesses cuz that's really fueling the business model. >> It's going to be no different than the Google Apple deal for search. Like, it's just going to be massive BD deals. Like, that's how it that's how this ends up. >> But each one will take their part of the bag." So >> well and the funny thing is is with the agents I wonder if you can run longtail BD so much faster. Part of the problem with BD is just like timeconuming like Amazon's clearly it's worth doing a business deal with Amazon. Do you think in the future, Dave, like these micro agents basically go out and negotiate deals more dynamically using like AI like at scale?

37:22 >> H >> cuz I think that's you could imagine Hilton >> maybe >> Hilton can do a small number of BD deals by themselves with humans. But if all of a sudden you have a gajillion bots running around services, you know, Instinct 6.0 and 7.0 and 9.0, they can't support them all. So historically like, well, we just want to do deals with you. Now it could be, oh yeah, yeah, Instinct 9.0, we'll do a deal with you. Like, let your bot talk to my bot. We'll hammer out the terms. They negotiate with each other. You actually can have a whole different business ecosystem of negotiated deals and treaties effectively that you could never have managed before.

37:54 >> I like this. Agent treaties. I think they're treaties. Maybe that's another branding we can do, which is we'll call them agent treaties. >> Yeah, we're going to start a treaty framework. Screw this standard stuff. >> I like it. It's like I think that the underworld of all the developers and VCs should do that. >> Standards were so 2010s and 2000s. We're moving to treaties. Standards exist in a world where you need standards because you can't do a bunch of one-off deals.

38:21 It's too complicated, right? It's too much work. So, you have the standard and you write to the standard, right? Fine. We don't need that anymore. Now, like with AI, everything can be a pair treaty negotiated truses, right? And so everything becomes the agent treaty framework, >> but only the agents are negotiating. So like Amazon's agent and Muse's agent will figure it out. >> It's kind of what's happening across all domains. Like >> it's kind of like modems, you know what I mean? Like allows to connect you like basically be like, "Look, I love this.

38:54 This is a thing. We should push this cuz it's like that's right. >> I don't see any commercial value in this though for us, Sam. Is this just like an ego play?" >> Well, okay. If we really wanted to make it into a business, a money-making scheme, you come up with, as Dave said, the Asian treaty framework, and then you basically would want some sort of like treaty platform, right? We're like, okay, we're going to sign a treaty and like you have treaty lawyers, bots.

39:20 We're going to have like a meeting place where you can put up, okay, I'm willing to do treaties. >> I think Treaty is just a good brand. I could see a character for it already. I think since now we're all just using characters as agents, like we should just own the treaty bot. >> Yeah, but we're just like there are no standards anymore, only treaties. >> What's interesting is a very large open source project that I may or may not be involved with. We have so many inbound PRs now that we literally just can't deal with it anymore. And so this is breaking in open source. It's breaking in bots. The whole like 2010s, 2000s way of thinking about this just is fundamentally not working anymore. So, what do you do about that, Dave? How do you fix that?

40:00 >> We develop the the treaty framework. Like, that's what I'm saying. >> The ATF is a thing. I love this. We got to spend time on that. >> The Asian framework. >> That's our next topic. Well, we Okay, I'm glad we got that covered. >> No, this is very good. This is why we do this podcast, by the way. It's not for the listeners. It's so that we can iterate on ideas. >> No, it's for us to come up with new ideas, for sure. Have you heard about this thing called Megalith? Michael Romano. Do you guys know Michael Romano?

40:26 It's a new type of fund vehicle that buys pro rata rights from seed funds that they often can't or won't exercise. >> I mean that's every seed fund. >> I think so too. >> Why didn't anybody else do this? >> They have no dry these funds have no dry powder left. Let me get more context. >> You can buy all of mine. I have infinity. >> They have prorer rightites but they have more no more dry powder. There's no follow on like mandate. And maybe they're just not organized for it.

40:50 >> No, it has nothing to do with not being organized for it. everything to do with that the LP ecosystem doesn't want to buy this right now. >> Some of them it depends but yes, >> here's the thing I find funny about this. So when Slow Ventures does seed deals, we always put Prouter rights in >> as does offline ventures. Yeah, >> they're very important. And whenever we do deals, I'm always like to the founders, I'm like, listen, I want to be really clear now and I'd love usually I like ask for written documentation, but I'm always I'm like we care about these.

41:21 This is not like haha you have prora rights. Some people later stage funds are going to tell you later that your seed prora doesn't matter and like we'll just like quash it and like not respect it because they don't need that. They already own plenty and like make the cap table work by like taking out the seed. And I was like listen that's going to happen later because it always does. It happened early in Slow's history with a deal that I did. I'll never forget when a well-known large VC that we thought we were friends with and literally is an LP of ours starts with the letter S.

41:52 >> I remember this clear as day did this to us on a deal. They're like, oh slow doesn't need their their prata. And I'm like, whoa. So very clear that prata matters is critical to the model. The interesting funny thing though about like selling praa rights is X may be slow and offline who take this very very seriously. CDK funds get out of their praa constantly >> all the time. >> It is in some play like we care a lot about this and make a big point of it and will never ever fold on it but like lots of funds do and so it's kind of funny to package that because I think a lot of people look at like early stage pratar rightites is like not worth the paper they're written. Well, I think to your point, Sam, I've always just assumed now ever since that experience, and I've seen it many times over now, that the series A funds, the big mega series A funds will just take away PRAA from everyone. They renegotiate the contract every single time. It's not clear. And the only thing that saves you is a very good relationship with the founder.

42:53 >> Well, or like the willingness to go absolutely nuclear. >> Yeah, totally. >> Because you can't sue people over it. >> Yeah. Nobody does though. Well, I've gotten close. Yeah, fair. I'm willing for for all those founders out there listening. I will sue you over my pro router rights, just to be clear. >> I mean, it's real money a lot of the time. >> Well, it's usually what actually makes the deal work. Like, that's the irony of a lot of things is like a seed fund without any follow on. The math, especially in these capital consumptive days, unless you're in like a crypto project, really is pretty damn hard if you can't follow on.

43:25 >> For sure. Truly concentrate. Well, the thing is that the megalith fund is letting the GPS have the bulk of the carry still. They're taking like a 1.25% slice or something of the carry and then taking the investment. >> That makes sense cuz why would you do this otherwise? Like you have to incentivize the GP to do it. >> Does that change your mind? Would if you had a fund that was out of capital, you couldn't follow on anymore. No parada, right? You still have PRA rights though.

43:55 You're still, I think, kind of buying market beta, like you're buying private market beta doing this because the fund's goal is to be very good at choosing the right ones to extract the alpha out of. And so, you're kind of getting it's like second pickings proa, they should just name this fund second pickings capital. If I'm an LP in a fund and you use my capital to find the thing, I kind of want you to offer it to me as an SPV before you give it to an entirely different investor stack.

44:21 >> That's what I was going to say. I think something like 64% of LPs now want co-invest rights. >> That's all they ever want. >> Well, I know 64. There's actual number now around it. >> I feel like half LPs in the world only do early stage investing for the option on possibly getting like co-invest later. >> And it makes sense if you do the math, but also they all want like no fee, tiny carry. It's the terms of these co-invest deals less advantageous than they used to be.

44:49 >> There is an interesting dynamic at play. It's not like core Silicon Valley, but I don't know if you guys have seen this, but I think it's kind of interesting. Like I think there's like two types of LPs in the world. There's the type of LPs in the world that are excited when the people the funds they've invested in make them money and the fund managers make money. Like that's good. Like that's the goal. I think that's like classic Silicon Valley is like we all want to win. There is this like I think other class of LPs though that maybe come from a bit of further a field or more from the finance world that seem like personally affronted if GPS make money. They want to make the money, right? But they're like, "No, we're going to like quash you on fees or whatever cuz like they're like and it's like seems like a it's an insult to them for the GPS to make money, which is an interesting dynamic."

45:31 >> H why? >> Well, at the end of the day, it's just a profit pool. And if you're not playing an iterative game and like then why wouldn't you be like, "Well, I want all the profits." I think that's one of the interesting things about private markets in general, especially in this moment where like all bets are off on so many things, is you know what the private markets really are? Is they're just the no standard. They're the treaty markets.

45:50 They're the standard markets, right? Like everything's just a treaty. >> Are we making treaties with LPS as seed funds? >> Yes, 100%. It's all treaties. It's treaty based. >> Could our agents just be doing this with their agents? >> It's treaties, not standards. That might be like the newsletter theme of the week, right? Which is like, >> what do you think is LPGP treaty of the future, Sam? >> Sorry, now I'm just writing my newsletter. The >> what is the I think I don't know. I think it's like fine.

46:21 >> You think that the current treaty of LPS to GPS, the general contracts that we sign are all fine. >> Sam, you were tweeting about hedge fund letters. What have you learned? >> Oh, hedge fun guys are so smart. >> Do people not think hedge fund guys are smart? Well, I just generally think that people who do venture capital, the average IQ is pretty low, right? That doesn't mean there aren't exceptions, >> right? >> Like us three obviously, >> but like the average venture capitalist I don't think is that smart. I guess that might be true of the average person trading on Wall Street, but the average hedge fund manager, I think, is just like smarter.

46:55 >> Why do you think they're smarter? >> It's very hard. >> They're dealing with more inputs and they're thinking more broadly about the world and balancing more factors. I think you have to qualify them against later stage venture capitalists. Not the early stage venture capitalists have to be like wizards and like artists with some intuition that like you can't even compare to hedge fund guys. >> Look, typically I'm an early stage venture capitalist. I think it's extremely fun. I also don't like the idea of having a book I have to look at every day because I like my relaxation time, right? Like I'm not I'm not envious of the hedge fund people. But you just read their letters and like they're pretty good. I don't normally read them, but like I'm an investor in a bunch of these funds, you know? They they I've just been getting a bunch of quarterly letters and I was like this world is so weird. I'm actually curious what these managers are saying and it's just really interesting. Like I think they have really sophisticated takes.

47:41 >> What's one smart thing that they've said recently? So, I think one of the most interesting points I saw from two managers is these are ones who are way way way up on like AI stocks or AI microns in the world. And I think they're all struggling with like what's the main street application of AI cuz AI is so hated by Main Street and by normal people. And the ones who are bowled up on it, the basic narrative is main street says they hate it, but then if you actually show it to them and use it, they love it. And so like it's there's this whole gap around like perception versus the real the narrative versus the reality of it. And like so and so's small town store has been around for a hundred years, you know, then you show them how to redesign the store with AI and their mind is blown and they're super psyched about it. You know what I mean? It's like there's like a whole narrative about like, okay, what's the real story on how people feel about this stuff and what's like story versus narrative? I thought that was somewhat interesting, etc.

48:37 >> It's funny that you bring that up, Sam. Like I keep hearing this from OpenClaw users. There's a lot of small business OpenClaw users that are are that store that you just described and it reminds me of the early Facebook days when in 2006 or 7 right after we launched Newsfeed there were protests outside of the building but if you actually looked at the usage numbers they were through the roof. So there was like all these protests like newsfeed is evil and I hate it and yet the actual numbers suggested that people loved it and so I think there's like an element of that going on right now. I don't know that it's the total story but there's a there's definitely an element of it.

49:19 >> They might not have loved it. They just might have been like addicted to it for other reasons. >> It wasn't the same as it was I mean it was very different back then. Well, I'm going to get the peptide refrigeration travel kit because I've got a lot of travel coming up, so I'll be set. >> Okay. Well, get ready. New Brit 2.0 4.0 is coming in the next 12 weeks. Wait, Sam, let me say goodbye to the people.

49:40 >> Thank you for listening and viewing and listening to the what's the treaty called? I hope you all >> follow the As agent treaty framework. >> Agent treaty framework. >> Sam and I are working on it. We'll come back to you later. >> Thank you for dealing with our antics. We will have Jessica here again next week to properly moderate. >> Thank god. >> Oh my god. I thought I did a B+. >> No, it has nothing to do with you, Brit.

50:04 It's that without her, the band doesn't work. >> Let's see what happened. Yeah, >> just relationship wise. I want to call out the rules reversed. And I was like, thank God Brit is going to be back next week. I would be like sleeping. I wouldn't even be allowed in the house. >> That's actually not what I meant. It had nothing to do with you, Brit. I meant the band was not complete. I just don't even like I'm like I'm floored. I'm floored.

50:27 >> Sam, he Dave is gonna have to say some nice things to me later today to make up for that one. And everyone just heard about it. >> This is not what I meant at all. Is your love language words? Cuz you could get something else out of it. >> Not what I meant at all. >> Acts of service, actually, Dave. So, I expect a clean kitchen and do something for the kids. >> Dave's never cleaned a kitchen in his life.

50:50 >> Dave >> clean the kitchen every night, man. He knows how to take out the trash. >> Yeah, >> he has like one job. It's fine. >> I will say on kitchens, guys, I will I I I gave Dave I sent him a mayor culpa this week >> cuz I was wrong about something. It was very nice. >> That rarely. It's like once a decade for you. >> Yeah, it was. Well, this was about a decade ago. But I was like really hot and heavy angry about investing in Zipline because it was like completely out of scope for our fund. And Dave was like, "We're doing this." And we did it, but we didn't do enough. But Dave, that's my mayor there. Appreciate it, Sam. Now, my mayor a couple to both of you is I love the mill.

51:26 >> The mill is amazing. >> It's like I was not expecting that device to last in our kitchen, but it's great. I agree for what it's worth. Like, we invested in it and I still was skeptical, but it's truly pretty amazing. I >> think we've got like a 12x on that one, but I I could see it at a 50 soon. So, they've got some deals in the works, so everyone should get a mill. >> I mean, they're unbelievably expensive, so I don't know if you can sell, but like I love it. No, they're going into enterprise and commercial like literally grocery stores and restaurants and all kinds.

51:56 >> Yeah, they're too expensive for anyone other than like like us to buy, but they're great. >> And did you know the founder is Susan Car's son, the designer of the Apple logo and the original Macintosh? >> That's why they're so pretty. They look like an Apple trash can, compost can. All right. Well, that's our plug for the week, everyone. Check out the mill. This was the longest ending of a podcast episode ever, so we're going to wrap it up now. Thanks to everyone for listening, watching. Please subscribe, like, comment, all the things. Let us know what you think about the ATF. And we'll see you back here next week.

52:25 >> I just registered a treatyframework.com. Should we get the dial? >> Another episode of more or less. >> Goodbye. If you enjoyed this show, please leave us a virtual high five by rating it and reviewing it on Apple Podcast, Spotify, YouTube, or wherever you get your podcast. Find more information about each episode in the show notes and follow us on social media by searching for more or less at Dave Morren at le.

53:01 See you guys next time.