How AI Is Rewiring Marketing with Profound’s James Cadwallader

0:00How do you think about what the internet looks like in five years? I think for the folks or the businesses whose product is content, it could be a very difficult period. Hi everybody, welcome to Inspired, I'm Alexa von Tobel. Today I have on James Cadwallader. He's the co-founder and CEO of Profound, one of the fastest growing new marketing AI companies, helping brands understand and control what AI chatbots like ChatGPT, Pod, and Gemini say about them. As agents become the primary user of the internet and consume 100x, maybe 1000x more content on the internet than humans ever did, marketing teams globally are going to need to create orders of magnitude more marketing.

0:39Before Profound, he built Kyra, working with TikTok's biggest stars to understand the ins and outs of human attention. Since founding Profound in 2024, he's raised four rounds, most recently$180 million rounds at a$1.8 billion valuation. when very clear that content and pictures and videos are created by agents. We've got to be really careful because the second what I create looks and feels like slop is the second people start to tune out. How do we avoid the slop? I'm so thrilled you're building in New York City. So welcome. Thanks, Fab. Let's go back to the beginning. Let's just quickly, in plain English, talk about what Profound's doing today, sort of how you stumbled upon the product, and let's go from the beginning.

1:19Cool. I mean in the most simple terms we're building the vertical AI company for marketing so you know what we're seeing is generative AI birthing big vertical opportunities across everything from legal finance obviously software engineering marketing is a top three line item for every business on the planet and yeah for some reason it's been ignored I think it's because it's somewhat it's something to do with it being like not less verifiable than software engineering or um so yeah i think it's it's more fragmented it's messier as a category yep um but also i just think you know the the smartest people in the world tend to gravitate more towards places like software engineering etc so you know marketing tech has for a long time been uh not the sexiest category it's a bit of a a crowded graveyard yep i feel like martech was pretty out of vogue for venture investing for about a decade too so yeah we've really noticed that as well as like you know in our fundraising cycles it's yeah i think a lot of investors have ptsd around marketing tech quite rightly um but yeah we you know what we've managed to do is build a very technical um very exciting pool of talent.

2:40And yeah, we've managed to build a team of some of the smartest people, at least here in New York, who are dedicating themselves to reinventing the category of marketing with applied AI. I love it. So let's go back a little bit. I think from 2017 to 2022, you were building Kyra, which was a media company really aimed at TikTok. First, just like, let's talk a little bit about that business and how did you go from there? So focusing on, you know, marketing inputs and skills for real people on TikTok to now being truly the fastest growing company in MarTech, um, at the bleeding edge of AI, helping machines figure out how to put in the right marketing inputs to help people.

3:26Um, Chira Media, what was that? I mean, Kyra was really the warm-up. Very grateful for that experience. It was a baptism of fire. It actually started in a very different place. We began in tech, so it was a consumer software business. We built a discovery app in 2014 that helped millions of people find cool things to do in the city. We started in London. We launched it at Paris. It was very successful. we couldn't monetize it well um so we went through like a sort of series of pivots yep it pivoted you know through a desire to monetize a consumer app we started meeting with advertisers which then pushed us towards video and uh that sort of pushed us into kind of going more down the media path but we're trying to use software to kind of enable um what had been a very human and capital intensive business um yeah and i i think you know the business is still going they're doing well um i stepped away in 2022 i come with that business to the u.s and stayed here in in the u.s through covid um and yeah it's really hard you know working with co-founders across time zones is very tricky.

4:49And yeah, I felt at the right time. So I stepped away in 2022. But really, Kyra, I think had I not worked on that business prior to this business, Profound, there's a lot. Profound's been a real rocket ship. Truly. I mean, wow. Yeah, it's been a wild, wild ride so far. And there's a lot of people telling me or us that I'm clever. and I think I would have taken that more on board and you know attributed more I think it'd be it'd be easy right now to start to believe my own bullshit and believe that this is some you know this is some stroke of genius on my part and I think if I were 24 doing this now I would and without that experience of Kyra I think I would maneuver very differently so I'm very grateful for that experience I think it's like it was it taught me a ton of very hard lessons that first business that have allowed me to keep my feet on the ground with this business and stay super paranoid and understand that you know things can get really hard and you know understand that we're very blessed to be in this position and that there's a lot of timing and good luck that has allowed us to get here and that you know when you get presented the opportunities like this by the universe or you could say god yeah um you know you repay those opportunities by just working super hard and meeting the opportunity with uh an appreciation and yeah a recognition for how lucky we are i appreciate that mindset first of all i appreciate the admission that a lot of life is luck and you know i've been lucky i i appreciate your point of it's luck and then meeting the moment with extreme amounts of hard work.

6:36And it sounds like having some things that didn't work as well helped you get to something that is working wildly well. So let's go back to quickly profound. The world is shifting pretty profoundly, dramatically underneath all of us right now. In plain English, AEO and GEO, just for people who maybe don't fully appreciate that. What is that? There was the birth of ChatGPT in 2022, and everything was very generative. It was writing an email to my boss it was you know help me craft this poem etc it was a generative work around 2024 end of 23 early 24 what we saw was this inflection where really is perplexity that pioneered connecting the llm to the web and allowing the llm to synthesize from the web to generate an answer replacing the need to go and browse the web as a human so instead the the tovr is what we saw in 2024 was people shifting away from blue link search and instead just asking an answer engine like a complexity question and then complexity will send agents we could say bots interchangeably there to the web browse the web browse the web and then pull together a synthesized answer with citations telling you where where it went to get the answer brought yep we went all in on that idea with the with the understanding that information retrieval is one of the most important primitives of our species yep um you know it began with sharing ideas around the campfire then went through iterations of like newspapers television the internet itself and you know blue link search dominates the idea of information retrieval or dominates the the category of information retrieval and seeing that disintermediated by bots i don't need to go to the web anymore i can just ask an answer engine i chat your complexity and it will send agents to the web and my behalf is this giant disintermediation and what that there's a second order effect that what that means for marketers globally is hey how is ai talking about my brand products and services and how do i help that and then the next question is what do i do about it yep and then if you start to go that then you can you know this we can tie this all together because there's basically three stages there's oh my god this is a problem because the world is now using Gemini, Grok, ChatGPT, Claude to do everything.

8:58That's where people are getting information from. So how does my business show up and do I show up? And then if you recognize that, then the next question is, okay, well, they'll log into Profound and then we can show these brands, these CMOs, these marketers, this is how ChatGPT talks about your products. This is how Gemini describes your brand this is the difference between claude and grok and this is how it's changing over time and this is why this is where these models are going in order to answer these questions about your brand they're visiting these websites they could be reddit wikipedia youtube your own.com competitors websites comparison sites etc etc etc so we can give you the what's going on and the why and then that leads us to step two which is they say okay that's interesting thanks profound what do i do about it yep my competitor's beating me or you know i'm doing really well here but i want to do it better or you know we're completely invisible in this topic and we don't want to be because this is important this is a new product launch so the answer to all of those questions is just do more marketing basically and do marketing a new kind of marketing marketing that is designed to be consumed by an agent rather than a human yeah it's like marketing for machines you have a great example um it was the showerhead example um where you were trying to buy a showerhead for your apartment can you just like give everybody that example some people get it it's uh it's i'm seeing i we moved into a new apartment around the corner was it the flat iron building um the shower pressure sucked obviously i don't care that much about showerheads so i just asked chat i'm like hey you know i'm looking for a showerhead um the Pressure is no good.

10:41What ChatGPT does in the background is it's saying it's 65 different web searches. Yep. I think I was using thinking mode. I use 65 different web searches to synthesize a selection of shower heads from there. And if you compare that to the human experience, you know, pre-LLMs, the experience might be, I would have been lucky to have visited two websites. Yeah, you're definitely not going to 65. No. So I think that the obvious thing that we're highlighting there is that agents are now consuming this content. And when they consume that content, they look for different things. Our minds are influenced by top of the range, best in class, these kind of platitudes.

11:21Whereas agents, when they're crawling, they're looking for all kinds of granularity. They're looking at weight to pressure, this, this. They'll go into the specifics of granularity and they're indexing on super. Would you say they're doing a far better job than we would do as... A far better job. Great. If I said to chat, research, shower heads, and no offense to you, I believe chat would far exceed your response. I agree. Yeah. We're limited by... I have three kids. I give up pretty quickly. We're very limited by our lack of patience. are the whole idea of SEO was this idea of just ranking in the top five blue legs.

12:03You didn't rank in the top five. You don't matter. You don't matter. Why does the top five matter? What is that? It's because humans get bored and don't want to go past the fifth. Yep. The bottleneck is created by human patients. I want to talk a little bit about a great example. And so your own data shows how real this is. You had a single chat GPT update that pushed IKEA's visibility up by 19 % while sending Travelocity and Priceline down. And I think the point of that is these huge LLMs, the ones we are all living on, they're updating hourly, daily, weekly. And a single shift in one of those platforms can dramatically change visibility for some of these big brands.

12:46And talk a little bit about, and obviously for, you know, IKEA to get a 19 % boost while its competitors dropped quickly, that's a pretty big problem. If you're running revenue and marketing for these big companies, elaborate a little bit of how big that problem can be. Because that's what profound solving. Yeah, each of the models will give you completely different responses. so there's this fragmentation that didn't exist in the world of search or seo um but more than that it's also just the the this whole term geo is plagued by the the fact that it rhymes with seo because the rhyming is kind of where the similarities end to an extent it's all said by the way you're like it sounds the same it's not at all yeah like how chat talks about your brand products and services is frankly just a function of all the marketing that exists on the internet and today chat we use 65 websites to answer the question about a showerhead i genuinely believe in the next three to five years that will be 650 websites maybe 6 500 websites it will just crawl the web as you know Moore's law kicks in and the cost of compute goes down it'll consume everything it'll just consume it'll consume everything and what that and the reality I'm kind of tangent this doesn't answer you whether yeah the model shift around a lot to answer your question but and you need to understand what's going hence the analytics part of our platform but the bigger yeah that's the tip of the iceberg the bigger issue that this is creating is that But as agents become the primary user of the internet and consume orders of magnitude, 100x, maybe 1 ,000x more content on the internet than humans ever did, marketing teams globally are going to need to create orders of magnitude more marketing.

14:52I mean, you're like, you have to keep up with that consumption. That's correct. And there's two ways to do that. You either hire 100x more team members or you use agents to help you do the work, to leverage the existing team that you have. And that is the premise of Profound. That is the whole idea, that we take those analytics and we take all of the data analytics and we pipe it in and you pull in context becomes super important. This is the stage three. Once the CMO agrees, okay, thanks Profound. I know what's going on and I know why. And I agree that I'm going to create more marketing using your platform, using Profound agents.

15:35The next big concern is, well, how do I ensure it's not slow? Yeah. How do you make sure it's not a bunch of garbage? Before we go there, because I want to come back there, I want to take a minute and just ask a question, which is your team at Profound, which is almost 400 people. And obviously your growth has been astronomical from a 3.5 million seed to now, you know, multi-billion dollar valuations recently. um how do you keep up how does your team in let me ask this question really clearly these huge llms are at war with one another to put it bluntly right they are all anthropic and open ai and gemini and everyone is building as fast as they can at a breakneck speed that honestly we haven't seen in tech in my career you live outside of that world because you're not sitting inside any of those companies and you have to keep up with their speed and their changes in their edits because your product thrives by doing so.

16:32How does that work? You know, how do you think about your tech and product teams and how they think in order to be able to keep up at that pace? There's a pragmatic response to that question and then a sort of philosophical. To start with the pragmatic, we collect from different data sources and triangulate. So we survey the models. So, you know, if you ask the models, they give you an answer. So we survey the models at scale. we also connect to our customers own websites to gather first-party data at the CDN level which is super valuable because you're able to see once you connect to a CDN you're able to understand when your pages are being cited via these AI models and then the third is we built a very large consumer panel of real user interactions we spent a ton of capital and time and energy building together this big consumer panel of real user interactions with AI that grows every month.

17:28And think of it like a big market research. Great, and that becomes a powerful asset. Yeah, billions of real user prompts, real user conversations, we collect the prompt and the response. Oh, that's so interesting. That's like almost a new Reddit entirely in some ways that you have internally. Yeah, we don't own the data sources. We license from multiple vendors, like you know the most well-known uh source of uh consumer marketing data is like nielsen they invented the idea of uh you know consumer research basically like years ago or something but now in 2026 there's there's so many of these that exist so we've gone and aggregated all of them into one big we patchwork quilt quilted one big consumer panel of real user interactions with ai um and that's super valuable so those three data sources allow us to stay on top of what's going on with the models and the consumer interactions and how that's changing but maybe to the more philosophical point of like how do our team stay ahead i think that's cultural and i'd say yeah we have you know i always say like violent aggression um i'm like i know what you mean by that exactly yeah we we're you know i want i i try to be the team that you wouldn't want to compete against uh i think about that often i think about you know if I were a competitor looking in, how would I feel?

18:51And I want to make them feel as bad as possible. Terrified. At all times. Yeah, we have an extremely paranoid culture. I don't subscribe to the manifestation thing. So I spend a lot of time talking about outwardly, not in a way that makes people miserable, but I talk about, I've done it all hands with the entire company where I've spoken through blow by blow examples of how we die. and everyone's face goes white and everyone's clearly freaked out. But then I remind people that, hey, this is our destiny to control. I'm going to highlight ways in which we can diet, and that's going to make you as paranoid as me because I don't think this is...

19:36I think people should know. I love that. I subscribe to the same... Truly only the paranoid survive. and to be in a position where you're growing this quickly, the world is evolving so quickly and we have watched companies grow really quickly and overnight also disappear. So I think that culturally, culturally strategy for breakfast. So I think that's a fabulous way to be a leader right now in this year of 2026. Yeah, I feel people when they come in, when they join, when people are interviewing, I'll say to them, hey, you should be abundantly aware that this company is doing very well right now, But there are scenarios where we fly the plane into the side of a cliff and this whole thing goes to ship.

20:22That is very possible. But with the same intellectual honesty, I also believe that there are scenarios here that if we execute on plan, we could build a company as big as Salesforce or Adobe. Yeah. So it's kind of down to us. Which path we choose? Yeah, it's an execution game. Well, James, in such a short period of time, you've had incredible progress. and I was chief digital officer of a Fortune 50 company, so I know just how much it takes to get into those organizations. As of today, you're already a third of the Fortune 100 are your customers, almost 20 % of the Fortune 500. That's a lot of adoption really quickly from very big players.

21:01How'd you do it? I mean, what I want to say is like building the best product. Yeah, I say unapologetically to the team that everything is downstream from great product. I think if you're a CMO of a Fortune 100 brand right now, you understand that this is the biggest platform shift in the history of marketing. And you have a choice. is okay should i rely on my legacy software who's introducing you know ancillary features to try and speak to this platform shift or do i want to partner with profound whose promises that they will keep us at the bleeding edge of one of the most important platform shifts in the history of internet haunting and that that choice that decision for 99 cmos is a no brain up and then yeah we have to reciprocate we have to ensure that we're living up to that promise which is to keep these enterprise brands at the bleeding edge so we are shipping ferociously we every day there's new features launching through our platform we have a super robust post sales function which you know we partner super smart domain experts with these marketing teams to be there as a thought partner and keep abreast of everything that's going on in this constantly evolving landscape.

22:23We also have a forward deployed motion as most good vertical AI companies do these days where we can go and parachute in a marketing engineer into the enterprise who will actually have descriptive takes on this is what we should do here, this is what we should do there. So all of these things, along with that it sounds boring but a lot of enterprise features as well like commissioning and everything they need to get through all of their own internal processes like all of these things start to become really important um when you're working with giant brands and you built all of it we built all of it yeah and it actually takes a long time to build all of that stuff and you want it to be you know watertight um and you know reliable so yeah but there's not one sort of silver bullet response to you know why we're winning the enterprise but It's kind of a combination of all those things.

23:13Well, it's impressive. Really, really impressive. James, I want to shift a little bit this just dead internet theory. I think it's just for everyone listening and watching, it's a really interesting one. And I would love just to know how you would, how do you think about what the internet looks like in five years, knowing that right now we're still as humans getting online, looking at our computers, but rapidly you're watching that shift pretty quickly. What does the internet look like in five years in your mind? For brands and marketers, I don't think this is a big concern. I think for the folks or the businesses whose product is content, it could be a very difficult period.

24:01I think this is existential to publishers whose product is their content. these large language models have been trained on cool content and we're seeing the disintermediation of human attention with that content where I just asked chat to give me the scoop on the latest happening in the world and chat will send an agent to visit one of those news sites or visit you know that there's so much volume of content that you know that an agent can get the answer for you from somewhere for these publishers who rely on a human visitor hitting their web page and then showing them an ad um to fund it uh i think yeah it's gonna be i i've spent a fair amount of time thinking about where this can go and it's it could be very difficult for those publishers only if you're only if your product is content uh so you're saying if your product is content, just to be provocative for a second, the question is, do you even exist?

25:07Is there a place for you at all? Perhaps. But then this is where the dead internet theory comes in because ironically, the large language models or just answer engines, generally speaking, if you want to call it that chat, Gemini, et cetera, they rely on the content in order to answer the question. Of course. So the thing that the example I always use is, say if there were, God forbid, an explosion downstairs on Broadway chat and we asked chat GPT right now and said what's going on on Broadway it doesn't know the answers to that question it doesn't know what's happened on Broadway what it's relying on is a fleshy human being experiencing the explosion downstairs and writing a blog post a news article something about the explosion and then it can go and rag that into its answer so humans play this very important role where we are becoming like APIs between reality and AI.

26:02That's so interesting. Humans are, that's a good one. Humans are becoming an API between reality and this agentic future. Which maybe replace, you want to get a bit Stanley Kubrick with it for a second. And I think that the sort of final destination is once you go into like humanoid robotics, okay, well now, or maybe drones, it doesn't necessarily need to be, you know maybe open ai will have human-eyed robots i mean the most obvious one here is xai having human-eyed robots scattered around the city that can experience that explosion and downstairs on broadway and alert the systems yeah or not even alert they would essentially sort of play the role of the reporter in that sense where they're like hey this is what's going on this is what's happened this is what this person said then the cops came at 11 58 and this and that it no longer requires a human being to act as this conduit between reality and the internet um but yeah in the short term yeah our ability to to translate reality to the internet yep um is very relied on by um these large language models or by the labs to create a good product i want to dig into a few more of your predictions because i think you are in such an incredibly unique perch of being able to see how the future looks.

27:28If you fast forward three to five years, what other things just seem obvious to you that maybe aren't obvious to everybody else? I mean, more in our world, in the world of marketing, a thing that we are very bullish on is that every brand on the planet will have their own eval or their own evals and their own set of evals that essentially determine what good looks like for their marketing and communications etc and ai will utilize those evals so let's say walmart will have an eval for like this is what good walmart brand marketing looks like this is what good a good pdp looks like this is what a good press release looks like this is what a good billboard looks like and walmart will own that eval and that is really the moat of these businesses in the future will be that that that eval becomes very important because in a world where intelligence has evolved to where we think it's going to to get to it will be able to utilize you know you'll pipe in the intelligence like electricity and it will be able to utilize that eval to create great marketing basically great because great in the world of marketing is very subjective and it's moving right it's not only just subjective it's constantly evolving that's correct like the you know the ice bucket challenge yeah did that today if we went into the ice bucket challenge right now it wouldn't go by at all in fact yeah let's not do that um yeah i i appreciate that.

29:04Silly question, but if I was trying, let's say I run a showerhead company, okay? And it's 2030 and I'm trying to sell my showerhead. How do I think differently than I would today? As I said at the beginning, you will create 100x more marketing and you will understand that creating 100x more marketing with AI does not equal slot. Creating great marketing with AI is a skill and it's you know we have thousands of customers using profound agents to create lots of marketing and what we see is not a one and a zero these people use ai to do their marketing and they're winning these people don't and they're losing we see it as more of a one to 100 and the the customers that are in the you know from the 90th the the 100th percentile yep they are doing wonderful things with agents and they're building marketing that is not parity with the marketing they were doing with humans it is beyond what was possible with humans they're creating marketing that wasn't possible um how do we avoid slop to your point and i've been involved with uh you know we have a bunch of companies uh shop my and others that are kind of on the edge of e-commerce where, you know, a lot of people actively are talking about how when it's very clear that content is content and pictures and videos are created by agents, our brains at some point are going to turn off.

30:40And even some of your biggest influencers are like, hey, we've got to be really careful because the second what I create looks and feels like slop is the second people start to tune out. How do we avoid the slop? Slop is a skill issue. Slop is... That's really well said. You don't know how to use AI to build quality marketing. So if you think the solution is to abandon generative AI and do it by hand, I mean, sure, you can take that approach, but just understand that there are very talented people out there that are using AI in way more sophisticated ways. The reason it's slop is probably... That is incredible.

31:17I fully appreciate your point, which is it is not to abandon AI. If it feels like slop, it's because it was misused, it's using improperly. Yeah, you put in a two-line prompt and hope for the best and then wondered why that generic AI created generic marketing. So context is the antidote to slop. Yep. And what we see is, I mean, this really speaks to the vision for the evals that I'm talking about. It's just the most extreme version of context. it's you are trying to in inject your brand voice your values your organizational structure your permissioning like understanding where the human should review where the human should approve like that all of these all of these tenants are what we're seeing companies in profound do now we have context manager which is a dynamic understanding of everything about your brand product services competition like competitive landscape market and all of that is piped into to a harness so that when you type in the one line prompt saying, help me write a blog, it's brilliant for all of that context layer to build a blog that sounds like you.

32:27And actually just to go a step further, what you're really saying is that magic of marketing that we all, it really is IQ and creativity and insight and having an incredible awareness of what's happening around us in the world and finding that one got milk ad that stood out to the point where we all know it. AI can do that and better than us. If we fast forward five years, is marketing going to get crazier? Like if we think about the things that grab attention right now and it is good marketing, it's an ad, it's content, it's an ad, it's uniqueness, it's compelling differences where you're like, oh, wow.

33:12I mean, the Geico Gecko, I think about that all the time. I'm like, well, it's so random, but we all know about it right the most random thing if you if you'd ever asked me to come up with that i wouldn't have um but somebody did and it broke out do you think we'll see more outlandish marketing to be able to in a world of 100x agenda gai content like is will that be necessary to stand out is my question what we're going to see from my yeah and my prediction of what we'll see is a bifurcation that starts to take place where we have like very agentic marketing, everything we've been speaking about.

33:50High volume, high, super fast agentic content that is orchestrated by a marketing engineer who is plumbing in all of this content and doing this new form of marketing. And then at the other end of the barbell, we will see very human marketing that start, that becomes more and more important. We're spending a ton of money on billboards, billboards, events, dinners, all of these things start to matter more in that world. I couldn't agree with you more. That is brilliantly said. You recently announced 180 million Series D at almost$2 billion valuation. Congrats, first of all. You're absolutely killing it.

34:33And as a fellow New Yorker, I'm thrilled for you. What are you most excited about? if we look forward to the next few years, what's getting you out of bed with an extra spring in your step? Working with smart people. Yeah, Profound Labs, the applied AI, the applied research that we're doing now is really exciting. Getting to partner with businesses like Mercore and Fireworks and building the first marketing benchmark and really being able to rethink this category through the lens of abundant intelligence and how we can apply this abundant intelligence to one of the most important industries on the planet um is super fun uh yeah we had a meeting this morning and yeah i think every vertical in industry will birth a a giant new new car yeah i see it absolutely clearly i see you as the next sales force that's very very clear uh and also not not only because you're doing something that's really hard and you're the fastest growing first mover um with already the hardest companies to get but you'll see everything that's on the fringe you'll be a front row seat to seeing what happens around you Yeah, I mean, we're super inspired by Salesforce and Adobe.

36:01They had this platform shift of cloud, and they were able to build giant businesses by that platform shift. I think this platform shift is arguably bigger. I mean, probably indisputably bigger than the shift to cloud. And I think, yeah, we have that opportunity, and it's down to us to execute. I think that I don't want to overglaze Ramp, but I'm very inspired by their ability to attract the best talent in the world to work on that problem I think is the sort of slept on miracle of that business I agree and Kareem did there was like nothing shy of a miracle like how do you attract these IOI IMO geniuses to come and work on that idea versus you know build god at open AI what we are doing is we are attracting that talent into our business yep to work on applying ai to the category of marketing and i think if we can everything is downstream from that talent yep um we have only a few minutes i have a i'm like i have a hundred more questions i want to ask you but i'm gonna make sure we focus on a few um because you talked about something so interesting which was that um the future of uh social platforms become biozones for human data.

37:18And that's a big idea. Say more. So future of social platforms become biozones for human data. The labs are going, in this world where they accidentally, well, in this world where it becomes very difficult for publishers to monetize their content because of that disintermediation. And maybe that doesn't emerge. Obviously, there's lots of deals happening with publishers and labs, but there's also just a type of content that we take for granted that exists on the internet like uh yeah leila's flower blog that she did out of you know she wanted to most people create stuff and push it to the internet for two reasons status and money um and when it's an agent visiting leila's flower blog to get the answer you like there's no benefit status and you don't make any money because advertisers aren't paying to show ads to agents so you know maybe for the fat head like the the new york times the wall street journal you know the big players there could be a path through by doing deals with the labs but this like beautiful abundant long tail of the internet i think might struggle in this new world because what's the point in publishing that website anymore so that all of that human exchange has to go somewhere and i think it you squint you can start to see like okay well maybe that will be the role of social media this biozone we're referring to is that later publish a flower blog on x for example and you know she'll get rewarded by an x if the post goes super viral so she could she has that financial opportunity and there's also just status like people like you know people we crave uh uh praise from our from our fellow humans and that will that i think that the role of social media will be we very important to block bots and agents from accessing that you know from partaking in the conversation it has to stay very human i follow perfectly it actually like becomes this really important outlet where it is leaning towards human validation yeah and protecting that will be important yeah interestingly as usual elon is kind of ahead of the pack in that i think what we might see is this verticalization of social media with the labs you know xai being able to train on access utilize all of that rich human interaction is going to create this flywheel um that yeah i I believe is going to separate them as the, in a scenario where the internet that we know and love starts to die.

39:58Yep. That becomes a very important moat for the likes of XAI. And we see this like verticalization happening all the way through to, you know, from humanoid robots collecting the data, you know, an API between reality and the internet. Then we have this bio zone for humans to be able to exchange ideas and interactions. all of that rolls out to create one super intelligence um again elon is very well positioned to to capture um two last questions uh can you give us just one contrarian take maybe a view that you hold that you think others don't um around kind of where the future is going so a belief that you kind of have or see uh that maybe others don't have at all i will short the idea that the labs will eat all industries and that all knowledge work will be captured by two or three companies once you start working with the enterprise as we do the more time you spend with the enterprise the more preposterous it is there will be one company anthropic that captures all knowledge work across all enterprises it's just it's hard to articulate but it's just not how it works wait i love that that's a great hot take yeah um that's a really good hot take okay so last question is you're on a rocket ship um in a short period of time an incredibly fast growing company uh what's one thing you're doing to just stay sane uh what's one new habit how do you how do you keep james leaving this rocket ship uh it's it's kind of corny but honestly it's my what oh that's a great one to end on i think she she's very uh a newly married in march married she she doesn't work in tech uh she's you know she's supportive and interested in the sense that she should be but like or that you'd like her to be she's she's interested in what i'm doing but she's not you know i i go home and we don't talk about the stuff that you and i've been talking about that's because she's a fabulous interior designer and so she just truly is having her own incredible life she's shatty and wonderful and like i tend not to go home and actually speak about my day much i actually want the opposite like when i get home i love hearing about her day and it's almost like a bit a wet wear reset it's a reset okay cool let's just talk about your stuff and she's yeah she's happy to to not want to like dig into it by the way a plus husband great answer i really love it and you're absolutely right having somebody that truly is not steeped in your world is a really nice reminder that there's a lot going on elsewhere.

42:43James, we're rooting for you in every way. Congrats on everything. Thank you so much for creating so many jobs around this ecosystem. And I can't wait to see where you go. So thank you.