LLMs Ate the Search Bar, Now What?

Listening to the Ostrich Report, the marketplace show no one asked for, but somehow you can't stop listening.

Ja, we're the— we are the baig who look at the cross-border chaos, weird retail strategy, and the stuff platforms would rather bury. And we talk about it anyway.

We're part data nerds, part gossip columnists, and entirely unqualified to stay quiet.

If it's got an acronym, an IPO dream, or a billion-dollar valuation built on vibes, We've probably got opinions.

Well, Hendrik definitely will. So buckle up, we're going global, getting nerdy, and maybe ruffling some feathers along the way. This is the Ostrich Report. Welcome, welcome, welcome. It is another episode of the Ostrich Report, the marketplace podcast you didn't know you needed. And you're still trying to figure that out as well, which is fair enough. With me today is my co-host, as always, Hendrik Laubscher from Cape Town, and joining us I think it's from somewhere near London, is Max Sinclair, founder of Azoma. Gentlemen, hello, hello, hello. Hendrik.

Hello, Vinny. Hello, Max. So this week for me has been something that I've been looking forward to for a while. So audience must just know Max and I had a conversation a couple of weeks ago and, it left me thinking about the future. So to kick off the conversation—

Okay, whoa, whoa, whoa, we gotta let Max say hello.

Oh, sorry, sorry, sorry, sorry, sorry, sorry, sorry.

I've also had the pleasure of speaking to Max before. Last time we had a giant gin in our hands, it was in Belfast. And your company has evolved a little since then, Max, maybe, and changed quite a bit. Maybe you could let our listeners know who you are and where you come from, as Cilla Black would have said back in the day.

Yeah, sure. Hi, I'm Max. I spent 6 years at Amazon before founding Azoma. I founded Azoma 3 years ago now. My co-founder is a PhD and postdoc in AI, and I used to work at search in Amazon, and we were looking at the very early GPTs before ChatGPT through the API. And became convinced that search was going to change and that LLMs would be the future of search. And that was kind of the genesis of the company. I think the company has changed. Well, we firstly rebranded. We used to be called eContent. Now we're called Azoma. So maybe that's what you're referring to, Vinny.

Yeah.

But secondly, I think in the very early days when we were kind of selling to customers, we were showing them science papers on calls and being like, hey, look at this science paper from Amazon. This is something you need to be paying attention to if you're on a rank and kind of very few people would listen to us. And I think now in 2025, it's become a very obvious problem for everyone. And as you say, the company has grown. We've 10x revenue in the last 10 months working with the likes of Mars, Arla, Zappos, HP, Colgate. So we've got some very great customers we're very happy to be working with. And yes, it's been an exciting time.

Hendrik said to me in our pre-call, he said that one of the differences with Isoma is that they're working with real customers, which you've just kind of touched on there as well. Just before Hendrik jumps in, he's got the first big question of the day. The ability to look forward when you're receiving so much information. So you being in search on Amazon, your co-founder being a PhD or doing his doctorate in, in GPTs positions you kind of differently. Like I know a lot of companies have emerged in the last 18 months, but you have many years of experience and foresight rather than hindsight into this. And I think that's obviously a huge advantage. So how much further down the line are you looking, I guess is my question. And Did this all happen in the same timeframe you would have expected?

We got investment from Techstars in 2023, and when we were out in San Francisco and eBay Ventures were the LP, and I remember being in the offices with eBay Ventures, senior VPs there who were advising us. And I was saying LLMs is going to be the future marketplace. So that was the initial hypothesis. I will admit, like the whole, like people shopping through ChatGPT was something that we didn't expect. Obviously we're now working on that. Like 3 years ago, we kind of didn't expect that disruption to Google. But we were absolutely certain that like LLMs would be the future of Amazon search, of Walmart search, of eBay search. It was just so obvious to me at that time, given how bad the experience was before the LLMs. And I had firsthand knowledge of that. And I remember saying this to the guys on eBay and they were saying, no, you're wrong. We're never going to disrupt the ads model because obviously if you have an LLM and people ask a question and they get an answer with like a Rufus experience now on Amazon, there's no space for advertising, right? There's no advertising as you scroll through 50 products. So it was a very counterintuitive view then. I think it's a very obvious view now. Did it give us an advantage? Yes, because we've been working with customers for 3 years on this. We have built 2 patterns on how to optimize for LLMs, and I'm sure we'll discuss. So when the big guys started to care about this, and this is when we rebranded from eContent to Zoma about a year ago, stopped focusing on SMBs, started focusing on enterprise because we had a product that we knew generates a lot of value and we're like, well, why shouldn't we go and target the biggest customers in the world? We have the right to do that now. And I think it does give us an advantage because we have like a very— yeah, we've got 3 years of tech. As we said, we've got 20 people now, we've got 2 AI researchers, 10 full-stack developers. We've got something that actually works, which I think is as you were alluding to, kind of unique in this space of let's just vibe code up something in an afternoon and like try and flog it.

Yeah. Hendrik, that seems like a reasonable segue into your question that you were asking me earlier on. Do you want to maybe frame it up for Max? Because I just see it coming from his last comment.

So Max, here's the million-dollar question. Let me just start off by saying we are an eBay-positive podcast in the sense that eBay is very near and dear to Vinny and my heart.

We're nostalgically positive.

And the investor, I, I, I've, I'm very near, like, I like eBay.

I, I, I, I know, I just wanted to make that clear. Um, but here's the, here's the question for me that our conversation asked, and that is that as a brand, there's a million things that you now need to consider. You had to invest in SEO, you need to invest in marketing, you need to invest in, uh, uh, geo and all these other big words. When you have clients, how do you tell them what to focus on? How do you make, help them make sense of what's next? And the reason why I ask you is you are one of the few company founders that were looking at LLMs pre-ChatGPT. So my question remains, how much does this cost and how do you help brands navigate this noise that's currently everywhere?

Yeah, there's a number of questions in there, so I'll try to take them one by one. I think the first question is, what's next? Is that— is that—

yes, yes, yes, yes, yes, yes.

So in my view, we are kind of, we're already in the LLM world, like whether people like it or not, we left kind of like the traditional SEO world. I'd say in like the start of 2025, in the last 10 months have been all the major marketplaces like Walmart, Amazon are employing LLMs in search, whether they're doing that, whether they're doing that via like Rufus or Sparky in kind of like a very obvious chat interface or it's just in the back end of the search, it's there. Like Cosmo, which is Amazon's algorithm, is rolled out everywhere. Walmart, similar. So we're in this LLM space and talking about SEO, before the search bar, Amazon used to work as a— like Ask Jeeves, right? Where you would kind of— it would be almost like a Yellow Pages and you'd go down browser node by browser node. You'd kind of click on home, then you'd go into kitchen, then you'd go into pans, then you'd find something. And this is kind of how the internet worked before Google invented, basically didn't invent the search bar, but caused the adoption of the search bar. There's no right for that kind of keyword-based search to exist. It was certainly better than what came before, which was really kind of long and annoying to click through everything you had to do. You guys probably remember better than me.

Veiled dig.

Yeah. And then, but there's no guarantee that that's the future. And obviously it's not the future. The future is conversational AI and that's where we are today. And I'm looking at your nice microphone, Hendrik. I bought one yesterday and I use Rufus because you're trying to describe the situation, which is I have a desk here with a drawer, so I can't have a clip-on. And if you put in microphone arm into Amazon, it's just hundreds of microphone arms all looking the same. And you say to Rufus, hey, I'm looking for a microphone arm. I need a clamp that is big enough to fit a desk drawer or to go on the leg of a desk. And it understands it brilliantly and it gives you, here's your 3 options. I go, thank you very much. You've just saved me scrolling through hundreds of similar products, which it's hard to see which makes sense. Next, I think we're going to have the browser chat. Atlas just launched from ChatGPT, Perplexity, Comet. So I think very soon in the next 6 months' time, we will all have kind of like agentic browsers. So we'll be having this conversation, we'll have an AI next to us taking notes or helping us to think of talking points or whatever, all sorts of stuff. And while you'll be surfing the internet, you may have that make purchase for you. I think following these browsers will have agents. I don't think agents work quite well yet. Agents, I would define as AI being able to take multi-step tasks independently, but it will definitely, of course, it's going to get there, right? And then you'd say, hey, well, I don't know if you need to talk to it, it'd be on your phone or whatever, and everyone will have like a super personal assistant and it will go and make purchases for you. So if you need to replace a piece in your dishwasher, it will go and do that. That's not something you want to go search for to understand it. It's complicated. You just say, hey, agent, go and do that, and it will kind of understand the setup of your home. I think beyond kind of like the agents, I think AI will start to become physical. So we will have you know, like smart mirrors or whatever, helping us to buy specific skincare for us, smart fridges that will help us to, you know, identify a healthy meal and maybe, you know, put in an order if we're missing some ingredients and have that delivered to us. Of course, this will be delivered by bots and stuff. Like one of my investors invested in a, in like a bot delivery company, which is live across Milton Keynes.. And if you've been to San Francisco, you'll see the robo taxis everywhere. So the future is already here, which is not evenly distributed, as they say. So I think we'll have physical AI, which is in our home appliances and helping us to get 1-second delivery. And then, of course, I do think the virtual reality stuff will— or integrated reality is something to think about. I can imagine a future where we're walking around with the Ray-Bans. Meta Ray-Bans and you walk into a store and it's popping up that there's a discount somewhere and it's like a personalized discount. So that's agents like physical AI and kind of like integrated VR stuff. And that's only what we can predict, right? Of course, there'll be a ton of unpredictable things that are going to happen. Who knows? But we know that unknown things are going to probably happen as well.

Let's take a pause there because there's a lot in what you said, a lot of Minority Report references that I'm taking cues from as well in my head. There's a great author, Kevin Kelly. He wrote The Inevitable. He's a technology anthropologist. I recommend anyone who hasn't read his book really to give it a whirl because you talk a very deliberate language, Max. You say, I predict, or I think these things are very foreseeable with the knowledge that you have. And there's only kind of one direction that it can go. That book certainly provides me with a framework to sit here and believe all those things you're saying. That smart fridge idea is really interesting because in the instance that I saw, the smart fridge not only placed that order for you as well, but it had talked to your toilet, which said Vinny has been living like a Viking for too long, drinking wine, eating red meat, and now needs soy milk. So he's lactose intolerant. So your toilet and your fridge are now in sync and they're trying to ruin your life and make you a lot healthier and live longer. But I certainly see that part of the future manifesting itself and it's just when and the speed of which. You mentioned early on in this discussion that you've got two patents pending.

We've got two patents approved. One is approved. I think one is pending. But yeah, my CTO is more in the details of that. One is definitely approved.

I'd love to hear a little bit about it because in a space where there's so much homogeneity and off-the-shelf purchasing and stuff. I think that's going to be interesting. And then the second part, maybe it feeds into this, but you said that Google search bar caused adoption. There's something about the way that ChatGPT has rolled out globally, every update all at once, nearly everywhere in the world. I think that has helped that adoption curve as well. And it's something they've done better than competitors. How is discovery— and I believe it will— discovery will surpass search as our default method. It's probably happening right now. You maybe see that already, but that's a two-parter. So one, maybe talk us through the patents that you have and are you helping feed into that causation?

So the patents that we have, I think there's now a lot of software platforms which basically use our patented technology and I'm not going to name names, but I think it's going to be very interesting in the future when we see mergers and IPOs in this like GEO space. Because we were literally one of the first companies to do this. We patented the technology. And what we have specifically is we will ping the AI engines hundreds of thousands of times. We collect the response data. From that response data, we see how the AI engines are scraping the internet in terms of web traffic and the response data. And we use that to build a predictive model to understand how people are actually talking to these AI engines. So we're basically simulating user behavior and we can do this on specific topics and personas of how people are conversing with AI engines and therefore what is a brand's visibility in AI for certain personas, etc. So that technology of sending hundreds of thousands of prompts or millions of prompts through the ChatGPT in the front end to build that predictive model. Something we patented. And yeah, there's many AEO, GEO companies now, which, I mean, again, obviously I'm not technical and I haven't gone through the details of what they're doing, but on the face of it, look like they're basically doing our patented technology. And I look forward to kind of seeing what we can do when whoever survives this and whichever companies come out of it, when they start to exit and IPO and all of that stuff.

It seems like your 20 could become 40 with a group of lawyers sitting in behind you. There's a few things there. So you can persona match on existing data. Is that what you're saying?

So we are like, the first patent is helping people to understand how AI engines are talking about their brand. So if you're Mars or Colgate or Arla, any of our customers, right, we help you to basically understand what share of voice does Colgate toothpaste have versus all the competitors and track that over time. The way we do that is basically by collecting data from OpenAI. But we're not just collecting data of random prompts. We're not just kind of saying, hey, we're simulating how people are actually talking to the engines based on how AI is tracking the internet, where they're sourcing the information from, and what the responses are like the bot traffic.

I saw an example. Sorry, I didn't mean to jump in. I saw, I was at a conference a couple of weeks ago and the presenting data was showing where, say, Anthropic pulls the majority of the responses from. I think it was like 40% Reddit and 16% YouTube and so on. That's the kind of mix that you're talking about there, right?

So we show that to customers. We call that citations. We show them what are the most important citations for their brand.

Yeah.

What I'm specifically talking about is like their crawler traffic. How the AI engines are crawling the internet. And if you see what the responses are and their crawler traffic on mass, you can reverse engineer how people are talking about them. Obviously, no one has the data of how people are talking to OpenAI apart from OpenAI. But in a similar way that Semrush gives you keyword data, which they don't have, they're they're building that off Google and extrapolating. We kind of years ago now tried to build that for an LLM and that's something that we patented.

And if I take that then and kind of go back to Hendrik's original 4-part question monologue that he gave us, it's to say who's making use of it? You moved into enterprise space. With very deliberate purpose. The sorts of things you talk about feel like they're shopper insight-led. Who's the buyer of your product?

Yeah, so we, I mean, normally it's a central team, so there'll be like a center of excellence, sometimes there's like specific search teams. So we're very lucky with Mars to be working with the head of search or director of search, some lead in specifically for search. Otherwise it's e-commerce, like a center of excellence team for e-commerce, like a transformation team. So normally our cycles work that, you know, and I'm hiring our first go-to-market person who will start in 2 weeks. But so far we've not really done any outbound at all. It's all been, as you know, like I've been talking about this stuff for years. We'd like have, you know, do stuff on LinkedIn. We have podcasts. Like these and kind of like we generate a lot of inbound from all of that stuff normally from that kind of that type of folk, but they don't own the P&L for rollout. So when we, you know, we've signed like a handful of annual deals now which are six-figure annual deals. When we do that, it's coming from the brand teams, right? It's coming off the brand P&L and therefore we either have to like prove that we're adding revenue during these pilots or that we're reducing costs by the content generation, using these insights to generate content, which not only adds revenue, which is obviously the goal, but also it's a lot given that you were previously doing this maybe offshore with an agency or whatever. We've just automated and reduced your cost like 100x or whatever it is, right, to just AI generate product listings and this kind of stuff. So it's the kind of brand teams who will we also need to build a relationship with and we work very hard to build a relationship with as well. But normally the initial pilot is like a central person.

Hendrik, you must be surprised by that. That's the key answer to the question you asked up top. How do you prioritize this? If this is coming under the brand P&L, this feels like the work will get done, whereas SEO has historically sat under a very technical component within e-commerce. Are you surprised by Matt's answer?

Yes, I am. And, and I think that's what Max has just communicated is the challenge that I'm increasingly seeing and hearing about, i.e., where's the money coming from and who's the decision maker? And that the fact that it's not aligned means that you have brands that are lost. And, and I say that it's with all due respect, but also you have people looking for solutions without actually understanding the value that's going to be generated. What, what Max said about the fact that they need to create, uh, like a value proposition in their pilot just highlights to me again how the P&L makes the decision. It's not because we're doing this for the betterment of the consumer We are, we are doing this because we want to make more money and we want to spend less. And, and that's where this, this conversation for me gets really confusing in the sense of that, and I say this with respect, the people that solve this challenge are people sitting in front of computers and looking at streams of data. This is not the same people talking and learning from marketing. So for me, that lack of alignment basically makes me believe that the wild, wild west that we are currently seeing is exactly why we are seeing it, because there's, there's too many decision makers that are not decision makers having this conversation with Max and his team. Max, just a, just a, before we, before I go on, Yeah, the big, the big, the reason why I call that out is, is that as a consultant, when I speak with brands, the first question I always ask is, where does e-com sit in your business? Is this a marketing function? Is this a, uh, uh, uh, uh, operations? Where does that— because as soon as you know that, then you understand, okay, how the business forms. And the fact remains is that And Vinny, the, the point that I was going to make with— in my long-winded way is that the brands are not made, they're not designed to do this in a manner in which it makes sense because you have millions of people that— you have 3 or 4 people that hold the budget strings, but the people that, that are required to show growth and get the money so that Max can speak to the CFO or whoever. They're not aligned. Max, just before I, before I forget this, when you were talking about the future, one thing struck me is that you mentioned browsers. I'm asking for a friend. Everything that you are mentioning is on a computer. Why are we not seeing these browsers being able to be browsed via mobile device yet. What's the challenge?

Yeah, I, when I, when I, when I talked about agents and like, like Tim Cook in an internal all hands at Apple kind of said, you know, we're never the first, but we're always the best or something to that effect. You know, talking about AI and I think there will be an agent on the phone where you can just take action for you. And as we discussed, I think we're going to start to see Meta Ray-Bans. They've just released a new version where you can have text across one eye. I'm definitely going to purchase it and I can let you guys know how it goes. But I think we will see AI across different devices. I just think the agents are not good enough yet to take action. And that's just the reality, which is they can't take a chain of action well. They can do one task very well. If you ask it to write a blog or email, whatever, it's going to do it really well or summarize this or whatever. It's going to do one task well, but I don't think AI good enough yet from what we've seen to take chains of actions independently. Like, for example, book my flights and my hotel to South Africa or whatever in 2 weeks. It just can't do multi-step problems where there's— but it will definitely get there. It doesn't take a stretch of imagination from where we are now to say we will all be booking via ChatGPT and Booking.com, you know, in a year's time, you just say, hey, ChatGPT, which may act as everyone's super personal assistant agent. And that's where Sam Altman wants to take it. You know, please book my flights, hotel, and Uber for my trip to South Africa. Make it as cheap as possible, but I want to have a gym in my hotel. And it will just go and do everything. So, like, we'll definitely get there. I just don't think the technology is ready yet.

With respect to the previous conversation, the conflict that Hendrik describes, I think brings up an interesting point. If it's brand owners who are now purchasing, they're trying to solve a problem, it feels like, which has been for years level up.

To be fair, it's different in every organization. So I'm like, it's definitely different in every organization, as you say. So I'm just kind of saying on general. That's been my experience.

But proof of ROI in a brand marketing activity versus advertising is a very interesting place to be. And particularly if you're an organization that is 30, 40, 50 years old and you're used to spending tens of millions a year on brand marketing, which may be useful or not, and you don't know for 18 or 24 months into the future if that advert was interesting or not. And that kind of Coincides then with a point you made earlier that there's no room for advertising in all of this. It feels inevitable. We know ChatGPT are trying to hire a head of advertising. How is that going to shoehorn itself into the flow? How ugly is it going to be?

Yeah, this has been my belief for a very long time, which is people are going to expect with answer engines, they're going to expect to ask a question and get an answer. And then with kind of these agentic, they'll expect to ask a request and have that request done. And I don't think they're going— I don't think people have the patience to have ads. And ads have kind of worked with the last technology wave because you kind of put in a keyword and then you look at the results. And if someone's paying on that keyword, the free market kind of makes sense. Like, why would you pay on a keyword that you don't care about? Obviously, whoever's paying on that keyword kind of thinks that they have the right to win that keyword. So it kind of made sense from a market dynamics point of view. But now I think people will not have the patience for it. So if ChatGPT have ads, people go to Perplexity or Claude, as we discussed, or whoever, right? Or Deepseek. I think there are a diversity of models and the winner will be those who have the best customer experience. And the customer experience people will expect is I want your honest and unbiased view and I want to just trust your answer. I don't want to— and the chief business officer of Perplexity has said this on a podcast. He said when you go on DoorDash and you put in best Thai food, you have absolutely no confidence that what comes up is the best Thai food based on DoorDash's algorithm. What it is, is a merchant who's paid for the 'pie food' keyword and that is not what they're trying to build Black City. I, you know, Sam Altman has said at multiple points that he thinks ads is ugly. I think if you're Amazon advertising would just change. They have like Prime Video, they have Alexa, they have other ways that you can do advertising or like, or maybe there'll be something which we haven't even thought about yet, which is like hyper-personalized generated ads that like just don't exist today. Where it will say, we've planned 2 weeks to South Africa. This is with one partner and this is not with a partner or whatever. These guys will give you like 20% off a meal. I don't know, like random stuff like that. And you might go, oh, well, I'll go with Travelpedia. They're going to just give me my free meal as well. So I don't know, but it definitely is not going to look like how it looks like today. And how it looks like today, it didn't look like 20 years ago. Is why would people assume that it stays the same?

There's something that's bothered me for years. We give away our personal data and we do it for free. And now in an agentic world, it's gonna get even more hyper-personalized. Is there a point in the future at which we can sell our data back to the LLMs and reverse the whole advertising flow?

Yeah. I like, this is a really interesting, this is an interesting point. I would say that there's a couple of things here. It doesn't really answer your question, but I'll say, like, first of all, when people talk about AI, we have had, like, it's kind of misleading because we've had AI for ages, just the previous, and the previous AI is a lot more dangerous than the current AI or GenAI. So like ChatGPT generating a poem or email or whatever, it's kind of like pretty harmless. The previous AI of an algorithm, which where like it will, hone in on what triggers you emotionally, just show you more of that stuff like a deterministic AI that is super dangerous. And we're living in the consequences of that now. On the personal data question, like, I kind of, like, I kind of feel like that those kind of algorithms are dangerous. And I would love— or just a side point— the bit like I think there should be the ability to reset your algorithm so you could say, actually, you know what, I've had enough of this rage-baiting stuff I'm seeing. Just clean my algorithm, start me again with new stuff. So I think that should be a point there. On the data side, I think it's kind of a fair trade, in my opinion, where Google says, we're going to give you this amazing platform for free, but in return, we're going to see your data on aggregate and it's going to help us with advertising. I don't think that's a problem. I think it's the former thing I discussed, which is more of the problem. And then on the personal data front, I think there's very little risk of Nike trying to sell you better shoes or I don't know, Spotify trying to find you a new artist that you really like. I think that's kind of fine. Folk in the know. Sorry? Folk in the know. I think it's the government is when the problem is. I don't think governments should, you know, and I never post political stuff on LinkedIn, but I did against digital IDs. I think What we need to be really, really careful here is government centralizing all this data and basically saying, oh, this person, when they were 15, got drunk and had a warning with the police, and therefore we're going to refuse them for university admission, or we're going to deprioritize them on that. That's where I worry of the government holding loads and loads of data on us and just making decisions that we don't agree with, like using health and school and driving or just combining all that data. But I think on private, as long as we can opt out of our algorithms for private companies to try and sell us better holidays, I think that's kind of benign.

I love that you have ethical and moral compasses here, Max, because sometimes you get into conversations like this and it's all just about the sale or the experience. But you've obviously thought quite deeply about this, even being able to frame the discussion as being around the dangers of advertising versus kind of the experience out of it. And it gives me hope as someone on the other side of it with 3 young boys, hoping that the world becomes a little bit better for them. And the digital ID thing, I agree with you on, to be honest with you. I think there is a lot of overreach in there, but I also feel that governments probably now are in a position where they have such such a lack of power and ability to speak to huge parts of the population in ways that previously they were able to. Now, and tech companies have far more impact in terms of the belief structures that people have. So I think there's this fine line, and I'm not into that.

I think that all comes down, in my view, of like, we've had a really dangerous AI algorithm, which is basically you can't opt out of. So if I was Prime Minister, I would say the law in the UK is that basically Facebook, Twitter, TikTok, all of you guys, you have to have a reset button. You have to let people change their own algorithms and give people the ability to say, I'm done seeing this, give me more fitness and health or cats or whatever. Right. I think that is where, to me, government should come in. And I think the— yeah, I don't think it's I think the big risk is what we're going to probably see now, which is like collating lots of data on this. I mean, this is what happens in China. Like, I don't want to be one of those alarmist people being like, the UK is going down the China route. Yeah, but literally you see where this, where this ends up. Like, you see kind of this social credit score and they kind of track where everyone goes and they choose who can socialize with who. And like, this isn't kind of science fiction. This is literally happening right now, just in a different part of the world. And we should not be taking any steps towards this whatsoever, in my opinion, in a democracy.

I didn't have Hendrik having disco music as a ringtone on my bingo card today. It's come at the most inappropriate time. You're saying something really important there. I'm going to have to take him out of the soundtrack, which I think I can, which is okay. But no, look, you raise really valid points. And I think as part of product discussions and industry discussions, either the sort of things that we have to have periodically to make sure that we're talking about the right things. Daniel Hume, he's head of AI at WPP, gave this great discussion a number of weeks ago and he posed a couple of questions. One was around what if the cost of everything went to zero because of AI? How do we exist in that world and how would you want to behave? But it was just the fact that he was able to pose questions like that, that we need to build towards and kind of think through. What he ended on saying, I'm sure this wasn't part of the sponsorship gig at an e-commerce event, but he left it by saying that commerce is one of those bastions where he feels that some of these ethical concerns could be played out and we have ways to manage those experiences and cultivate them and use regulation in a way that you've described, like a simple thing like a reset button or an opt-out. I think we just need to educate the users better on what these sorts of options are for them. And then for the next wave of designers and founders that they have these embedded philosophies as well. Look, guys, we're coming close on time. Max, you've been very kind and given us loads of our time. Hendrik, if you've got no more disco songs, have you got any more questions that you'd like to run by Max before we let him get his day back?

I think just from a commerce standpoint, I think the one question that I have to ask Max is that So currently, if you shop on ChatGPT or all of these companies that have raised funding that are using, you know, chat interfaces, I'm thinking particularly of Daydream and others, the business model is currently is affiliate generation. You know, we'll take a percentage of your transaction fee. Yeah. Is this gonna be the future of commerce or, or how, how, how do these companies make money?

Because yeah, I think so.

Really?

I think so. I mean, look at Stripe. Stripe is a billion-dollar company. Most people use Stripe. We use it. They take a, I think it's like 3.5% rev share or something. It's a lot of money they take relative. I think this is absolutely going to be the future. I think you will want to use the agent that gives you the best answer to my problem of where can I, how do I find a microphone stand that fits on my specific desk? I'm going to upload a photo to the AI agent to just show you, like, this is my setup. I need a microphone stand that's going to work with it. And you're going to want, you know, like, you will go to the AI agent that gives you the best answer to that. My view will be Rufus. I think Rufus will win. But, you know, ChatGPT and Shopify and Walmart's kind of cool and sexy, I guess, as well. And then, yeah, they'll take a referral fee. Amazon already takes a referral fee, right? They take like a like a 15% referral fee, whatever it is. And I think it'll be the same with Shopify and ChatGPT.

It's amazing. You think about PayPal's strategy and it seems kind of all over the place right now, but in the world that you've just described, it makes absolute and perfect sense for them. So the more we laugh at them, the more they laugh at us. Max, you've been incredibly grateful with your time. I only ask that in maybe 6 months you come back on and expand our minds some more. It's been absolutely fascinating. And Hendrik, this surpassed all of my expectations. I don't know about you.

It did. It did for me as well. And to your point about Max coming back, I think 6 months is fantastic because the speed at which these, the sector has evolved, it's mind-blowing. So Max, thank you for your time. It was really interesting to hear your thoughts. You've clearly thought about the sector a lot more than most people that write on LinkedIn about it. But yeah, thanks for your time.

Thank you, amazing Max. Thank you so much. See you guys.

Cheers.