Death of the Search Bar

Welcome to Applied AI Australia. I'm Roand Rodriguez and here's the truth. Ignoring AI is probably the fastest way to fall behind. This show cuts through the noise and delivers only on what matters, growth, margins, and time. Each week, I translate AI into practical outcomes leaders can actually use. Today, I'm joined by Anthony Middlemark. Anthony is a former PWC partner, former chief technology officer with Fujitsu and now serves as an advisor to boards and exec teams on AI strategy, data and customer experience. In 2025, Australian online retail reached approximately $65 billion. Now, AIEL search is expected to influence 20 to 30% of that. So we're talking $19 billion in sales are going to be redistributed across the market. The customer journeys organizations have spent years and millions of dollars building will now need to fundamentally evolve or rapidly lose relevance. Let's get into it. Thanks. It's a very challenging topic. All the rules are changed. The legal rules are changed. how you make money, how you protect a new definition of IP, it's all very very different and a lot of companies are challenged by it. All right, execs, board CEOs out there in 2025, $6.8 billion was spent over the Black Friday, Cyber Monday weekend. the customer journey that you've spent millions and millions of dollars in investing in over the last decade decades, if it doesn't have a significant faceelift, it's going to have to go in the bin. In September this year, I spent $7,000 on getting a decking done for my house through Plexity. I didn't even go through to the website. A few weeks ago, Open AI has now built in capabilities to do that. Anthony, what's your view on how this is going to disrupt organizations, not just in retail, but customer journey all together? Ramon, I think this one's going to be a bit of a big bang. I mean, if we kind of take people back to when Uber launched, you remember that kind of simple bit of differentiation where you could see the driver on a map and it it was kind of a nothing function, but it was so important, right? because it was about I didn't want to move. I didn't want to come out of my house. I I wanted to know when the driver was close and then I'd go straight out to meet them. And it made it convenient. But what it did was it shifted our expectations really quickly. You know, I want to see where the plumber is on the map. I want to know when someone's going to deliver something. I want to see them on the map because I want to be able to plan around. I don't want to sit at my home for the 4 hour window you gave me and wait around. Gen AI search and and answer engines um driven also by AI. They are creating uh a big bang disruption for how consumers want to interact with things that they buy. And it's not it's not like seeing a driver on a map. This is a very very powerful and very very fundamental shift. I'll I'll I'll give you a kind of an explanation of it. Mostly when 3II when people bought stuff they would keyword search. So even if you typed out a sentence in a search engine it would look for the keywords and then it would match. And so understanding relevance based on a keyword you know sometimes it works well. If I if I go brand, if I go Nike, Air Force One, that's fine. But when you when you say things like best home loan, it had to interpret it. Um, and you know, and then you would get a list and you'd have to click through and you'd have to do the research yourself. What what Genai does is it provides curation. It allows you to ask many many questions which curates the experience in the prompt. Um if you're if you're asking oneline prompts uh it'll it'll refine and refine and refine but and if you're asking one paragraph prompts you can probably get a lot closer. um perplexity as you mentioned will say do you mean this do you mean that because their refinement props prompts are actually advertising products for them. Um and uh if and there and the other factor is decision support where you're you're you're trying to buy the the deck but you don't you're not an expert in decking, right? So it goes do you want a wood deck? You know, do you want a cement deck? you know, do you want and and and so it it actually helps you refine the decision and then and then it gives you results back. So, um, as you mentioned in the Black Friday sales, there was a very very significant amount of referral traffic from Gen AI search and um, who who appears in those searches right now is a bit of a black box, but we know that generative engine optimization and answer engine optimization, which is the practice of making your data relevant to AI is a is a is a key thing that you can do to make yourself found by those AI engines, but it's the fastest shift. It's the fastest behavioral shift we've probably almost ever seen. Um, social was pretty quick, too, but this is even faster. And um and and the issue like Uber is that customers are going to have the expectation that you as brands are going to offer this ability to curate my experience or provide decision support. And if you don't do it, if you don't provide it, like you said, you know, if you don't if you don't rebuild your customer, your front-end customer experience, you're going to drive more people into starting their journeys with Gen AI search where they can get the experience that they're going to be demanding. Now, I mean, this started in 2023. It's soon to be 2026. This is what people are going to come to expect. Yeah. Yeah. So, it's fair to say, you know, if we just put this back to basics for execs and boards, guys, if I asked you even now, no matter what your product or service is, would you want to be found on Google, whether regardless of you want more sales, whatever it is, you're going to want to be found on Google, right? So, the same question I'm asking you today, do you want to be found when consumers are looking for your products, goods, or services? Because right now there is a black hole. And if I asked you, what do you appear for on Google? Nine out of 10 of you hopefully would know. But if I asked you the same question for Open AI or ChachiBT or Perplexity, what your visibility is like, most of you aren't going to be able to answer that. So what that represents is a real risk, Anthony, isn't it? It's not like any other tech shift that we've seen. It's obviously been a lot faster, but curation is a whole thing in terms of customers expectations just completely shifting. You know, I I'm really confident that there's going to be a big drop in 2026. What do you think? I I I think you're right. I mean, the the percentage of usage at which, you know, retailers or financial service companies or any company which utilizes the internet to get a customer should be worried. We've already passed that. There are industries which get 20 or 30% already of their traffic from chat GPT. It's the leader. But Gemini is coming on really quick. And when you consider that Gemini and Gemini is integrated across Google apps which are feeding it data to help it understand what your interests are um what you're looking at, what your demography is and a and a whole bunch of of of other implied information. um that ability to curate is going to become so powerful that it it it's just going to be second nature to expect that kind of experience. Basically, if I go to a website to buy something, you know, like let's just use the the sneakers example and I go, I have flat feet. What are the best thing? And and it just doesn't do anything. of course I'm immediately going to go to Perplexity or or or check GPT and and and use that, right? And that's that's the issue because if you delegate the customer experience to those AI platforms, you're going to have to pay to get your own customer back. And I don't I don't think it's going to be like bidding on keywords. I think it's going to be probably a bit more expensive. Um, Complexity's pricing model is already the the highest price you could get for uh a a a click referral from a refinement prompt, right? And that that's just testing the waters. When Yeah. When chat GPT knows that its ability to curate ically increases the conversion rate, they're they're they're going to want some some money for that. So that's this is the this is the environment. And and and to your point, the longer you wait to figure out how you're going to exist in the AI economy, the the the harder and more expensive it's going to be. Yeah. It sort of kills me, you know, that we're still having to convince people that this is happening. It's not just a fringe amount of consumers, customers, users, Australians that are using AI, you know, generated search or AIE search. It is completely mainstream, you know, and it's just going to be a real shock, a rude awakening to businesses one day when they wake up and see that their traffic has dropped off 20 to 30%. And some people are going to start to notice it in their analytics right now. But if you're starting to see, you know, more sales and more products that you can't necessarily attribute back to where they came from, that may be the start of it. But what's the day today? You know, today is we're yeah, December 6. So take this as your warning, guys. We need to get on board for a big year ahead. What sort of expectation do you see from what consumers are expecting today verse what organizations are able to provide? Do you see a gap? I I I to be honest, I think a lot of retailers and financial service companies are totally aware of the stuff we're talking about, but they either don't know how to go about doing it. They're worry about the risk of the investment to make it work. Um uh you know, they they want to react tactically. And in AI tactically, we'll buy new three months. you know that the cycles are so quick. The big the big issue is that the data for these company is siloed, right? So they might have be running six applications, they might be running 56 applications, right? And so the tendency is to get an AI tool and connect it to one or two of that of of of those data sources and then provide an outcome. But the issue with that is like a unsophisticated chatbot. You open the chatbot and you go, "Um, I I want to buy this." And they go, "Well, our operating hours are 8 to5. Come back then." That that's not a chatbot. That's just nothing, right? And and that that frustration pushes people in into the Gen AI tools as well as the as the tactical approach. So yeah, I mean you you you got to decide. I mean the there there's kind of two ends of this. The longer you wait, [clears throat] it's going to get to be a kneejerk about just, you know, retaining volume, right? You're you're you're either going to try to advertise your way out of it, you know, which is going to put you in burndown. you're you're going to have to keep investing to just grow more logs on the fire to just keep the flow coming in. Um or you know you can you know start to look at a path to AI first. I mean I I agree with you. I think by kind of 2028, if you're not architected to be AI first, you're going to be minimally at a distinct disadvantage. Um, it's also going to probably impact your ability to be acquired. Um uh yeah it's it it it it it's going to be rough because this just this basic thing you know it you think it's basic but just this basic curation and decision support capability that these um generative tools provide is such a fundamental is such a fundamental difference. So yeah I I I I think like you said the clock is ticking. Yeah. Yeah, and just to take it a step further as well, Anthony, I think, you know, we're speaking very much uh, you know, customer awareness, consideration, conversions, you know, outcomes. But the thing that's even further to that is I've heard, you know, firsthand experiences and it was actually, I think, one of the major four banks. If a if you don't have visibility what's going on in an LLM, even if they're not talking about your particular brand or service or not being recommended, even if your industry, if it's a niche and it's being particularly uh thrown apart and consumers are starting to build infrastructure of their lives, particularly Gen Zed around what these LLMs is saying and informing them about. So, it's not just about the conversion point. I think it starts way way before that. you know, what is it actually saying about the category before someone actually wants to buy something, they consider it. They need to be aware of it. And there's a whole heap of uh a customer journey that goes through there. And if you're not aware and savvy to what those conversations are looking like, you may be losing the customer before they even think about you, you know. Yeah. Yeah. I I I agree with what you just said. I mean, when when you're just investigating the category, you you could be asking five prompts which completely rule out a a company that thinks they're the right provider of that service, right? Because they simply cannot answer the the questions that you have. So, yeah, I I I agree. And I mean, and and that's actually a really really great point because I'm talking about the point at which you've decided to purchase, right? And you're talking about that whole bit of consideration that comes before and that's really important. Yeah, it's very interesting. If I go back to the decking I was doing, so let's not spend too much time on this, but originally I was going to go and get a pagola. It didn't even end up going down the route of actually considering build building it, but I went through the route of decking because through my research, it looked like it was going to be cheaper and it was going to last for longer and it would look aesthetically better in my garden. So, I didn't even get the opportunity to start looking for pagola builders cuz at the category point in time, I was influenced. That's the power curation. Think about that on scale, you know. So, if I just move on because there's quite a bit here to unpack. Anthony is friction. What do you think about friction between what consumers want and technical limitations of some businesses UX, website platforms? Yeah. Yeah. I mean, um, so there's no such thing as zero friction, right? There's always going to be a little bit friction, right? You always got to do something. But, you know, obviously what what what you're going for here is that AI removes most of the friction. You know, a lot of people like oneline prompt, but as people learn to write more sophisticated prompts, they're going to remove huge amounts of the friction that would have been associated with purchase uh before. And the other thing is with with agents, when I have an agent on my phone and I go, I want to go to Paris, I want to visit the Lou, I want to stay here, I want to do this, I want to do that. Go find all that uh and then book it for me. That's going to be another level of removed friction. And the and and the other thing you're you know in on this topic is if you're a company and you're not prepared to interact with agents. You don't have a channel for agents because they are a custo that's not only going to be friction that the agent's going to hit your site and then just bounce off and move on. And I think hold on just just say say that again. Say that again. So you're saying we're not just talking about customers now that are humans. We're talking about agents are your well they're potential customers. Everyone's potential customers are going to be agents. And they're going to want to negotiate on every single aspect. When how much I I don't want I I interact all the time. I want a discount. I I I don't I don't care how you think about it. I want to discount on this purchase. And if you don't discount so if if you I there was a cost of acquisition for me anyway, right? So to get me to your website, there was cost of acquisition. So if you don't discount and I bounce off, you paid that that cost of acquisition anyway. it's cheaper for you to consider the cost of acquisition and apply that as a discount or partially as a discount so that I don't go away because if I even if I do come back you've now paid a second time right so that's consideration one and consideration two is you are going to have to be very flexible on how intent changes along the value stream so I I I want to buy something I figure it out. Then I get to the transaction. As I said, I want a discount. Then I get to delivery and I'm going to go, "No, you know, I know your I know your company's only a couple Ks away or what whatever whatever it is you're thinking about and I and I want a discount on that or I I buy stuff every month. I want free shipping this time." Those those individual considerations are going to just become how it is. I don't I don't know how quickly this is going to come but for the companies that take this early once again a very very distinct advantage agents are going to shop and look for things in a completely different manner. We may be looking for white pairs of Air Force ones but criteria and what we select filter in and out is going to be different. Right? So, you know, in terms of technical capabilities, what the user experience is like and how things are set up is going to be more and more important. So, it's it's interesting and I'm looking forward to seeing how all this plays out. I think it's going to be next year. When, I don't know, but it's definitely coming for sure. If you think about we're talking about Air Force Ones, if we just continue on the road of big brands, there's obviously a lot of successful brands around, you know, Nike we're talking about at the moment with Air Force Ones, but it can apply to anything. So competitive big successful brands, how do they lean into strategic foresight and ensure that even though they're competitive today and they're making an impact in today's market, how do they ensure that they can futurero themselves for tomorrow for the new economy? I don't know. You know, this discussion of brand is really interesting. I I think brand becomes in its current form way less valuable. I mean, you know, when you learn all TVs are made in the same factory or all all computers are made in the same factory, you know, kind of what are you really buying with a brand? I think brand will come to equal the utility that it provides the customer, right? So, a brand will be considered a good brand if it can answer all the questions, right? Right. And I think, you know, brands that don't answer questions, like if I go to Nike and say, "I want Air Force One," and it can't tell me how, you know, I've got one foot bigger, slightly bigger than the other, and I have flat feet, and, you know, is that the right sneaker for me? If if it if it doesn't answer that question, I think where I discover a new brand that I've never heard of that does answer that question. And it is also recommended by the Gen AI. I will buy that. Right? And I think this is where a lot of startups will really start to gain advantage. Right? Because where they are thinking ahead and and preparing themselves to really figure out what is the ecosystem in which they operate and be able to answer those questions. When you go, what are the best sneakers to play basketball in? It goes Xbrand, you know, Nike and Adidas, and then you've never heard of X-rand, but it's rated number one. Of of course, I'm going to have a look, right? And so, [clears throat] I mean, I do do I think Nike will will catch up? Yeah, I think they will, but they're a big machine and they're slow. And I think that we will see a lot of brands disrupted and it will be based on this concept of of utility, the utility that they that they offer the customer. Yeah. Yeah. And I really do agree with that. And I would probably even go even further to the right that I fundamentally believe that there is going to be decentralization, you know, cuz if you think about what is possible now and what will be possible, what took a team of 10, you can do with a team of seven. So the things that were impossible is now a maybe. And a maybe is is a definite. You know, the forecasts are out. next year at some point in the next 12 months we think we're going to see the world's first man or female worth a billion dollars. So if you think about that in terms of capability and enablement decentralization is going to be happening at a category and industry level. So, it's really interesting what you say like, yeah, Nike is a huge brand, but you know, they're they're slow, but then again, not everyone is Nike. You know, they're are the 1.11% of the world. So, it's interesting. It's interesting. Yeah. And I mean, um, to your point about the one the one person billionaire, the one person company billionaire, um, everyone knows where all the sneakers get produced, right? If you are utilizing AI across a distributed supply chain and you have the ability once again via AI to design sneakers, you definitely can be a very few person company that is scalable, which is the point. the the the the point about having a lot of people was that's what gave you scale. I mean obviously technology as well but but you needed people like they were the human agents which is we're going to see a big a big shift there now in few of your publications and white papers Anthony I've seen intent resolution so I'm really keen to unpack with you what that means to this shift that we're talking about today. Yes. So intent is the term for uh you know the way you express intent to purchase, right? And um and there's explicit intent. I want a bicycle and there's implied intent. I want to get healthy. It doesn't it doesn't tell you exactly what the way is, but it it gives you information about the direction, right? and and and intent changes as you move through the value stream like we were talking about with the agents. Um first, you know, I I want to get healthy. Uh then I decide it's a bicycle, then my intent is to purchase a bicycle. So curation becomes really important and then my intent would be maybe to get the bicycle delivered and if it's built to get the bicycle put together, right? And then and then my intent changes after purchase to how often am I going to ride? Do I need an app to give me a program to do that? And so intent changes and the the whole business model of Google, which for many decades, which is where they made the most money in Google, was about understanding intent, right? and they built their search engine to help people define what intent was and then they would provide the answer and that's how they got paid right and the open AI was after trying to get in on how Google makes all its money because Google's a very very profitable company and they have the approach that we've been discussing for this whole time this genai approach around understanding intent so the difference really is when I express intent in a boolean way, which is how Google works. It's very taxonomical, and I'll just stay with sneakers because it makes sense. So, you know, you go brand Nike, style, Air Force One, colorway, white, size 10, and price, right? That that used to be the consideration way to express intent for sneakers. But as we've been discussing this whole time with Genai search, I don't need to start that way if I don't want to. I can go coolest high tops, I highest performance high tops. And there's basically like between 45 and 50 categories of data for sneakers. And I think there's between 750 and 800 distinct data points, right? And these are the data required to manage all the parameters of intent expression for anyone who wants to buy a sneaker. So, a kid who wants to buy a sneaker, someone who wants style, someone who just wants comfort while they're walking, someone who has flat feet, you know, I think of someone who's crossraining, people who are running, you know, so all the sports, all the styles, all the cool, you know, and and um and when you think about being required to be able to answer all those questions, it's very very challenging. But the value of a business that can answer all those questions is significantly higher than the value of a business that can't. So the resolution of intent is is is like a is like searching for gold basically. the more you resist it, the more it's going to be very hard to continue to build customer volume. And the more you embrace it and learn how to be creative with it, the better it's going to be. Um, and it's very non-trivial. Like this is not an easy thing to do. Um, especially if you look at the context of financial services where they only want the transaction, right? So they arrange the answers to questions or the resolution of intent only around that. But that's likely not what your goal is at all. Like when you go for a home loan, your goal is to have a home somewhere that you like with your family and you know as safety and security and all these other words around intent for a home loan. And basically when you can't match that customer's goal, things boil down then only to price and that's going to be a position that you're going to want to be in. Right? If I put myself in the listener's shoe right now, they're probably thinking there's obviously a problem. We needed to do something about it. But transformation takes time, Anthony, generally. So is this a threeyear thing, a 12-mon thing, a six-month thing? What sort of timeline do organizational leaders need to be putting on this? How quick do they need to act? Yeah, I think there are ways to get going right away. I mean, I think the first thing you have to do is get Gen AI tools in your business and get people to use them and ask people what are you prompting. It's less about the outcome. It's more about how are you thinking about prompting, right? because you want people in the organization to be getting efficiency or getting learning or something valuable from these from these interactions. A lot of people report that they think using genai is cheating so they won't use it. You can't think of it like that, right? And then I think the second step is to do some tactical AI projects so you can learn about prioritization and sequencing of a project. When I talk to companies a lot of time they there's no reason they just go I want AI. There's no they don't ask why enough about the projects that they're choosing. Right? And they also don't do the work to figure out what it really takes to give a really high quality outcome utilizing AI. You know, you've got to look at what you have foundationally and what that's going to take. You know, if your core source of data is an ERP, that's going to be quite problematic, right? And if you have 56 sources of data, that's going to be quite problematic. And then the third thing you can do is go through an exercise of figuring out what does the business look like AI first. Then you can go from stage one to stage two to stage three and that limits your risk. Is it a year? Maybe if you're lucky it's a year. But if you start in 2026 and you build capability and AI fitness, you know, in your organization and and in these decisions and the procurement that you have to make to move from unintegrated to strategic use of AI, that's smarter than crossing your fingers. Yeah. No, I agree with that. I don't think you have a lot of time. I kind of agree. I think you know you're in Australia, you're protected a bit because the level of competition is only the people that are in this country. Although global companies are more and more relevant. I mean especially if you buy online, you don't care where it's coming from now. Used to we don't anymore. Everything comes in 48 hours. But but I think yeah I mean you got to get on to it. you you've been intellectual work now and if you need support, you need to find people who can support you and you need to be very very careful about the deals that you sign and um and how you utilize your data. There may be there may be IP in your data that you give away. There may be IP in your context engineering that you give away. In some cases, you may have to because you're trying to equalize the customer experience of Gen AI. So, you may have to give it away. But in a lot of cases, you won't want to. And very, very shortly, the valuations of companies are going to be determined by your ability to utilize your data. You know, it'll be at first people will just do a little bit of technical due diligence, ask five or six questions. I mean all you have to do is look at an architecture diagram and I can tell whether you're efficiently using your data or not. But in the next, you know, 3 years, these practices of assessing your ability to use data are going to become sophisticated and valuations are going to get discounted based on an inability to utilize data. They're real threats and opportunities that organizations need to be aware of. So if we think about practical outcomes that we sort of led to it, if we were speaking to a CEO today, you know, a chairman of an organization, what are the common things that you think most organizations are doing today? And if they needed to stop or start something right now and put them in the right direction towards being AI first, what would you suggest? How would they get ready for this shift? Number one, data audit. Find out where all your data is. Find out if you even own it because in a lot of cases they don't. Find out if you can extract it because in a lot of cases they they won't give you access to an API to pull your own data out. And then understand all your options in terms of making the data accessible and usable. There are many strategies. There are tactical strategies. There are some strategic strategies that are not end points, but they're really good. And there are some endpoint strategies. The the problem the problem with the most sophisticated strategies is they're potentially risky to to to try to move from A. It's from A to Zed. It's not from A to B because it's a big shift. Um, and and they're expensive, but there's a lot of ways to do it. And what I find, and I think this is the key thing for your listeners, is if you don't understand how to procure it, that's going to be your first mistake. Find people that can explain to you the process and the difference between each of the strategies and the impacts of making decision A, B, or C, and then act. spend a week in a room on a whiteboard and really get yourself across it. And CEOs and boards, you can't delegate this decision often to CTOs because they don't know either. And even if they know the how, they might struggle with the why. If your why is maintain volume at a lower unit cost or increase the value of this company that may not mesh with their how right so it's very important to really really understand and you can't you can't do this exercise and then go back to firefighting or tactical. You have to have some capability in your business pursuit because if you wait till the last minute, even if you're even if you're a company that has a lot of money, you are going to punt a lot of money. The best example, Walmart. Walmart sold, you know, they had huge revenues, but they were never as valuable as Amazon. and to catch up with Amazon to to be considered in the same category as Amazon, they spent billions and billions of dollars and and I was aware of a lot of the uh prototypes that they built which they just built and threw away. And we're not talking about couple million dollar prototypes. We're talking about hundreds of millions of dollars invested in something they thought would work and it and and they and and they just trashed it. You don't want to wait till the last minute, right? Because if you do, then you over capitalize. And I know we're we're at crunch time now. And to your point as well, Anthony, like this isn't a technology problem. This isn't a marketing problem. This isn't a procurement problem. This is a business problem and an opportunity. Yeah. So I'm a firm firm believer with what you just said that this isn't something that the CEO or the organizational leader can just give to someone and then put their leg in and put their leg out because the organizations in the next era which we're stepping into now we are already in this is going to be what your valuation and value is going to look like. So CEOs, they don't need to be developers or software engineers of any of those technicalities, you know, like you've got to be savvy and pull in the right people at the right time. I do speak often on the podcast about empowering your internal subject matter experts and I do firmly believe that. But with some things, if you have the capability, fantastic. But if you don't, you need to just find the right people that do. So, it doesn't matter where you get it from, but just make sure you've got the right people guiding you through this next wave. I 100% agree. And most companies don't have this capability internally. Um, and a lot of companies want to promote people internally to be their chief AI officer or head of AI. If this person is articulating the big objective and you think they have the capabilities to get you there, then great. But if they they don't, I don't I don't know how that helps you. Yeah. So, we've got through a few things here, some pretty important topics. If you're a listener right now, Anthony, what is the one thing you would recommend people to go away after listening to these episodes and get into action straight away? The a really really basic thing I think you can do is if you work at a company that sells something, go to chat GPT and Perplexity and write out every single prompt that you think leads to your product or service and see if you come up. Know your prompts, right? Because if you're writing prompts that you think, "Wow, that that is exactly what I would ask if I were gonna buy this thing," and you don't come up, then go back to your to your website and just adjust your content. So, it does answer that question. This is a very quick, very tactical thing that you can do to just be more relevant. But this practice of knowing your prompts lifts the capability of the whole organization. I would go to the technology people and I would go what prompts are the ones that we need to understand. I would want everyone in the organization understanding what prompt and what prompts do not work at all. And I'd be I'd be fixing that right away. That is a short-term thing considering those three steps that we talked about. unintegrated AI for lifting capability. What's the tactical AI that also lifts capability but helps you start to build a practice around understanding how to utilize and get ROI from AI and then what does the organization look like in the context of AI first? Those are the two big recommendations. Get moving. If there's one thing that you need to act on after listening to this episode, it's the customer journey that you've built will not survive. Maybe some parts of it will be salvageable, but most of it's going to need thinking from the ground up. My name's Ramon Rodriguez. This is Applied AI. Thanks for listening and I'll catch you soon. Now remember, you can't control the speed of change, but you can control your gap. Yeah.