Matías Muchnick y NotCo AI: “Hoy competimos con Anthropic, OpenAI, Microsoft, Palantir y Nvidia”

This is still NotCo or is it simply another company that just happens to have the same name? Look, it's NotCo. This is season 3 or 4. Netflix just renewed the series, and the season is called 'Selling Mayonnaise in Jumbo, NotMilk in Hoods, the JB with Kraft Heinz, to win an AI contract from Palantir for one of the biggest companies in the world.' I can say it was all for something, and it's for where we are today. It would have been very different if AI hadn't had the growth it had in the last two years. Thanks ChatGPT, thanks Anthropic, thanks Claude, thanks Microsoft, and thanks to everyone, man, because you're making my life so much easier. Because if that hadn't happened, maybe NotCo would have been screwed. So thanks, because it brought my soul and body back for two reasons: because telling the whole world 'I told you so' that artificial intelligence was going to revolutionize physical, mental, and scientific work on a structural level. And the second is because I returned to a NotCo in a Series A stage, back to the Mati UDA founder with 10 years on top, starting over for the first time. Today, with artificial intelligence, we are serving more than 30 of the top consumer goods companies in the world because they were convinced that a Palantir, a Microsoft, an OpenAI, and an Anthropic don't have the data moat that NotCo has to make them produce better, more efficient products with a much higher probability of success than in the past. So today, we are the chosen partner of the consumer goods industry to the point that we can access formulas from 100 years ago.

Thanks, Juana Pa, happy to be here.

A ton. Probably more for you than for me, professionally speaking, at least. There are many things we talked about there, your origins, the origin of NotCo, the story you've told a thousand times with Jeff Bezos, etc. Today I don't want to talk about any of that. I want to talk about NotCo AI, what it is today, and also the questions that have been raised about NotCo, which I imagine you know how to answer. So, you were living in Chile at that moment. I think you weren't in Chile at that moment.

In New York. Yes.

It was a pivotal moment of change. I'd say more 2024 than 2023. In 2023, we were coming to understand how the venture capital environment was being managed, which accelerated with everything in 2022. Everyone had this scare: was it a bubble, was it not a bubble? The inflow of billions of dollars to Latin America was unprecedented in the history of humanity. And in 2023, it almost fell to less than 100 million dollars, you know? So everything changed, from the priorities of venture capital to the priorities of startups in this venture capital world. And I think it hit not just NotCo, but all companies with a moment of adjustment: what the hell do we do? What will our company be in the next two years? And for CEOs like us who didn't have prior experience, because it's different to go to an Airbnb or an Uber or a MercadoLibre that lived through the 2000 crisis, the dot-com, another type of crisis, the 2008 subprime crisis. We, or I as a founder, hadn't lived through that. So there was no investor or board advice to change your mindset. For me, a lot of coming back from New York to Chile had to do with that, with reconnecting with reality, with being the first to have to change that mindset that the things that mattered before don't matter today. In fact, the opposite. The things that were once spectacular or cool are now the least cool, and some of the barbarities I did in the past. But that takes time, you know? And you think you adjust and you don't adjust enough, and you think you changed, but you didn't change enough. And no one has to suffer that more than the founders. So it seems to me that these three years since the last podcast have indeed seen a lot of changes, mostly directed towards maturity. And as I was saying, there's no better analogy than life, right? At the end, well, at seven years old, all you want is to play soccer for the rest of your life and you think you'll be able to do it. And then you go to, I don't know, you're in your last year of high school, and all you want is to go out to parties and explore new things. Then you go to college and you're this revolutionary guy who wants to change the world, but with a brutal level of progressivism. And then you start working and you say, 'Wow, you know what? Reality is completely different because what I studied for four years has nothing to do with it, and the things that used to be cool aren't cool anymore.' And look at us, I don't know about you, but if I'm in bed on a Friday at 11 PM, I'm happy. And if the Mati of seven years old saw this, he'd say, 'What's wrong with this guy?' You know? So, nothing, a lot of changes, a lot of maturity, a lot of mistakes, but what I highlight most is the learning and how exciting the world is today, and looking back and saying, 'Wow, luckily we started 10 years ago doing something that seemed completely crazy and logical, and today to say, look, I told you so.' This, this in that line, this is still NotCo or is it simply another company that just happens to have the same name? Look, it's NotCo. This is season 34 of NotCo. Netflix just renewed, and the season is called 'Selling Mayonnaise in Jumbo, NotMilk in Hoods, the JB with Kraft Heinz, to win an AI contract from Palantir for one of the biggest companies in the world.'

As I was telling you, look, at the end of the day, it's not that we changed the pitch of how we presented NotCo at the beginning. It's just that 10 years ago, no one knew what artificial intelligence was, what it could do, how it augmented us, what things we could now, as humans, dispense with doing, the effort, the hours dedicated to certain things. I don't know, making a song, generating an image, editing things, making an Excel, making a business plan. Today, an investor deck, a level one, Anthropic does it for you in 3 seconds with seven data points you gave it. So everything changed. The pace at which it changes is infernal. So I would tell you that what generates the most in humans today is anxiety, right? Everything changes all the time. Anthropic beats OpenAI in a day, and then OpenAI releases something else that beats Anthropic, and you're like in a constant psychosis, and you have to get on that hamster wheel because if not, you're left behind. So what I'm telling you is that my head is obviously not the same, and if it were the same, I shouldn't be where I am. I think three years ago, or two years ago actually, when I went to propose to the board that there are two paths for NotCo: NotCo Foods, let's say the NotCo retail that you know, and NotCo AI. They are two similar paths but very, very different in many things. The team in Foods is not the same one that has to write the AI chapter. The AI panel is very different from the Foods panel. And we, you know, we were too early. We invented, created NotCo, founded NotCo a day before OpenAI. So we were saying AI was the future even before OpenAI, before Elon Musk, before Sam Altman, you know? But it applied to an industry that mattered a lot to us, which was the food industry, the consumer goods industry. So for us, it has been nothing but hyper-reassuring to say, three years ago we started the NotAI path, and today, two or three years later, NotCo has generated the most powerful vertical model in the global consumer goods industry. And there, but when you presented it to the board, wasn't there anyone who questioned you or told you this isn't going anywhere? Or were they already on board? Look, I think NotCo has a special board in the sense that they don't point you to the 1,000 things that can go wrong, but if they see you super convinced with the right arguments, everyone gets excited about the idea that if it goes well, even if it's a 1% probability, if it goes well, wow, this could make NotCo a $10 billion story company again. And that's been the last two years. Today, with artificial intelligence, we are serving more than 30 of the top consumer goods companies in the world because they were convinced that a Palantir, a Microsoft, an OpenAI, an Anthropic don't have the data moat that NotCo has to make them produce better, more efficient products with a much higher probability of success than in the past. So today, we are the chosen partner of the consumer goods industry to the point that we can access formulas from 100 years ago, consumer panels, focus groups, terabytes and terabytes of consumer data, and also understand that there are tools that allow you not just to understand or make a model understand what flavor is, what aroma is, why we like one texture versus another. Because I like things that you, Juan Pablo, even though we're similar, come from similar backgrounds from the same country, you like some things that I won't like. Those differences today, for a human, are incalculable. Then, imagine all the social, cultural patterns that make a seven-year-old in Bulgaria like things diametrically different from a child in China or a child in Germany, you know? So all that information, we can now put into an algorithm. So today, the famous Giuseppe that the market often laughed at, let's be honest, maybe it didn't reach you because it also happened to you that maybe it was like the emperor's new clothes. Do you remember that fable? No, no, no. Well, there was an emperor who had a great parade and hired the best tailors in the world to make him a suit, and they basically scam him and pretend to cut the fabric, but they let him walk around naked, and he goes out naked and the whole town applauds and 'long live the king, long live the emperor' until a child shouts, 'He's naked!' And that's when it's over. So maybe it happened to you a bit too, that they told you, 'Hey, no, they're killing it, man. They're on fire.' To me, the work comes. Generally, I get the nice things startups are doing. There's the occasional gossip who tells you, 'Hey, watch out for the layoffs here' or 'watch out for this.' So maybe I lived in that world, and many times people called me and said, 'No, man, Giuseppe, man, they're stuck on their Giuseppe.' Today you're seeing Giuseppe being revalued, so tell us about that.

That, that you already achieved.

Okay, but basically then today, to understand the structure of NotCo, you have NotCo Food and NotCo AI, and basically NotCo Food is like your Trojan horse, which for a long time burned cash like crazy to get these companies you mentioned to look and say, 'Hey, what is this NotCo? What is this thing they're talking about?' Yes, I think there is a synergy that is too powerful in the sense that NotCo AI wouldn't exist without NotCo, just like Mercado Pago wouldn't exist without MercadoLibre. They are two companies that are a consequence of each other and feed off each other. There are a lot of synergies that, whether you like it or not, even if a consumer business is distant from an AI business, when you start, the most important thing is to validate what you're thinking, right? We validated it with products on the shelves, we validated it by taking market share from competitors. It's being able to innovate at a speed that the rest isn't used to seeing. We had 130 SKUs in supermarkets in Chile, Argentina, Brazil, Colombia, Mexico, the United States, and Canada in 5 years of commercial execution. Not just any company does that. So clearly, we were able to show that there was an underlying asset, which was artificial intelligence, that was allowing us to do this in a different way. Also, let's say 400 million dollars, a bunch of other things. But the important thing is that people know your company, know what you're doing, and see that there's something there that interests me. I don't know what it is, but it interests me. And that's how the JV with Kraft Heinz started. It started because they saw us, let's say, taking market share, gaining market share in the mass market, from something niche, etc. And well, it was our first, without knowing it, our first AI partnership, right? Which later led us, let's say, to the division into two different business units, NotCo AI and NotCo Foods. But you realized there, the JV was the catalyst. It was wow, like that, yes, because it was wow for us and it was wow for the market. It was wow for the market because it said, 'Ah, wait, there are other ways to partner with NotCo.' And suddenly we get a project that wasn't to launch an animal imitation, we get a project to reduce the amount of cocoa in a product, maintaining the same properties, maintaining the beneficial properties of cocoa, the same color, texture, and aroma, etc., for a brand that makes like 18 billion dollars a year. And it was like, of course, the logic of our artificial intelligence can do practically whatever we want in a formulation because we understood that a formulation is a multidimensional space. If you change something, you change everything else, right? Predicting how the rest changes when you change one variable is extremely complex. The vector space in which you can move in terms of potential solutions, especially if you don't know all the ingredients in the world, the combinations, the chemical reactions, is practically impossible for a human to predict how the rest changes if you change just one thing. So the clients even led us to understand what things Giuseppe could do that we weren't even selling. Then came a sugar reduction, then came better electrolyte absorption in the body. We stepped from project to project to project, and today we must have around 80 projects running with 30 of the largest consumer goods companies in the world doing all kinds of projects of the same nature, right? So that allowed us to say, well, today we don't compete with consumer goods companies, today we compete with Anthropic, today we compete with OpenAI, Microsoft, Palantir, Nvidia. And that's what led us, in one of the companies, to be in a pilot where they told us, 'Obviously we are comparing you with models from Palantir, Microsoft, OpenAI, and Anthropic, and the winner will get our AI transformation contract.' Imagine the game we put into that. That was our startup. If we achieved that, everything else stacks up. So we can be the Harvey of consumer goods 100%. When Microsoft found out we won this AI transformation bid, they said, 'Okay, we'll distribute all the technology. We have more than 40,000 consumer goods clients with Azure, with Cloud, why don't we get involved?' And there I saw that little graph, that exponential graph. I said, 'Wait, we're back, we're back.' Not only that, inbox, mail, mail, mail, all the partners who knew us before, who didn't end up investing because NotCo's capitalization as a tech company wasn't there yet. 'Hey, what's up? We heard this, we saw the Forbes article, AI vertical specific, you guys are number one, you've been at it for 10 years. I want to invest, I want to invest, I want to invest, I want to invest.' So to tell you three years ago, we're never going to get up again. Suddenly I start talking with the CEO of OpenAI and with the team of Mustafa Suleyman from Google DeepMind, with the Anthropic people, with Nvidia, and I said, 'No, no, wait, this company is another company.' So 8 years of feeling naked, 8 years of turning a deaf ear, saying we have to get there, we have to get there, and it wasn't happening. We could have gone under before. And I think that super conscious, intentional decision we made three years ago in the board, with the risk of looking like an idiot in front of the board and saying, 'We're going to open our AI branch,' was the best decision we made in the history of the company. And having suffered what we suffered to get to break-even in consumer goods, I can't tell you, and I don't want to make it easy or make it look easy. I think founders, when we go through hellish waves, we make it feel easy, right, man? 'Wow, yeah, we rowed with everything, but here we are.' And man, tell me about waking up at 1 AM, at 3 AM thinking with your head, man, walking around more anxious than hell, saying 'I'll solve it this way, not that way.' Panic attack, it happened to you, 100%, and it happens to everyone. It's just that we make it feel easy, man. And that's a serious problem, because a lot of people get into entrepreneurship without knowing what they're going to suffer. Man, I'm the first to say, 'Hey, watch out with entrepreneurship.' You get on a hamster wheel, you take on brutal responsibilities. Like having kids, man. I'll throw that back at you. Three years ago you didn't have a kid and weren't married.

No, today I'm clear, I lose certain... I have more responsibilities, I can't, I don't just call the shots anymore. No, man. And what you could do before, you can't do now, but you learned and you have more desire. You don't see gray hairs on me, but I do have a couple. I have a couple. But it gives you, it really gives you, at least for me, I don't know how it's happened to you, maybe I've been lucky that my wife is a rockstar and the baby turned out well-behaved, but it gives you a different energy to do things like, man, I don't do this for myself anymore, I do it for others, and that's good. We're not going to get into that, right? No, man, if we want to talk about you, we want to talk about you. We want to talk about you. But hey, speaking of that same thing you're saying, I want to emphasize the peach, in fact I remember you talked about mental health and you talked about the side of NotCo. Yes, yes, and it was very entertaining and different from the others. You have that quality of storytelling that few founders in Chile have. In Chile, the Chilean is bad at storytelling, very bad. Some better than others, but in general, it's very bad. I've talked to more than 1,000 entrepreneurs and in general, it's bad. It could be. And I want to talk about that moment, 2023, 2024. In fact, I remember we met off the record, so we can't talk about it, in a Starbucks. Starbucks, but I left and said, 'Wow, this guy is having a terrible time.' Yes, very bad. And I remember you were telling me like, 'Man, we're rowing hard, but it's extremely difficult,' and that was probably before all this boom. Obviously, I want to talk about that moment when you came back from New York, when you closed the headquarters there, when you did a big layoff, and I think luckily you had raised 450 million, otherwise you wouldn't be sitting here. Look, the 'ifs' are very dangerous, man. It's true, because in the end, you make decisions with the information you have. But we were indeed in a moment where there was little clarity, and I always say that it's okay for CEOs not to know sometimes where they're going. You have to communicate it, you have to say, 'Guys,' and I think in the Town Halls, I tell you, I think there are nervous people today, surely at NotCo watching me saying, 'Man, look, I think it's this way, but I'm not so sure.' And things can change super fast too. I think transparency today is underestimated, man. Because when you operate with transparency, in the end, it can give you more fear, but at least you have nothing to hide, man. The worst is like saying, 'Man, yes, we're going that way,' but you're not really going that way, and you know you're not going that way, and you have to be really two-faced. I can't. I have the virtue or vice of not being able to pretend I'm okay when I'm not okay, you know? And you saw it in that Starbucks, right? It was a moment when I was coming back from New York, the New York office was moved to Chicago, in the Craft Heinz building. You know, as much as I wanted to see it as a win, because in the end it was like, 'Wow, the New York office is gone,' you know? Maybe for a greater good, but wow, I came back to Chile, but I came back to Chile because I could have gone to San Francisco without a problem. The NotCo HQ is there and it's today the center of artificial intelligence and it's where, I don't know, 90% of the companies we're working with in the AI business are there. I said, before this, I have something to solve, right? That I don't know. I'm not a founder who took a company to profitability, I don't know how to do it. I don't know how to do it and I need to surround myself with people who do know, right? So I think I had a plan at that moment, I didn't know if it would work, I didn't know if I would be able to solve the consumer goods problem, that is, get consumer goods to profitability at the same time I was building the foundations of the artificial intelligence business. I didn't know. Now, if there's one thing I am, it's optimistic. Man, and we were discussing it a bit before. I don't see the billion things that can go wrong because they will go wrong. I see the one thing that can go well and change everything. And I aimed for that. I said, I'm going to make a profit. I'm going to generate an AI business that brings back the dream of the $10 billion company, and I'm going to do it in the next two or three years. Because I have to do it, because if not, I'll fire myself, you know? But, but you had lost that dream a little bit at that moment, like, how am I going to be a... selling mayonnaise on the shelf, right? No, of course not. Of course not. Of course not, because the world moved. Because the goalposts moved 3 km. The goalposts were here, you know? And they moved 3 km, and not because of you, because of the environment. Everything changed. The dynamics of consumer goods changed, private labels took over a percentage of the share they hadn't taken before. Three fundamental things started to happen: consumer preference, even politically, wokeness changed diametrically from being fully committed to sustainability, to a better world, to caring for the forests, the Amazon, etc. 'Ah, man, I'm drinking raw milk,' you know? RFK Junior comes out saying peptides are the best. So everything changed, and there are things that don't depend on you. What does depend on you is not dying, right? And you have to adapt and change, and basically make a list of the five things you have to change with agony, Juana. Because in the end, you also say, 'Wow, I think I'll be able to change.' Changing isn't easy and it takes time. So the anxiety you get when you say, 'I'm changing everything and nothing changes,' is brutal, you know? And that happens, but I had to do it in person, and that's why I came back to Chile, because in Chile, in the end, it was like, this is the alma mater of NotCo, and we grew to Argentina, Brazil, Mexico, the United States, Canada from Chile, right? I want to go back to the origins, to the old building, to sit in the office, to feel the team, because there are things... staying behind a desk is a very dangerous place to make decisions, and going out to the street and seeing the shelves and what's happening and talking to the consumer and talking to clients. I had to go back to Walmart, back to Jumbo, back to Cencosud, back to Pão de Açúcar, back to Carrefour. It was like, where did we screw up? How do we change? etc. So it takes time, it's hard. I had to get on a plane again, three times, four times a month, and fix things and understand what we had to change. Well, I made the decision to change some things, and 90% failed. 10% worked. Back to another cycle of decisions, 90% failed or 85% failed, 15% worked, and so on, fixing things. And three years later, after a lot of change, a lot of fixing, a lot of decisions, discontinuing some products that I loved, but because the margin didn't work, we simply had to discontinue them. Those kinds of decisions that the Matías of three years ago was incapable of making. Today, when they present me with a P&L, no matter how much I love the product, the brand, the aesthetic, the creative campaign, etc., I say, 'Let's see, show me this, does it have above 45% gross margin? Okay, I don't care, I'm not interested.' You know? So it took time, and it's directly proportional to the time it takes one to change their mindset. So you're talking to a more mature Matías. Who were you talking to at that moment? Who did you ask for help? Wow. Man, look, the entrepreneur can ask for help, but in the end, we're also quite stubborn and think we know more than everyone else. I can guarantee you that. So you can receive 25,000 pieces of advice. This is like love, man. How many times did you give love advice, man? And the guy says, 'Yes, yes, man, tomorrow I'll break up.' And then he posts a photo on Instagram, man, with a bouquet of flowers, man, in Viña at the hotel. You know? This is the same thing. We can't be more human in that sense.

And I have the virtue, and I think I've also discussed it, of having a close family, to whom I ask for extremely a lot and very frequent advice. I have an older brother who I think is a crack, a crack. In fact, he helped me several times at the beginning, especially to raise capital. He was my rational part until José arrived, who is my CEO. My brother was, let's say, the one who helped me in that sense, and my dad. And they are two people who have a lot of business experience, who have a side that I don't have, which is a bit more numerical, the 'this isn't going to work', the 'these are the decisions you have to make', and they do it with love too because they consider your human side, which is a bit difficult for an investor, right? An investor obviously has the obligation to return the money to the fund, to make a 10x in the next 10 years. So, they also have an interest. There are some investors who wear more of the NotCo director hat versus the hat of protecting the obligation to the LPs of my fund. You know that's also another thing that isn't discussed much when chips are down. That's when you really see who are investors and who are your business partners. And I've always said it too, Nico Szekasy of Kaszek Ventures, one of the founders of MercadoLibre, and Elio Leoni, who is a board member of Kraft Heinz, of Ambev, of Barry Callebaut, who has a lot of experience. I think I got very close to those two because they know about crisis, they know about change, they know about adjustments, they know about a lot of things. So, I surrounded myself with them and I surrounded myself with my team more than ever. I think showing yourself vulnerable to the team and sometimes having your back covered is key. 'Joe, man, please, take this play. I'm too stressed out, man. Take the operation of this thing, please. Help me decide this.' Uh, man, I don't know, INE, our marketing team, which wasn't obvious it was going to be our marketing team. I said, 'This is the one I need.' INE covered everything for me when I couldn't attend to the mass consumption business, CPGs, I was too focused on closing this $30 million contract, another $20 million contract. I needed someone to have my back. And you can't do everything. So you find that team where you articulate your strengths and weaknesses and you know perfectly who is backing you up.

No, look, talking about numbers seems to me that we are also normalizing private companies talking about numbers. You don't have to say what your sales are or how much you burn or the cash position, those things. I never saw any founder of Airbnb, Uber, MercadoLibre say that. I mean, you saw Galperin and he never in his life until he went public said how much he sold, how much he burned, etc. So, I maintain that right, but at no point did we have it bad because there was an existential risk. Now, you never want to operate with existential risk because the amount of screw-ups it's going to send you will be infernal. Imagine, you're with that danger of dying. You can't think, man. You can't sleep. You make all kinds of bad decisions because whether we like it or not, we are in a risk business. You can't be a zero-risk business. They wouldn't have invested in your company. That's why. You didn't found your company or your startup just to be profitable. You created it because you wanted to change the world, because you wanted to build the $10 billion story, because you wanted to change people's lives out there with something. And suddenly you feel like no, because you're on this hamster wheel. And for me, the best things that have happened to me in my life as an entrepreneur aren't the photo in the club, the photo in Candelaria because today, man, you're a crack, man. 'Damn, you changed me.' Those are comforting, obviously, but when a guy comes up to you and says, 'Man, I had an accident, the only thing I could eat were your products and you changed my life.' It's like, man, damn, in those moments... So anyway, one doesn't do it for the credit, one doesn't do it for the press, sometimes it gets confusing. I think one also confuses success with how much you appear in the press or how many billions of dollars, or success is something else, you know? And also in this process of maturation, you start to favor some things that you didn't before. Because you were always probably one of the founders with the most press, for better or for worse. I think so. Didn't it ever happen to you that you said, 'Man, I'm an imposter because I'm out, the media calls me all week, but I have a giant mess here all the time.' Or [laughter], I mean, the impostor syndrome is the most normal thing an entrepreneur can suffer. It's crazy because you feel like an imposter all the time and it requires other guys to tell you, 'Hey, man, stop, look at what you've done. It doesn't matter how this goes. Look at what you've done.' And one is incapable of doing that. One says like, 'Man, yes, all good, but my goal...' You're a nonconformist in a sense, I think, quite negative, because in the end, your moments of joy are very few, because closing a round only generates more anxiety, because closing a contract only generates more fear, because you can't deliver, because you promised something that will very likely fail. Do you know how many times it happened to us with the artificial intelligence projects? It's like, man, look, the most sophisticated teams of the most sophisticated companies for 15 years haven't been able to solve this. Why the hell would we be able to? We said yes, we moved forward, and we hoped that 1% probability would appear. And it appeared many times, but other times it failed, you know.

Uf! Man, I have many, it has hurt me a lot, many things, huh? What hurt me the most? Good question, man. This one I didn't have planned. What hurt me the most? Uh, no, I think what hurt me the most, at the same time made me grow the most, is realizing that the things I thought were cooler than shit, the ad on Fifth Avenue, in San Francisco, in Phoenix, man, at the door of Danone, of the Danone office in New York, that it was cooler than shit. Damn, I think about it and say, what was I doing? You know? But it's crazy because you also say those were other moments. So, on an individual level, when you analyze that, you say like, 'Oh, what pain, to have spent the money to do that thing.' But then you say, 'Man, I think the guys from Danone or Coca-Cola or Ferrero, who today are our clients, saw that.' So, as Steve Jobs said, when you connect the dots, it doesn't seem so stupid anymore. But the pain of having had to change that mindset was tremendous, it was abandoning things that you might regret. And that, I would tell you, connects a bit with my little appreciation for politics. A politician can't do that. They get screwed. This guy who said this comes out, now says the other. You're screwed, your political career is screwed, you know? But that's how the human being works, man. So, that's why I'm never going to get into politics in my life, but what a regret. Afterwards I'd make a good minister of the economy, you know? But anyway, changing is very hard and realizing you screwed up is very hard, but then you look back and when you connect the dots, everything happens for a reason and it's all a domino effect and nothing is disconnected from your story. What I would have regretted is not having changed and having continued the line we were following, and having followed the advice that many people gave you. 'Hey, man, stick to your knitting, stay on this path, don't leave, these are market adjustments, everything will come back.' No, they probably told you that several times, man. But there's nothing wrong with selling either. I mean, when you reach a point where your babies... it's like, man, what parents must feel when their daughters or sons leave the house or get married, or sometimes you also have to say, 'Man, it's time to mature and companies have cycles too.' But it was on the table, I mean, it has always been on the table, there are always companies that are very interested in grabbing what you built, because they can't build it themselves, or because you talk to an audience they can't talk to, or because you have an incredible brand that can say things none of their brands can say. And there are those moments that I think every founder has to always consider this thing about dying with your boots on. Why? Look at Dani Hajnal and Oscar and Cuevas, happy, man, you know, of having sold. Happy and on to the next thing, you know? Because you're an entrepreneur, because just as you changed one thing, you can change another. And you're a person who also, back to politics, you can go down a path and think something is extremely important and it's layers and it's what the gringos call the calling. Your calling is one and then in another stage of your life the calling is another, you know? So not being afraid of that seems fundamental to me and always having the maximum amount of opportunities to continue and several paths. Why not? Why not? Let's go to that calling, man, the current moment.

Correct.

I think so. Look, to tell you why we were able to beat Palantir, I have to go back a little bit. To help understand a bit what we've built with the NotCo models and Giuseppe. It's the first platform trained and created with mass consumption data. For 10 years we've been extracting molecular, spectral, physicochemical data and contrasting it with sensory training from human beings to a model and an algorithm. When you find the underlying patterns that exist between the molecular components of food and human perception of flavor, color, texture, and aroma, you can basically predict how to formulate better. Not only that, but also when you start integrating some consumer differences, social, demographic, structural, behavioral. And I'm going to add this which is super cool, which is lying. Humans lie a lot. You're capable of telling me you don't like something you do like, or that you like something you don't like. Okay. For a sense of belonging, and especially with younger generations. When you go to an older person, when you go to your dad and he doesn't like something, he doesn't like it, man. It's like, 'This sucks, man.' Fantastic, because the older guys care less about your feelings than someone younger. Generation Z is very intrigued with what their reference likes, right? Social media changed everything. So, there are people today who are consuming things and buying things they don't like, but because someone else consumes them and to belong to a group, they consume them. We also, apart from all the logical and rational stuff, had to get into Giuseppe's brain that people lie, okay? For a lot of things. So, what's called behavioral finance, sometimes we say you made the classic mistake of creating a solution for a problem that doesn't exist, okay? And that's the most expensive thing there is, okay? So, we have to train this algorithm a bit with all these notions of the human being and consumption dynamics. So, from what the consumer thinks to how formulas react in industrial processes. And today, you simply give me the architecture of your production plant, I can predict how a formula will behave inside your production process. So, in the specificity of variables that affect mass consumption, NotCo is the company with the most data and the most specific data globally. Palantir is extraordinary, extraordinary, but only with the data its client has, right? We have basically built the data that no one has. The beautiful thing about Giuseppe is that if you have data, you put it into the platform and the platform gets smarter. And why am I today able to offer artificial intelligence to competitors? It's because their data will be a different version of Giuseppe. Okay, so Coca-Cola will potentially have a Coca-Cola that will be completely different from Pepsi's. Mondelēz will be very different from Ferrero, Ambev will be very different from ConchitaTorres, you know? So that allows us to basically have an algorithm 10 times more solid and proven than any Palantir out there promising this change. Also, NotCo has been extremely intelligent, and that's why I tell you that NotCo Foods was so important for NotCo AI, is that all the models were developed with pain. We realized that the formulation problem wasn't the only problem, man, the problem was much deeper than that and it wasn't just an R&D problem, it was a marketing problem, it was an innovation problem, it was a sales problem, it was a finance problem, it was a fundamental workflow of these companies that was broken at absolutely all information transfer nodes. So, I'll give you the example of cocoa. Cocoa is one of the most fundamental raw materials in the world. Cocoa basically grows in West Africa, particularly in Ghana, okay? And in Côte d'Ivoire. And for 3 years it hasn't rained like it used to. The futures were trading with brutal evidence that a cocoa price crisis was coming. The finance guy knew the cocoa mess was coming 3 years in advance, but because the finance guy doesn't communicate with the marketing guy, with the innovation guy, and with the R&D guy, companies started losing a margin level that was absolutely detrimental to the attractiveness of the business. So you see the top publicly traded mass consumption companies trading at minus 60% versus the S&P, and it's a transformation problem and it's a technology problem. So, NotCo allows these companies to regain their competitiveness. It's using their data, connecting all the workflows, saying, 'Man, I'm never going to make a product or even think about a product that won't resonate with the consumer I want to launch it to, that the formulation won't work in the production plant where I'm going to make it, that the regulations of the countries where I want to launch it are completely different, that the consumer where I'm going to launch it, the formula for New York has to be different from San Francisco. Because the guy in New York likes something different from San Francisco. So, for the San Francisco one, add 2 grams more of this, and for the New York one, 2 grams less of this. In this one, remove an ingredient. In this one, add two ingredients that sound more natural.' Those differences today are managed with Giuseppe. So, not only that, and it's a bit of what I'm going to show you too, which is the first time we're going to show how Giuseppe really works, is a simulation. You're never going to have to launch a product to the market without first testing it in a virtual world, okay? Just like Runway, you know? It creates images, I mean, it needs an infinite amount of images to be able to represent the real world, the reflections, the lights, the sun, the movements, etc. We do exactly the same thing with people. What do they share? When do they share something on social media? Who do they share it with? How do they talk, do they like the brand? Don't they like the brand? Why do they like the brand? And I can simulate a conversation of why a person liked or didn't like a product, a particular person, someone sitting in Ñuñoa, you know, who is 19 years old, is a woman, rides a bike. I have today the ability to interact with that brain, right? On a sensory level and on a cultural, social, and demographic pattern level. That is the power it has today. It's allowing companies to test products before even launching them. And I'll give you one more example before we get to the question. There's a company, the biggest in the world or the second, the biggest, I think today it's the biggest in the world, that has the coolest brands. And they told us, 'Guys, we want to add a functionality to a brand that will make people who don't consume us today consume us.' It was a nighttime sleep song for women, etc. And we put Giuseppe to think about what it could do. It offered us 30 routes, okay? Of the 30 routes, we chose one, but we left one in the drawer that seemed to me, it seemed to me something truly demonstrative of what Giuseppe is capable of doing today that a human being might never have thought of. We went for the most obvious path, but this opportunity remained that I'm going to tell you about now. Giuseppe said, 'People in hospitals today are dying not from the disease they have, but from malnutrition. Today, nutrition is the leading cause of death in hospitals. Why? Because when you're that sick, it's very difficult to pass nutrients, okay? And the best way is through ice cream, frozen things that are melting.' So, the idea was to launch the first ice cream brand for hospitals to save the lives of many people who were dying of malnutrition. When I saw that thing, I said, me, who feels like a very creative guy, man, who feels like someone with a, I don't know, man, I can feel white spaces, I would never in my fucking life have thought of that. And well, it's going to be a category that we're going to develop with this brand, but that in the end tells you, man, of course, Giuseppe today is able to offer things that a Palantir, a Microsoft, an OpenAI, an Anthropic, and a Google DeepMind still can't because they consider or because they have training that is more limited than ours. So that's what today, and that's what today basically puts you back in the ring of the $10 billion company. Because going back to the Forbes interview that puts you at $1.5 billion with a revenue of $75 million, with your percentage too, which they threw out there, threw it all out [laughter] unofficial. Okay. Here we can see if we say it or not, but basically it seemed like a 20x multiple that made sense in 2021 or 2020 in a hype and that makes sense with everything you're saying today, but from 2023 to 2025 it probably didn't make sense. If you look at it in terms of a multiple of the mass consumption category versus estimated revenues, right? A fifth-year commercial engineering guy can build you a model and tell you it's completely overvalued. What happens is the market never saw under the hood what was there and what we were creating in the artificial intelligence part. That's why I told you that for like 8 years I felt naked, because also going out to show and demonstrate that Giuseppe was very powerful didn't make sense at that time. Today, a lot of the marketing coming from NotCo is with Giuseppe, I mean, it's 100% Giuseppe. What I want to tell you is that, of course, the valuation, the millions of dollars raised, while it went on the billboards in New York, an important part of the Z location, it also went on GCMS spectrums in $3 million machines that tell you the concentration of molecules that different foods, different products have. That path wasn't visible and maybe we sinned a bit by not showing it, but in the end, it was another path, it was... So today, indeed, everything I'm telling you, maybe 2 years ago I also said it, I also said, 'We're going for it, we're doing a B2B platform, etc.,', but it was a story, I mean, it was a vision, it was literally, man, we aspired to do that. Today I'm telling you that we are being 10 times bigger than what we were even projecting 3 years ago. And even an optimistic guy like me, I'm seeing the results and I never imagined we would be here. So, it's not about the billion dollars, it's about having an artificial intelligence business that has 80% gross margin, that we are closing with the biggest companies in the world and that we are beating Palantir in the vertical specific platforms.

Let's play with Giuseppe. Okay, basically here what we have is the world of Giuseppe. This is what eventually a company that hires NotCo AI can see, right? Correct. And it's what we've mapped, right? That other companies haven't been able to. So, this is a graph that basically indicates the things that in your brain are going to be important to be in a product. This is your brain. This is the brain of a person. So it's interpreting basically what things are going to be important, what things are good, what things are bad about the formula I'm creating. So, if I want to launch a product and this is your brain, I have a way to predict if you're going to like it or not, or how much you're going to like it and what things you're going to like and what things you're not. Okay? So it's a bit of the representation of the human brain, okay? And this is for a particular person. I mean, you could do and repeat this for many other people. So, you see from how certain molecules, vitamins, etc., affect your overall liking of a product, okay? Which is nothing, as I tell you, a vector space of decision-making, right? That involves many, many, many variables and in different priority, okay? And there are some things connected to others and there are some things completely separate.

You put a prompt.

The first thing is you have to consider what matters most to the target consumer. Without that, it doesn't matter what product. The most important thing is to know what the consumer you're talking to is going to like or not. So, once you have the representation of the target consumer, you basically come and say, 'What product do I have to launch?' And Giuseppe gives you the product you have to launch, okay? With all the characteristics, with absolutely all the considerations, and then you can ask it and predict to the consumer, and there you have the prompt target, you can say, 'Okay, man, what are the most unexpected findings in the simulation?' And there, on a consumer level, you're asking the person, 'Okay, this is already in the brain you analyzed, right? Okay, perfect.' Exactly. So, once you're polishing the concept you want to launch a bit, and this is the coolest part by far, then you had the product, you had the potential consumer, and you know more or less what matters to them. Here you literally have real human beings and this is a simulation of a synthetic world and it's not only what it does. We're in London. We're in London. Exactly. You launch a product and you not only simulate the perception, you simulate the interaction. If they share it, they don't share it, who talks to whom? You have people in this radius who are basically interacting with each other, doing all sorts of things, going to work, coming back, getting on social media, and what you can do is I launch this product, how does this product behave in the entire population? If you start seeing, yesterday was night, so they're sleeping, okay? So you start to see how the product interacts with people and how people start talking about your product, which is by far the most fundamental value that Giuseppe has, which is called Giuseppe World. It's literally a synthetic representation of consumers and the interaction between them when you put a product in that world. But then you also know this neighborhood, we're looking, well, we see Oxford Street, Westminster Bridge, etc. You also know the idiosyncrasies and no, that's not the word, but the people who live there, I mean, look at what... that, I mean, you also run demographic data. Totally. Here, this is what's happening behind the scenes. Literally, you're there making a representation at the consumer level of cultural, social, cultural, social, demographic, and not just sensory patterns. Today, with the information you have from social media, you know how a guy in this quadrant is behaving, what he shares, what he doesn't share, how many people in that sector live under the age of X. So, and how much they lie too, how much you have the information of what they say they like but what they really buy, you understand the differences. There are people, for example, that we've realized are much more dishonest in these super... In the United States, there are some geographies where people are much more hypocritical than in other places. New York is more straightforward, right? In New York they lie less. They don't care about telling you to your face that they don't like something or that, man, the fashion thing is very strong, right? But it happens more with clothes than with food, right? Okay, in New York particularly, and in Nebraska, in Arkansas, in Dallas, people lie three times more than in New York. So you have to clean less data in New York than you have to clean in Dallas. So basically, it's an algorithm that represents behavior and that's the cool thing about this simulation. Why? And here is again my more entrepreneurial side in the sense of if I can do this with food, why can't I do it with a bicycle? Why can't I do it with a clothing brand? Why can't I do it with a restaurant? Why can't I do it with a cream? Why can't I do it with a service? If I can simulate demand, why not... like a diagnostic tool? You know? Giuseppe World is a synthetic representation of how the world behaves with structural patterns, behavioral, social, cultural, political, etc. A bit of the last thing I would tell you that NotCo has, which has been by far the most relevant of the recent clients we've made partnerships with on the AI side because of these things, you know, Ferrero is calling you, there are many brands that have been mentioned for this, obviously many other things too, but this is what makes the platform something much less R&D.

This is... like, you changed the way I think about launching a product forever.

This simulation has been run many times before launching a product. Yes.

There are three products launched today that went through this simulation and were curated by this simulation.

This is the result. This is the result. After the simulation, basically, you want to see a general result of how your product performed in the market.

Sure. So, you wanted to get as close to 10 as possible. And the good thing is you know why you're not at 10. So this is like the final report, you know? So, all the decision-making latency, this product was obviously launched for this type of people. So, how did it resonate with some? And sometimes, man, you want it to resonate with fitness buffs and you're okay. Now, the overall, because you considered these three tribes as the most important, you got a 7.4 out of 10. I'm taking creatine, so I feel like a fitness buff here, I don't know. I'm going to run a marathon. It's difficult, [laughter], you know? So, you can run this eternally, right? And the indicators that tell us if this product you're potentially going to launch will be successful have been 8.3 and up. So, if you have an 8.3, the probabilities are 80% with good commercial execution. It's tremendous. This reduces the 2 years that a new product innovation cycle has within mass consumption companies to literally 3 months. And from this, the product can't say which ones are on the market. We can't, unfortunately, but because... because it's not known that they were made here or because... all the above, because the companies, let's say, the mass consumption companies are quite picky about the information they give about product, about formulation, and about strategy. Because if you're working on replacing cocoa in brands that are multimillion-dollar or multibillion-dollar, you more or less know what that company's strategy is. If you reduced 35%, 40%, or 80% of the cocoa or the application of cocoa in a product, you're also showing some strategy. So, we tend not to say how much, not to say what product, etc., but they are products that are going extremely well and that have caused... and that also, let's say, you have to be aware that there are also ways to partner up. One way is a project, let's say, fundamental but very particular, is to say, lower the cocoa by 30% in this product without, let's say, any consideration of differences...

Sí, correcto. Hoy día estamos yendo por una estrategia mucho más inteligente, ¿ya? Ni siquiera, viste como nosotros vamos a hacer el caballo Troya, y hoy día, viste, no hay necesidad. Antiguamente lo que pasaba era que éramos una compañía de consumo, por lo tanto la relevancia de la marca per se era extremadamente importante y si nosotros no mostrábamos que era NotCo entonces perdíamos mucho. Hoy día ya nuestros clientes saben, nuestro cliente no es el consumidor necesariamente, nuestro cliente es acompañado consumo masivo. Entonces ya nos conocen, ya saben lo que hacemos. Entonces en ese sentido soy mucho más como un Palantir que como un qué sé yo, un Coca-Cola, ¿entendés? Entonces ese también es un cambio de vuelta.

Espectacular. Estamos preparados para gente como tú. Huón, créeme que los son más mañosos que la mierda.

A ver, una de las crisis fundamentales que está enfrentando el mundo, que para mí es la cagada que la gente no sepa tanto, son la disponibilidad de naranja en el mundo. Gran parte de la producción de naranja era en Florida y Florida tiene problemas heavy de clima y tiene uno mucho peor que es un virus, ¿vale? Las plantas de los naranjos están con una cuestión que se llama greening effect, que hace que las naranjas sean mucho más ácidas y amargas. Entonces, si hoy día haría una compañía de jugo naranja, estás en problema. No sé si habéis visto el precio del jugo naranja.

Todo el mundo habla de café, todo el mundo habla de naranja es otra. Aceite de oliva hace poco. Bueno, naranja es otra. Entonces, teníamos que encontrar una manera de generar el sabor a naranja sin necesariamente tener naranja. ¿Por qué? Porque en 10 años muy probablemente no vamos a tener suficiente naranja para tomar desayuno con jugo de naranja todos los días. Así que esta es una de las fórmulas más importantes que hizo Giuseppe a nivel estructural del mundo, ya. Y aquí para explicar a la gente que nos ve en el fondo, que nos escucha y no nos ve, mucha gente solo ve audio, o sea, escucha audio. Aquí tenemos una cajita con cáscaras de naranja.

Adorno está bien, pero que te tira harto olor.

Ah, okay. Si queréis lo sacamos.

Ya. Y tenemos dos cucharas que son lo mismo. Son lo mismo. Exactamente lo mismo, que en el fondo sabor naranja. Claro, es una base. Esto es una pasta de naranja sin tener absolutamente ni una naranja dentro. Pasa por un proceso de fermentación de una base de ingredientes que son super baratos y que permite a ti tener exactamente la misma experiencia que una naranja sin tener naranja.

Ahí.

No, no está diciendo trampa. Mira, exactamente naranja. Y te digo una de las razones interesantes que nos lleva a poder llegar a esto. No sé si hay seguido un poco la historia de los diamantes sintéticos.

Ya, los diamantes sintéticos lo empezó General Electric para digamos los chips. Se hacen con silicona y son muy ineficientes versus diamantes. Los diamantes eran muy caros, por eso no podían poner diamante a los chips. Pero todo va a cambiar porque hoy día fueron capaces de, digamos, hacer una ingeniería reversa para entender cuál es estructura molecular y físicoquímica y el modelo 3D de que hace un diamante un diamante. Si tú no sabí lo que es un diamante a ese nivel, es imposible recrearlo. Nosotros hacemos lo mismo que la comida. Es entender a nivel estructural, molecular, físico, químico, qué es algo para poder aplicarlo. Y esto es lo que está cubriendo.

Tenemos ahora chocolate. Acá tenemos chocolate de leche. Y cacao no fue afectado, digamos, por esta formulación. O sea, en esta formación cacao, no hay cacao.

No hay cacao. Creo que la gente tiene que entender que hoy día la tonelada de cacao está en 8000. Al principio de este año estuvo en 15,000 y hace dos años estaba en 2300. Yo entiendo que viene ahora un fenómeno del Niño que lo va a disparar también. Ven, estamos en grandes problemas si nos gusta el chocolate y lo que está haciendo NotCo es que nunca te quedes sin la experiencia de chocolate en tu vida. ¿Entendís? O que no tengáis que pagar 15 lucas por tu barra favorita. Y eso lo has visto en el supermercado también, seguro.

Tiene el olor.

Claro, sin mencionar marca. Esto fue un chocolate hecho para una marca que quería probar hasta qué punto podía llegar en la reducción de cacao para generar la experiencia sensorial y los beneficios.

De acuerdo. Y eso nosotros lo cateamos, nosotros tenemos descrito como baby vomit. Es un sabor, puede ser por ahí, es un sabor a leche rancia. Si tú vas a la historia del chocolate, una de las maneras de preservar leche en polvo era a través del cacao. Por eso se... Bueno, escucha el Acquired podcast de Mars.