De nieuwste AI-modellen helpen Ikea nog niet: 'eerst moet de basis op orde zijn'

Return on AI wordt medemogelijk gemaakt door PVC van AI-ambitie naar impact. Leuk dat je luistert naar Return on AI. We nu op podcast van Belair over misschien wel de belangrijkste transitie voor bedrijven in tijden. Welke AI-strategiehouterorganisatie ook in de toekomst winstgevend? Hoe krijgt je het werk neemers mee? En waar is nou de grootste efficiënties lach te behalen? Die belangrijke vragen stel ik aan de bestuurders die deze beslissingen moeten maken voor de grote bekende bedrijven die we allemaal kennen. En deze aflevering is bij ons gast inkaartgroep of te wel Ikea. Inkaartgroep is de franchise neemer, de grootste franchise neemer van Ikea. En ik ga niet alleen die vragen stellen. Daarvoor heb ik aan mijn zijden, AI-experen en ondernemer Robert Gaal, opregten we een onder andere AI-start-up climb, maar ook voor Google gewerkt en nogal meer start-ups heb je op je naam staan, hè? Het is een bevijf door alweer. Wat is het met die eerste vier gebeurd? Soms de succesje, soms niet, zo werkt het. Oké, ja, dat je hoeft niet nog allemaal in de luchtenhoud. Nee, nee. Het vijm heeft nu jou vol aandacht. Het is vol aandacht. Robert, waar gaan we het over hebben deze aflevering? Nou, we zien dat Ikea heel erg bezig is met het leerde van nieuwe vadergeheden aan mijn personeel. Daar hebben we natuurlijk allerlei vragen over inge hele stijl. Van Ikea hebben zij ook heel veel opmerkingen over hoe ze het allemaal inrichten. Het sanitaire, wat hun nodig hebben, qua vadergeheden en qua gereedschap. En ik denk aan het einde hebben we nog even over privacy. En wat dat betekent voor Ikea, in z'n wereld waarin heel veel data hoeft vergaard. En wat doe je met Ikea op dat gebied? Ja, en vooral ook het aandeerde van nieuwe vadergeheden. Dat betekent dus ook van, ze gaan ze niet bezig heen ons slaan. Als mensen niet iedereen is op eens vervangen bij Ikea. De eerste afleveringen nemen we speciaal op tijdens internationale E-Ai-conferentie Human X in de rij in Amsterdam. En we zitten hier in een exekje te vlaan. Van PVC, Robert, jij staat jezelf ook op de vloer, hè? Met je stardepje? Ja, klopt. Maar ik ben verhaalelijk in de wandelganger natuurlijk. Want het echte verhaal is misschien niet op het podium, maar de positieve dingen worden gedeeld, maar ook echt die negatieve ervaringen. Ja, wat gaan we nou doen deze week? Er voor mensen in de wandelganger eerder te zijn? Ik denk dat er wel frustraties zijn over hoe kan ik het beste gebruiken. Dat is van een groot bedrijf tot een klein bedrijf. Ik ben zo de gereedschap, dus we kunnen hierover leggen. Een kleine bedrijf, die je hetzelfde gereedschap hebt, ben ik dat ook dat zij dan weer toch nog een bedrijging kunnen gaan worden voor die grote bedrijf. Absoluut, meer dan ooit te voor in de geschiedenis van de mensheid is te starten op een bedrijging geworden. Dus ik als stardepvander, zie dat ze weer positief in. Weet je daarvan tegen Goliath, he? - Heerlijk. De komende zes afleveringen nemen we hiervandaan dus op met zo'n wel engels als Nederlandse gasten. De komende weken kun je het allemaal terugluisteren in je podcast viet. En daarna gaan wij natuurlijk door, Robert in de BNR studio elke week op maandag. Absoluut, gaan we voor. - Ja, dus abonneer je nu op deze podcast, zodat je niets misst en je kan ook reageren vragen stellen. Het kan allemaal via de BNR app. Dan nemen we dat mee in een van de volgende afleveringen. Genoeg gepraat, laten we snel naar daar de gast van deze aflevering. Inka Group is de grootste franchisenehmer van IKEA en zit gewoon lekker Holland's met het hoopkantoor in Leiden. Het bedrijf Runt de IKEA-winkels in ruim 30 markten is verantwoordelijk voor IKEA.com en de app, en boek te volg boekjaar 41,5 miljard euro omzet. Dit jaar verlagd IKEA de prijzen van onderde producten, maar is die prijstaling met AI terug te verdienen. Onze gast is Parac Parac. Hij is de Chief Digital Officer van Inka Group en verantwoordelijk voor de tichetale transformatie van de IKEA-winkels en de website. Parac, welkom. Dank u voor het gevoel. -Dank je wel. Ik was vond het al in de vroeg en ik was voor IKEA-objects. Ik denk dat iedereen een koffie-koffie, een koffie-koffie-koffie in de house, in de house. Maar ik heb het niet gevoel. -Nou, we moeten naar de sponsoren te veren. Dit moet je gevoel. -Dat is het. Right off, in het interview, je kan je de goede gevoel. Let's start off. Ik ben een goede gevoel. Als je op de IKEA-office op je gevoel is, 5 jaar dat we nu nog niet verantwoord zijn. om hoe IKEA werkt, omdat de AI is. Ik denk dat het heel interessant is. Ik denk dat we op een koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie. om een businessperspektive koffie-koffie-koffie voor ons. De eerste, die is voor veel organisatie is rond groot. En wat is de opportuniteit de AI-office voor groot, besides de regulaarorganic groot dat we hebben? De seconde belangrijkste is de IKEA-office-koffie, die er echt aan een fordabiliteit is. Deze interdactie, die je het op ons gebruikt, is de grootste voor ons om te strukturelijden. om bij een relatie in de NEC-office-koffie. om hier te weten wat de possibiliteit dat de AI-office-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koffie-koff eerst. Ja, ik ben er echt erg vergeten. Ik heb er ook nog een effect van dat. Is dit ervoor dat je ervoor gebruikt? Dus ik kan bij de cheapest table bij het leven. Ik zou het zeggen, ja, ik weet het niet. Er zijn er een paar waar we nu de eerste roos hebben. om van de AI in de gegeven moment te gebruiken en especially als we het in de N2-end-eco-systems en supply chain hebben. Maar er zijn er een paar waar we het gewoon starten. Maar als ik een van de exemplen van een ROI-perspecteur. en ik denk dat we het grootste ROI nu niet uit de gen AI gebruiken, maar meer uit de traditionele AI gebruiken. Wow, ik denk dat je ook in alle feestjes in dat sens. Waarom gaan we de supply chain en de fulfilement van? Er zijn er ook nog een van deze gebruikers starten. zoveel de eerste GPT-schermeld naar leven. En er hebben we in de praktisch en we hebben optimisten. En vandaag, when someone places an order, AI is smart enough to then tell us. where is the best to fulfil it from based on a few parameters that we want to really focus on. So there we start seeing the biggest ROI come from the use of AI to supply chain. But then there are some aspects where we feel our opportunities on AI. Maar we have just about got through out of the gate, and we see engagement, but not yet to a full scale from an ROI perspective. So if you dig into that a little bit, because we're in an age now where the models are improving and improving. And previously, you've said that the productivity gain in other interviews is about 5% to 10%. With what's been happening recently, so we see new models coming out. We see a lot of other capabilities, not just in the quality of models, but also in how they're used and how they're deployed. Is that still true? That's 5% to 10%, or is that old history? Is that. are we seeing higher productivity gains in the use literally between you and your colleagues every single day? I think I admit what we see with the AI models, you see a drastic improvement in the last 12 to 18 months. That said, I think one of the bigger challenges enterprises like us are mindful of is, in many cases, we still don't have the foundation. Yeah. I call it the plumbing to be able to leverage what these AI models are. So sometimes I have to be mindful not to be obsessed with the next versions or the upgrades. But really, how do we go back and fix the right data? How do we go back and fix the right processes? Because only when you do that is where we can start scaling and start seeing the ROI. We had this example to say, if we are doing a certain process in 10 different ways, and I apply AI on that, we'll probably have 10 different ways faster, yes, for sure. You now have 11 different ways. Exactly, and probably faster, but that doesn't help the goal of taking the cost out in the ecosystem. The idea would be, how do we say, what should we think differently on this process? Because maybe a few years ago, we couldn't think about it. And then based on that, how do we apply? Once we have standardized and harmonized, how do we apply AI to then really drive the structural difference? Yeah, so we could have the pace and development of AI is super fast. But you are saying we are actually trying to take it a bit more slow and look what we have already. And in that way, I stay ahead of others, because that sounds kind of like a contradiction. How can you stay ahead if you're not moving as fast as others? It's basically slow down to go fast, and this is also the third topic which we have anchored into our ambitions. That simplicity, we need to start embracing. And AI really has the possibility to take complex things and help drive simplicity in that. You mean simplicity for your colleague, for the team member, or is it for the customer? Where are you putting that weight? I think it will be across the board, right from customer capabilities to, let's say, everything in the supply chain, capabilities, but everything back office as well. And if I give you a concrete example, when I say customers, we say for AI is technology something that is better. experienced when you don't experience it. But today, if I look at that customer journey of how the traditional e-commerce, there is still a lot of friction. So if you are searching a certain website, you probably have to type four or five things, you probably have to navigate to five or six pages till you really come to where you need to be. Oh, so frustrating. Yeah. And then you have to go through the whole process to be able to finally put it in a basket and might be realized, no, it doesn't meet to the timelines, et cetera. It's easy shopping online, but there's so many choices. It's so complicated. Yes, sure. It takes time and effort. But with AI, I believe we now have the possibility to hopefully make it simpler and less frictionless for you. Yeah, because I mean, we're now used to using a cloud or chat GPT to ask everything, right? So everything from renovating our house or doing interior decorating, and there's now facial AI as well to do that for me. So take a picture. How does this billy closet look in my living room? Right? So are you finding that customers are already there? Is that a genuine source for you? Like, hey, customers are in those environments. So we need to be there as well. No, and then that's probably one of the biggest change we start seeing where the customer journey is transforming. So where historically, maybe an Ikea was selling us so far or a table or a chair, we are now saying, no, we are rather in the business of selling our life at home expertise. And really, what the life at home means for you? So we all of a sudden, we are asking, show us your living room, a picture of your living room. What's your theme looking like? What's your taste? What's your budget preferences? And based on those and based on our experiences of having designed many living rooms in the last 80 years, all of a sudden, we can provide you with here are the few options that you could really look into. So it's no longer take a tape table or a sofa or a chair, bring it home, and then discover, no, no, no, this is not to my like and bring it all over back. But all of a sudden, we give you a first impression through what we call Ikea creative, how the setup could look like. Yes, I've been looking at that. It was like I was playing the Sims. The game I used to play when I was a little girl, that you can organize a house and put furniture everywhere. It's it's almost like that, right? But then without the people. Yeah. And especially when you start looking at that, it starts, this is where I think the fact that the models are improving, you start seeing a tremendous improvement on the possibilities and what you could achieve. And of course, the key for us is how do we marry technology with our expertise and experience of having done this for many years. And that's where you see that some of the best outcomes. Because that was also when I visited an Ikea store and still do, then there's someone there who can help me with these exact same questions, right? So I have this hide in my house, what should I, what kind of closet should I get? Now that you can do this with creative, have you got any numbers on how this helps you earn money as Ikea? What does it mean in the numbers? Because you don't need all these people in the shops anymore to take all the time to help you out, why you can do it easily at home online. I think we are seeing a tremendous increase in engagement when we look at, because we today say 85% of that customer journey starts online. And this is where we engage, connect and inspire with the customers. But we also say one of our best experiences as Ikea is at our store. So if you still want to touch, feel, let's look at this experience. Maybe you have started to build your living room. We welcome you to one of our stores to touch and feel that room. But the good news is we also have the possibility of allowing you to speak to the many co-workers who can continue. The human assist to what has been started as an AI journey. And that's where we also see a tremendous engagement. You can actually bring your first designs, which you have started online on your sofa, bring it into the store and the co-worker can start that journey from there on. Okay, so the stores aren't going anywhere. They're saying we, Ikea, do believe that store is where our peak experiences. This is where you get to touch, feel, be creative, but also at the same point in time, we believe people are our biggest assets, bringing the human connection and the human engagement to it. And that's where we feel that this is where we can try the biggest experience. So we always say store is where you see the peak Ikea experience. But then, of course, based on your personal situations, you might decide to meet Ikea and we will show up in some of those meeting points, whether that's online, app, web or social. So that's where we have the A.E. parade of Ikea, because Robert, do you know the book Ikea, the Billy? Yes, I know. I'm an Ikea client since, since young, of course. Yeah, we live in Sliedrecht and that was the first Ikea. First Ikea, all right. So I often go to the floor and I'm very curious how that then evolves with the Ikea. Well, the Billy, you have to put it all in one place. Did you also have it? Sure, everyone. Yes, who doesn't enter that? Number one, Robert Hoezerhen, chatbot. So we've got to hear about what the Ikea chatbot does in terms of Ikea, all for difference. Around every experience of returning from Ikea, PWC has an insight of the researchers' Ikea ambition to be able to build a tonnable impact. Today, with Edwin van Boumou. A lot of organizations spread their I.E. budget over individual use cases. Coplopers in I, do it others. They invest as investors, focus on a few financial transformations and companies. EI, everyone talks about it. But how do you conceal the value of EI? We see what choices need to be needed for EI ambition to be able to build a tonnable impact. Go to www.pwc.nl. In 2021, the Ikea Group, the Billy EI chatbot. Of course, the Billy Cast, who was in touch with the I.E. chatbot. The chatbot came in in 2022, so they were already there. But we also saw the children, so we had to look at it. We had to look at it with the Chief Digital Officer in Ikea Group, Barack Barrack. It's Billy, just a chatbot, like every other customer chatbot. What is it, my goal of Billy? What we have defined is, there are many different intents the customer has while engaging with Ikea. And we are looking at each of those intents to say, "Where do we believe there is a possibility to really drive digitization and automation?" In this intent, when a customer is reaching out, questions like, "When is the store open?" "Where is my order at?" All of these are intents, where historically we had to spend a lot of time and effort from a co-worker perspective to be able to answer. And those are the intents where we are trying to say we digitize. So we do it by intent, and yes, the idea is to really put that at the service of the customer. Hopefully, making their lives easier, but also our co-worker lives easier. And almost everyone has a chatbot now, a I.E. chatbot, even probably most. What makes Billy now so exceptional that you make is the most important A.I. show piece that you wanted to talk about here. Probably I start with maybe a little bit of, we asked ourselves, sometimes when we look at technology to say, "Where can we use this technology?" And I think at Ikea, sometimes we ask ourselves, "Different, how can we put this technology to the service of what our purpose and mission is?" And Ikea speaks about the three roads. It's about affordability, accessibility and sustainability. When I speak affordability, I think we have some amazing products and we have made big roadways on product affordability. But when it came to design services that we were speaking earlier, we still are limited by the number of co-workers that we can offer in design services and are still limited by the time it takes per every room. So historically, if you had approached Ikea and said, "I want to design my living room," it would have probably taken us 5 to 8 hours minimum per room to design, right, from translating a 2D to a CAD to a room design with what we spoke creative, but applying the same technology to the capabilities that a co-worker uses for designing. I go from 5 to 8 hours to 30 minutes. Well, let's talk a little bit more about that co-worker and specifically re-skilling, right, because this is something that a lot of organizations are dealing with at this point. Is that, especially, for instance, in support or in other, yeah, more knowledge work. That is increased so they're not a domain of engineers. Any more AI that knowledge work is also being augmented with these new models and these [BLANK_AUDIO] to LLMs. How have you handled that? 'Cause I know that you're very focused on reskilling instead of replacing, what's the dynamic there at IKEA? Like from day to day, what are you doing to either give people those skills or replace some parts of the process? - No, I think part of it is where do we say, how do you create the awareness and give them the basic tools for that? You take the anxiety of AI away. So it's really the basic tools, access to co-pile, it's access to and get them to start using AI in there every day. So they start feeling comfortable and I'm sure everyone of us have these jaw drop moments where we have seen something AI do and I'm like, this is possible. And when the co-worker start having that, then they start thinking about, how do I apply that? - Right. - And then the other part is, maybe in some cases, some of the skill sets are no longer valid and you really need to, so in this example I was giving, previously when a customer came with a 2D version of their room layout, the co-worker had to spend time to create into a 3D version in a CAD setup. This can be done in 30 seconds at the click of a button on AI. So all of a sudden, I don't need the co-worker to learn CAD skills. - Oh, doesn't mean it's such a relief. - It becomes all of a sudden. And rather, they focus on what their expertise is on life at home experience. And there again, with the fact that we have done this over 80 years and designed thousands and thousands of living rooms, I can also give them a step-by-step view and call out in our internal tools to say, have you asked this question? Is this in the consideration and you make life easier? So if I go back with Billy, all of a sudden applying AI has allowed us to offer design services cheaper and to many more people than historically what we might have been able to. And that's where we have said, if we digitize some of the topics that a traditional customer service organization focuses on, we are freeing up enough time and effort to start helping on home design and services, which then also helps us on having that being affordable and accessible. - Well, what's the friction there? If I'm one of your colleagues and I love designing, I love that's the majority of your colleagues probably. How do I feel if AI now does that for me? Is there, what's the friction there? Can you specifically name an example of that friction and how you've dealt with it? Because I think that's good for listeners there. - It's a mix of AI assisted home design when we look at the co-workers, where we use like the translation of 2D to a 3D design. Then by AI, a quick eye from the co-worker to say, okay, does this look right? Is there, because AI is still learning. Many cases you find the windows in the wrong place, et cetera. But some of those inputs are necessary to make sure we continuously adjust the algorithm and course corrected. So one, you have to train them to say, it needs a second eye. Two, then AI sort of is making a first proposal. But the aesthetics is where you start bringing the human angle to it. To say, you now have spoken to the customer, you know what Robert sort of likes. How do you bring the emotional aspect to home that's been designed? Where could the pots go, where could the frames go? Is there something that needs to be adjusted? And marrying this, the intelligence, the AI intelligence, with the human intelligence and having a final version of it is where the opportunity lies. And is that-- - I believe that will always be the case is that the human will always be crucial for that step. This is where a lot of the friction is, right? With re-skilling and with-- - This anxiety perhaps comes from the idea that maybe in one day, it will be-- - Precisely. - Your face, right? - And this is where we try and address the anxiety. And as Ikea, we believe in the value that humans bring into the discussion. And this is where we still say, store is our peak experience, not just for the showrooms, but for the many co-workers that you get to speak to. And that's where we believe the differentiation is. For Ikea, we continue to believe that yes, our people will be our strongest ambassadors on what Ikea means from a culture value, but also what are home furnishing spring. And hence, we intend to continue to have a human in the loop as we continue to also digitize the organization. - But how do you manage that? Because you can tell this to people, but that doesn't make them less anxious, probably. People do not like change in general. They like to keep the things the way they are. So how do you actually manage to get those people along and keep them also engaged in this transition? - And I think this is where you practically have to engage them. So I look at this at three years. One is what we earlier spoke. It's about creating awareness and addressing the anxiety. That's probably the theory of it. Second is giving them the basic tools to start engaging. And we speak about earlier, we were speaking about, today we have 26,000 of the co-workers who are daily active users. How do I continue to make sure that more and more of the Ikea co-workers are active users on using AI in their everyday? Because when you do that, you trigger the curiosity on what if and what can be. And then if you do that, we have now started experiments with or what we call pilots or prototypes in Germany and in Portugal, two of our countries where we say, here is you are the digital ambassadors in the store. Let's look at some of the store processes. And together as a co-worker community, could you start rethinking what are the topics we can start applying? And the engagement we have seen, this is where the penny drops to say, wow, here are a few things we can change. Here are a few things we can implement. And then it's really a buy-in without having to do a top-down buy-in. And you see them engaging in the idea. So what you're saying is that adoption and really is stuck to collaboration, that if everyone has a chat GPT account by themselves and is in their silo, that that is not the way to do this, that it is really about like making that, you called it plumbing earlier, where I would imagine the right knowledge is in the system, right context, but also that the right tools are there for someone to consume it. Is that all built from scratch by you? Or are you looking outward for that? What is the. I think we are learning as we go as well. So I wouldn't say we have found the silver bullet on how best to do it. But we did start with the concept to say, we will have to bring our co-workers along this journey. Along the way, there were a couple of missteps as well. Well, the co-workers were also loud enough through our internal share points. Talk to me about that, what were they saying? What were they like raising a rat flag for? Like, this isn't working. The whole topic around anxiety, the whole topic around what this means for us, the whole topic around we need to be faster in bringing this, et cetera. A lot of callouts, which then you listen and then you adjust. Have we nailed it? Have we got this right? Probably not, but are we moving? And they're, yes, I feel so because we constantly measure to say, what's the active engagement? What are the movements we are doing? And then that's more the bottoms up. And then, of course, top down. We also have key matrices we use to say, what's the impact of AI on growth? And what's the impact of AI on cost? And we also start measuring it and reporting it organizationally, more as the matrices towards management as well. Sadly, we are already nearing the end of so many more questions. But one more thing is what I wanted to know. You tried AI plugins, for instance, and IKEA shop in chat GPT. So how do you think about this? Do you think this is maybe the future of shopping, but would be your position? I think what we call the space of agentic commerce is evolving. E-commerce was a discovery for a lot of us. E-commerce continues to be a discovery as we spoke earlier. And I think agentic commerce is one of the ones which will also evolve. You see the two or three key players in there. There's the customer who decides how and that experience should look like. There is the models and the engines, which own the search engines, whether it's the Googles, the open AI's or the world. They are the third player. And there are brands like us who are the third player. And I think in the next one, two years, you will see a little bit of an evolvement of how this shapes and what this means like and how will that settle. So I do believe that it's going to change, going to our discussion earlier on how the commerce journey is changing to take friction out for the customer. And once you take the friction out, the customer sort of starts to embrace it. And based on that, both the frontier. engines but also the brands like us will sort of need to adjust and evolve our approach accordingly. Any expensive mistakes you'd like to share about that IKEA made on this journey? I think we continue to learn on an everyday basis. Some of the examples when we early launched our co-workers saying no, we need help with A, B and C. A couple of areas where we are also drawing the guidelines but maybe some countries did go out and venture exploring AI use cases on topics which probably bother privacy where we had to also pull the plug to say no, not on this topic or not on this topic. Are you trying to behave or are you looking for where the frontiers, the borders of what we can do? No, we have started to be very clear on guidelines in the organization on where we will leverage AI and how we will leverage AI and we will share this as a small video also to say that's where we'll play. These are the frameworks it needs to start with our promise to the customer and our promise to the co-workers. Being reliable trustworthy is fundamentally critical so we have laid the frameworks but sometimes being the big organization you realize that maybe a couple of countries have experimented differently and that's where we have had to go back and say not in this space, we will not experiment with AI. It borders a little bit into privacy. This is not our use case and we need to start. The plumbing has to include the guardrails. It has to include exactly. It has to include the guardrails and then you decide where do you keep it standard and ask the organization to say this is how it is and where do you really ask them to be? How do you say entrepreneurial and go bonkers? Well, good luck on this journey, Perraq and thank you so much for being here. Perraq, Perrek, Chief, Digital Officer at Inca Group. I here, of course, thank you. Pleasure to be here. Thank you so much. Robert, where do you need me? I have a question for you. Now, how big are you? Let's talk about that. Yes, as a former customer who wants to buy the IKEA kitchen. There are just a thousand things that happen and you have to do everything. This plant doesn't fit in, you have to go back and I think that's the GAUS. That blinder panic in which you are as a customer. That must be the GAUS. Was it really a GAUS? Yes, that kitchen, yes, you have to get it off. And the plant is not worth it and then you don't have to help it. But it's also a role for this kind of GAUS to be launched through it and I think. And you see it back in Billy, but I think that in the future that you will also see it back. I think before I buy the kitchen, how does it look like? It fits well in that corner, because the big part that the kitchen can't work was because I've forgotten it. So yes, it comes out of the kitchen, Robert. What do you think about your own company now? Yes, I think what he says about the plant. I really think so. So it's really the way it is. It's not about the new model used and it's really about how it's implemented today. And it looks more like someone's drawing something and says, yeah, I have anxiety. So I'm going to draw something a little bit more. And then I want to say, well, these are the colors that you can move. This is the framework. If you want to leave it to me, I think that will have to happen in every big organization. This was Return on AI, the first episode of HumanX, the International AI Conferencing in Amsterdam. In the next episode, we'll hear you again, and then with Patrick Mans, head of AI Engineering at Philips. What's so interesting about that company, Robert? Are you happy? Well, in the meantime, there are a lot of questions that you want to ask. But the impact is also gigantic. And who is better about this than Philips? Yes, we've been looking at a 135-year-old startup in the data. What can we learn from that? You can now download the application as a full-length application in the browser, surely. If you have any questions about us or about a company, send them in via the BNR app. And subscribe to Return on AI in the following episodes, don't miss it. What you also thank, of course, and see you next time.