 Well, I paid for this party, so I kind of got a seat at the table twice. But right now we're going to share the stage, and I'm very excited about this topic, because I'm sure with all her magic happening with chat GPT, Google bar, we're all getting hyped up, right? We've seen this movie before with other things such as Web3, NFTs, crypto. And we're at this peak right now, kind of questioning, okay, but what's really next? And what are some of the real applications, especially in product? So I'm here with three product executives that are going to give us their take, the real take. We have a pretty sophisticated audience here, so we can go deep just to make sure that this is for real. As you know, I'm the founder and CEO of Product School, and I would love for maybe the rest of the audience to, the rest of the panelists to introduce themselves and add one more thing, which is from one to five, how excited, if you had a crystal ball here, what do you think is going to happen with AI? One being, ha, it's just another trend, it will fade out, five is like, oh, my God, this is happening, it's a paradigm shift. So maybe you, Jessica? Hi, everyone, I'm Jess Hall, I'm the Chief Product Officer of Just Eat Takeaway. Great to be here today. I'm definitely at the top end of the scale. I'm going to go for a five. I think it's super exciting. Iway. My name is Iway, Chief Product Officer at Talabad, and part of Delivery Hero, very similar sentiments. So over the last two or three months, I've been diving deep into the space. I'm a solid five as well. Shabayu? Yep, my name is Shabayu, and I'm the Vice President for Product and Growth and Tier, and I would go for a six. Oh, wow, how doing us all. Let's ask the audience. Let's start with the optimist. Please raise your hand if you are a five plus. All right, maybe 50%, four, not bad, three. Still decent? Okay, pessimist, where are you? Number two, still some, one, who thinks this is just a trend that is not going anywhere? Okay, I see some. You can change your mind during this conversation. So let's go to the nitty-gritty of it. Maybe you guys could share a real use case for AI. Maybe you are already applying in your current company, or that you see could also be applied for other companies in product management. I mean, for me, it is less about the specific use case, but just the absolutely exponential pace of change in technology that we're going to see as a result of AI becoming more and more used. I think right now we're still learning what those opportunities could be. And I think that's exciting. It's the scale of the opportunity, which is amazing. Where we are focusing in just each takeaway is on really providing training and opportunities to engage on this subject, because we want to be more efficient as a business. We want to find those opportunities, but we also really want to upskill the people in our teams so that they understand and that they can make the most out of this technology, and we want to make sure that we're developing our talent for the future. So for me, it's really about the scale of the opportunity, and I just think that the change over the next five years is going to be immense. Yeah, very similar sentiment. At Talabad and Delivery Hero, we operate in a very complex world. We serve our riders, our customers, our restaurants, and what's really amazing, and for me, what distinguishes a technology that's a bit of a hype and a real technology is how quickly you can start prototyping and testing and implementing new technologies to see if it actually makes the lives of your customers better. And what's been really exciting over the last couple of months is that we've been testing, we've been trying a bunch of different things, and whether it's using chat GPT or a bunch of other technologies, we've tried things like, what would it look like if we helped restaurants made it easier for them to market their products? What if we made it easier for them to write a menu description using some of these AI tech? And very quickly already, we've been able to take things to market. And that for me is the exciting part about AI. Not too long ago, it was a bit of a scary topic. I remember we would have our data science team trying to explain the concept of a neural network and how it all works, and there was a small fraction of people who understood it. But now it feels tangible, it feels usable. It feels like we can jump in there and actually see what it can do. We can run tests behind it. And I think that feeling of realness is what I'm the most excited about and giving this a shot over the next couple of years. Yeah, for me, it's about like chat GPT is everywhere, right? So basically it's about text analysis and how we kind of enable growth. People are mostly looking at it from a marketing perspective. I would like to see more in the deep learning space, the CNN in particular, where we are working on geospatial imaging. And in the medical field, imagine the power of AI that could no pave ways for cancer treatment, right? We are so fixated on the text base, but I think there's a lot of scope there. All the believer being a part of tier and sustainability for good. And I hope that this field catches up pretty soon. Suvai, you were at Amazon before, right? Yes. And they have a big team that works on this technology. We're facing now the classic buy versus build with AI. There are models out there that can be implemented already. Chat GPT, me, one of them, but they will see engineering teams really working on their own proprietary systems. So maybe starting with you, what do you think are some of the common misunderstandings in AI for product? What is the opportunity versus the fluff? Yeah. So I think one of the things as product leaders is we get scared like a general public. We get very scared with the term AI because it's kind of a black box, right? So I think first to Eva's point, like be bold, right? Dive into it and understand, and then you'll probably realize that it's not that scary. Most of the AI development happens from a solution led first, rather than a problem space led first. And then we try to fit in the, which problems does the solution fit into, right? As a product leader, I would like to have mechanisms and ways of working to manage that problem. And I think in Amazon, we really nailed it. There are AI PMs, tech PMs who know exactly how to kind of do that. I would like to also see large scale implementations of it. I think I would also add there that some of the misconceptions around this are that it is this miracle thing and maybe that's why some of you don't wanna go with the five score. It is imperfect because it's trained on the data sets that we're giving it and it's trained on information from society. So I think one of the other really important parts of the product role is to be aware of that and to be taking action to correct for those biases that are in the data sets that are in society today because the speed at which we will advance with AI means that if we're not cognizant of that and we're not making decisions about that, that there's potential for those things to become even more embedded. And I think all of us as product managers, as product leaders have a duty to make sure that we are really thinking about that as we develop products with AI. So just to follow up on what you just said, yes, I see a lot of startups now throwing the magic buzzword. I said, yeah, right? Back in the day it was like, we do machine learning. Now they are like, we do AI. They're kind of throwing AI to whatever they do and even their domain is .ai instead of .com. So from your perspective, what are some of the real wins that some of the companies can leverage today to start showing real value for these AI applications? I mean, I think it very much depends on the business and the customers that you're interacting with. The first place that I'm really thinking about is that I guess there's a bit of a fear isn't there that AI is gonna replace jobs. And I don't think that's the case. I think it's something that we've got to learn to work with and I think AI allows us to scale our cognitive abilities. So I'm quite excited about how we can use it to improve like technical tasks that involve human interaction today and remove that and allow for like more purposeful, exciting work for people. Like that for me is the beginning and the understanding. I think there's loads of applications in customer service in providing great testing and other elements that help us really learn quickly, learn quicker than we do today. Imagine if you could use AI to run experimentation at like way faster pace than what we can do with customer testing now that might take us two or three weeks or longer. So that's where I'm really looking for the value in the beginning. And I think it's gonna grow from there. I think it's very important for almost every function in an organization to stick their finger in there and see what it can do, right? Between a finance team, between a customer service team, between a sales team, a marketing team to write copy. I think it's a real opportunity for all the different teams to sort of figure it out. And I think over the next couple of years you'll see a verticalization of a lot of the technologies. I think you'll see some successes, some failures. Some companies making some very funny errors along the way, but that's really part of what breaking edge technology looks like. Yeah, I think for me at the risk of probably repeating myself, but what I'm really passionate about is when we are always trying to monetize AI, we're always trying to optimize the funnel, right? For AI. And that's fine, though I call it greed, but at the same time, we never thought of ourselves as customer, as product people, right? How does AI enable us? And I think that's power, right? And I think we should kind of, that's why we should embrace the technology and do that. I make sure that our team spends most of the AI into discovering the problem space and not moving into the solution. So kind of one of my principles is if all the AI is a deployed customer facing, there's something wrong going on. I remember 12, 13 years ago, I was a computer science student, I had a subject on AI before it was cool. And that was horrible. It was literally all about coding and trying to apply some intelligence to a video game. And now it's cool. And to your point, Yiwei, around sticking the finger on each team and see how they can leverage AI today. And hopefully there are ways to do it in a non-so technical way. So I'm curious from your experience, for product teams, how can some of the product folks in the room who might be thinking, okay, how do I learn AI? Do I need to code? Are there any other ways for me to kind of experiment with it without breaking the system too much? Like, what's your take on that? You know, actually, I mean, you were just discussing the importance of re-educating your team and your workforce. And I think we're on that journey. And I think the beautiful part about it right now is how accessible the technology feels. ChatGPT for its highs and lows and pros and cons makes it really easy for you to boot it up, throw some prompts in, throw some commands in and see the potential of what it can spit out, right? And so, you know, funnily enough, as product people, I don't know how many of you guys have tried to actually use it in your day-to-day jobs, but you know, when I walk down the hall, I see product managers with ChatGPT open and, you know, may or may not be writing their user stories and PRDs using ChatGPT. But I mean, I legitimately think that by testing it on our own, we'll be able to see the potential, right? And, you know, not in the not-so-distant future, I'm sure there's gonna be a customer service rep who also thinks, hey, I can make my job a lot easier if I throw in a customer query into this thing and see what happens, right? So I think a lot of that sort of skunkworks, bottoms-up initiative needs to happen and organically because of the accessibility of the technology we're seeing that already. I completely agree. I mean, ChatGPT is the one that everyone's talking about, but I'm talking to people in my life, including my dentist recently, who was telling me about all of the tools that she's found and she sent me a list of things that she found that were helping her with her website. You know, these are people who are not computer scientists or even product managers who are really getting in and experimenting with this technology. It has become so accessible that actually, it's in the hands of our customers and I'm a big, big proponent for getting close to your customer and getting different perspectives on things. So compared to other advances in technology in the past, I think this is so different because it is really in the hands of everybody and it doesn't take a great deal of specialist skill to start to interact with it and even go beyond just prompts in ChatGPT to, you know, there's all sorts of Google Sheets that you can download that will, you know, write you a talk or not that I wrote mine. Or do your SEO for you or, you know, whatever. So, and I think that's really exciting too. I think that's probably one of the misunderstandings today because it feels rushed suddenly because ChatGPT released their public version. Surprise, surprise. Week later, Google releases their own AI bar, right? And now a bunch of people are talking about AI as if they had to because it's hot. However, we know from experience that there are a lot of companies that have been working on this type of technologies for a long time and that is starting to emerge. So for those teams who might not have the resources to invest in AI from scratch today, they don't want to hire AI engineers or AI PMs. Have you seen any specific APIs or tools that would allow these companies to start showing the value without over-investing? Yeah, I mean, there are so many open source AI's available tools, especially like Google. When you Google up, you can find a lot of geospatial information about a particular location for customer, if somebody is using a mobile phone and you have a GPS location, you can have access to about that location instantly just through API plugins, right? And then there are so many platforms that enable you to plug in APIs like million APIs, right? I don't want to name the companies, but then those are the smart decisions as product managers we need to kind of or as product leaders we need to take, but there are so many. And even Amazon, Google, most of them, they're all about open source APIs, so yeah. And if I were to make a bet in the next five years, we will see a massive rise of companies that are able to make this very easy and accessible and verticalized. So you want to dig into locations, there's probably a company that will have a verticalized version of this that solves for that problem. You want a solution that solves for customer service and make sure that people respond in half the time to extra quality, there will be a service for that. You want one that helps you detect for fraud, there will be a service for that. I think the next five to 10 years we'll see a remarkable rise of just AI technology being massively accessible that I think, there's a good chunk of this room that will somehow end up in some of these product roles and a not so distant future. I've seen this movie with product tools before and it became a commodity. Now we don't have to reinvent the wheel to build a roadmap or a data analysis. So I agree with your perspective. Hopefully it gets to a point that we don't have to talk about AI because it's so embedded into our products that it's just a default. For the other two, what are your predictions for the next, I would say five years? So one AI is just, so like as we said, last two decades has so much happened and what you're seeing, the chadgivity, et cetera is an outcome of it, right? But don't forget the series that I'm the, sorry, I'm all about Amazon now. So the series in Apple, right? So those are also ones which came. So every 10 years you would see, but I would expect product as a service where you would embrace to you as point to all the APIs and all the instant plugins to instantly understand the customer will become so easy and then that will enable small cross functional teams or two pizza teams to unlock extreme potential. That's the space I'm looking out for. I mean, I agree with everything that's been said so far. I think we will interact with AI every day in five years time and the key skill we will have needed to develop or the key skill we are developing is how we interact with it, how we use it. There are tools already out there like co-pilot. I think we're gonna see more of that and to touch on your earlier point, I think we need to start thinking about AI as just part of everything we do and not a specialist subject and not a skill that sits to the side that we sometimes bring in. If we're really gonna achieve the maximum potential from it, we need to be thinking about it in all of our products and all of our discovery and all of the different applications. Like you said, it's not just about customer facing, it's also about how we operate internally and what we can learn that way. I love what you said about AI as a service, comparing this with software as a service. Hopefully it becomes that type of understanding for builders. In the few minutes that we have left, I want to use this opportunity to talk to the people that gave us a one or two in terms of their... optimism here. What if it doesn't work? We've seen movies where talked about virtual reality. It was going to change the world and it's still unclear. We saw NFTs not so long ago, we saw blockchain not so long ago and now it seems like we're throwing a new term out there called AI. So what is the worst case scenario? What if this doesn't work out? I mean, what is the worst case scenario? If it doesn't work out, we're going to move on to the next thing, right? But the difference for me here is how accessible this is compared to blockchain, for example, which the general public, people who don't work in tech, struggle to understand or really had to work hard to understand. Right now we're talking about a website that literally anyone can navigate to and put in a prompt and get an answer. And they don't have to go very far to find, learnings on how to make that better. And I think that's the crucial difference here and I think that's what makes it different from some of the examples that you've given. And I think one beautiful thing about the solutions that we're seeing today is that it's complementing work that we're already doing today. It's not a new frontier. It's not a new way of... It actually just solves a lot of problems that we're currently having today. I think that's very powerful. So I think for those who maybe gave a one and a two, my observation as well sometimes is that, I mean, chat GPT today says some pretty silly things. Make some mistakes and we've seen public demos from really big tech companies that didn't quite work out. But I think that's all part of the learning process and as we hone things in as some of these models are able to fact check themselves and it will just come in a matter, I believe, of months. I think we'll see a much higher bar of quality of products that will come out. Yeah, for me, it's like it will not work out if we go on, like we will replace everybody. AI will do everything. It will definitely kind of not work out in the next five to 10 years at least, right? What will work out is if we have realistic practical goals, right? Of leveraging AI to their strength and leveraging humans to their strength. Our strength is our connection. Our strength is building culture. Our strength is vision, right? I don't know if that's a strength for AI. So we play to our strengths. AI strength is processing large database and if we learn to coexist, to me it's a win-win. Thank you all for sharing your thoughts, the good, the bad, and the ugly on AI. Please just give it up for Jessica, Giwei, and Subayo. Thank you.