Building for the Age of AI with Yahoo

AI enables a different way of thinking than we've traditionally interacted with tools. That mindset shift is so significant that you have to really respect it in thinking about transformation of a workforce. If you create too much structure around AI literacy early on, people manage to the metric instead of really kind of deeply immersing themselves and kind of celebrate the culture of experimenting with and trying these tools across the company. Welcome to the Coursera Podcast where we have conversations with renowned experts and industry leaders to explore global trends impacting the future of education and work. I'm your host Arunab Senna, Vice President of Global Communications at Coursera. Yau has been a household name on the internet for nearly three decades, but it's not standing still. Today the company is reinventing itself with Generative AI at the center of how it builds products, serves users and supports its workforce. From male to news to ads, it's about creating smarter experiences and a stronger learning culture. To discuss all that, we have someone leading this transformation, Matt Sanchez, who is the chief operating officer at Yahoo. He oversees co-products like Yahoo, Mail, News, Search and many more. He spent his career at the intersection of technology, media and advertising and today is helping Yahoo chart its next chapter by rethinking how products are built and how people grow within the company. In this conversation, we'll talk about Yahoo's product transformations, it's approached to AI native design and how the company is investing in the skills and mindsets needed for what's next. Matt, welcome to the Coursera Podcast. We are so excited to have you, especially as a former Yahoo who spent nearly six years at the company. Thank you, Arunab. I'm very excited to be here and looking forward to the conversation. Absolutely. Let's dive in. Matt, I was looking at your background. You have a fascinating journey from entrepreneurship to a product leader to now CEO at Yahoo. What's been your focus in stepping into the role earlier this year? A lot of that background has been tremendously helpful and kind of instructive in thinking about how to not just, you can breathe life into a 30-year-old company that has tremendous scale and storied products that hundreds of millions of people use every month, but had tremendous opportunity for investment and reinvention. That was really kind of at the core of why I came to Yahoo and why I think Apollo got very excited about the opportunity to take the business private and invest in the business as well. A lot of what I've been doing, stepping into this new role as CEO has been a continuation of the work that we've been doing over the last several years. We've been working hard to just scale the investment that we're making back into the consumer business and really kind of breathing life back into these great products that people are using daily in their lives to help achieve their goals. We saw lots of opportunity there and we're excited to be relaunching and reinventing these brands. That's been about a belief that reinvestment can drive growth and we're seeing that in these businesses and what I've been focused on is just how do we continue to do that good work at scale and make sure that that's really kind of connecting for users across our entire portfolio. That's incredible. You rightly said that you have some of the story products and some of the most recognized internet brands. As part of your role, you also oversee Yahoo Mail, Yahoo Search, Yahoo Home, News. Each of them are serving as a key point for millions of users. How do you use AI to personalize content and improve discovery? I know that Yahoo has been home to Hadoo and a lot of other data science innovations which have become mainstream have gone on to become multi-billion dollar companies themselves. What lesson does that teach you about building AI that really understands people? It's a great point. Yahoo was one of the earliest pioneers on the internet. There were very big problems that had to be solved at scale that led to things like Kadoop and other fundamental technologies that helped power the internet. The story history of Yahoo contributing to the overall technical foundation and growth of the internet. It's a legacy that you're honored to be a part of and it's exciting to build on. All of that has been in the service of building products that were helped guide users through the internet and help them achieve the goals and accomplish goals that they were trying to manage in their digital lives. That was true in 1995 and that is true today. One of the most exciting things about this chapter and one that wasn't even real when we took the company private but is now very front and center in what we're doing is the role that AI can play and just the capability that exists because of this massive compute opportunity that allows us to create magical experiences for consumers. What we're thinking about is how do you take the job to be done or the core user needs or use cases that we are trying to solve for within each one of these products and imagine what could be possible with the capability of all of these all-elems that are advancing at rates that are just mind-boggling. Coupled with our understanding of what users are trying to solve in their lives and how could we then create experiences that just feel intuitive and solving for their needs. That might look like a quick actions out of your email so that you can very quickly accomplish a task. It might be sending a reminder or responding to an email or checking into a flight or you name it. There are things that used to be many steps that a user had to take and we can start to really collapse that into a much more functional capable way to engage with our products and get things done really effectively. That might be helping understand an article or the context around an issue much more richly. It's taking the articles that you're reading and actually just expanding on that or building more context around that with timelines or summarizations or other things that are possible today that would have taken a tremendous amount of work before it to add to an article. There's opportunities across the portfolio that we can really just supercharge and then make more intuitive and capable the use cases and problems we are already solving. Can you actually talk about how some of the concrete use cases in terms of what are some of the most exciting ways in which you are using AI to improve the experience of users and customers? I think some of the work that we're doing in mail is one of the places I'm most excited about. We just want to catch up feature that allows you to really quickly move through the emails in your inbox and with a quick summary, almost sort of gamify your ability to quickly sort through and either throw out or mark as red the messages that you can move through quickly so that you're really saving time for the message that need the most focused actually respond to. Anything else we can get you through that in a very efficient way or like I said, quick actions or prioritization of emails starting to think about this kind of agenteic capability that there's lots of conversation about where we can then kick off actual task management and take steps for you on the basis of your email. Those are all ways that we make you a more effective consumer and save time and just ultimately help we call it talk about it as managing the business of life. That can be really complex and there's a lot coming at you and there's lots of things that you might miss or this is so much power in these tools that can actually help manage all of that for you and simplify that down into something that you just feel in more control of. In fact, on Yahoo News, I believe you, have the key takeaway feature that has made it easy for people to get informed quickly. How do you think Genette is changing the way people consume news per se and how is Yahoo leading into that shift? It's a great point and actually one of the interesting things that we see about the key takeaways of summarization is it actually leads to more engagement because sometimes a whole article might feel overwhelming but if you can quickly get into a frame of mind of, okay, this is interesting to me now I want to learn more. It actually is an entry point into diving deeper into an issue. In news, you have this product challenge of scanner to engage or of reading headlines and then deciding if you want to actually engage in an article. There's ways that we can then actually help you step into that and create more of a glide path into actually going deep on a topic. From there, we see opportunities to what we think about as derivative content or other modalities you might want to actually consume that content. That might be a slideshow or a video that is related to the content of the topic that you're consuming. Being able to quickly present that type of information to you or if you are interested going much deeper into an issue and doing research or chatting about a topic or looking for a timeline, all of those are capabilities that we're looking at as ways to more richly surround a particular topic that you need to understand as a reader or that you're interested in with all of the contexts that's going to help you dive into that issue. These kind of summarization technologies can be very powerful on one hand that can be in hook to what's consuming some broad set of content. On the other hand, that can also lead to some challenges in terms of how the summarization has happened, what kind of misinformation that can and what editorial policies are in place. How do you strike a balance? Because one of the things which is unique about Yahoo News is a highly trusted brain and that's a responsibility as well in many ways. So how do you make sure that when you're deploying these technologies, you have a very strong checks and balance in place? That's something that we take very, very seriously and are very much a consumer centric in how we're thinking about what we're delivering in the product and the kind of value that we're putting forward for the consumer. On these types of issues, there's a lot that's evolving in terms of the way that you can put safety features around what it is that you're doing and round your response or your workflow in either the article or other articles or content corpuses that will help make sure that you're actually delivering something that's useful. Also lots of signaling that we're doing to get users to respond to that in flight. There's a whole set of things that we're doing both in terms of continuing to evolve prompts and workflow and the way that we're actually getting to an output that we think is something that represents the value that Yahoo is trying to deliver to a consumer. Also that feedback loop of making sure that we're then flagging anything that we need to look at just like you would any type of content process. You're absolutely right. That's a rapidly evolving field and I think one where the rate of change and even the capabilities of these foundational models is moving unbelievably fast. Maddit, there's one company who will get it right is going to be Yahoo because I remember some early work in the span detection, for example in Yahoo Mail. I think that became sort of a pioneering effort that guided the standards for the entire industry and I'm sure that you know, yeah, we'll do a wonderful job with the news as well. If you take a step back, how do you see AI changing the way people interact with digital prompts more broadly and what role will Yahoo want to play in that future? It's a great question and one that we talk about a lot internally because the power of these tools and the rate at which they are becoming more powerful and useful is hard for us as humans that really, really contemplate. I'm not sure that in our lifetime we've seen anything move as quickly or as substantially as what's happening with AI. And so kind of at its core, that means that all kinds of things are possible that we wouldn't have imagined real before. You know, whether that's dynamic UX or things being created on the fly very tailored to what it is that the user is actually looking to do or the problem they're trying to solve or you know, the gentick workflows that can actually go and accomplish tasks for you that you otherwise historically would have taken a lot of time to go complete, make your reservation for dinner or planning a travel excursion or whatever those things are. And so a lot of what we are trying to do is go back to our roots of how can we best serve as that trusted guide to the internet for users and how do we, in the categories that we're playing in, deeply understand the goals that our users are trying to accomplish and how we can, how we can play an important role in helping them achieve those goals. And then from there, you know, that's then about that understanding of the user and what they are doing on Yahoo coupled with all of that functionality. And we're trying to then imagine, you know, what are really out there experiences that might actually help serve that mission and do that in a way that feels just intuitive and like you've almost kind of anticipated that that users needs and are solving that for them. And I think this marriage of the depth with which we understand our consumer and all of the ways that we're helping them, whether it's kind of understand what's going on in the world or manage our inbox or manage their portfolio or deep fandom for their team or their fantasy league that they're in and then build very rich functionality on top of that. It's marrying that data and inference compute or the kind of foundational model capability in ways that we think are going to totally change how consumers understand consumer products. And the same way that when we moved to the iPhone and suddenly we expected great design and we expected like completely different level and depth of interaction with the apps that we were using from what we were used to on the Blackberry, I think you're going to see a similar kind of thing with consumer apps that are AI powered in very different ways. And that's not about just a bolting chat on. That's about deeply embedded capability that just changes what's possible for a consumer to accomplish in a particular app. So a lot of this technology innovations will be driven by the talent within the company. So I want to shift gears to that a little bit and Yahoo has a long history of supporting employees and learning including technical upscaling and online learning partnerships. As Genie and other technologies keep evolving, how are you helping people keep growing and whether that's through training or sort of internal mobility and shifting of roles or opening up new career paths. How do we make sure that Yahoo remains a premier place for talent to come and work and innovate on behalf of their users? I think you have to start with the sort of foundational premise that AI enables a different way of thinking than we've traditionally interacted with with tools. And so that mindset shift is so significant that you have to really respect it in thinking about transformation of a workforce. So it's not just about saying, here's a tool or go read this white paper and you're going to be good. What we're finding is it takes actually getting your hands dirty and really trying these tools and working through problems for yourself both in your life and in workflow tools you're using work and starting to kind of shift perception of what's possible and then experiment. And so we're trying to both make as many tools as we can available as well as almost kind of celebrate the culture of experimenting with and trying these tools across the company. So we just challenged every group in the organization to within their specific teams actually go try one particular application that they could kind of learn and talk about and kind of engage with. And so that's everything from the comms group to engineering organization. You know, it's relevant across all of that. And then one thing that we're finding is particularly useful is taking the teams that are furthest along. We kind of have a generational model of where we're seeing teams, particularly in product kind of understanding and then integrating and then actually shipping and iterating on the AI capabilities. And where those teams are furthest along, we're starting a rotation program. We're actually bringing people into those teams for two months, giving them a chance to really understand how differently that particular team's working and then bringing that back to the team that they've been historically working with. And that actually is the thing that I think is having the most success for us right now just because it gives people exposure to a very different way of thinking about solving some of these problems and then they can kind of bring that back. But the zero to one challenge here is very difficult just because the mindset shift is so significant. We have to surround everything with resources and tool availability and the space to really experiment and kind of challenge norms. But then also a little bit of exposure to ways that people are working that are very, very different so that you can then kind of think about how to incorporate some of that learning into your own practice. How do you measure the impact of these sort of AI literacy implementing your workflow type initiatives? Are there some productivity goals that you have assigned to this? Are you looking at your rate of innovation and how fast you shift products or clearly absolutely new sort of AI first ideas that you want to bring to life? We haven't necessarily formalized it as explicitly as what you're framing. What we are trying to do is create more of a culture of curiosity of people really embracing this in ways that's surprising us in some of the adoption and ideas and ways that people are aging these goals. I think this sort of challenge of the other side of the coin on you get what you measure is that if you create too much structure around AI literacy early on, people manage to the metric instead of really kind of deeply immersing themselves in what's happening here. But I think where you then start to see that is velocity. And so you can start to spot the places where there is acceleration of product development or kind of innovative ideas or capabilities that otherwise wouldn't have been possible. And then you really just have to celebrate that and reinforce that and that helps drive transformation. That's actually a very instructive point for companies which are looking to adopt AI. Every now and then, right in the beginning, if you start putting a lot of hard metrics, as you rightly said, it kills the spirit sometime. You have to allow to an organic process of adoption in caragement, culture of curiosity as you mentioned. And as you move along, then you start putting some metrics behind and some outcomes and impact data behind this. So that's a very useful framing. Thank you very much. Let's talk a little bit about the pace of change, which is only accelerating. And it's just not just AI, but also everything that's happening around the world. You know, some of them require hard skills. Some of them require different kind of skills. What skill do you think will matter most in the years ahead? And how can, you know, not just people at Yahoo, but also workers in general around the world can start preparing for them? It's a very important conversation that everyone from rising students to folks that are deep in their career and the workforce have to really engage with. And I think that we are increasingly in a phase where some of the meta skills are going to be much more important. And, you know, the two that I think, you know, for me, are probably the most important are curiosity and resilience because of platforms like yours and YouTube and others, there is so much information out there about various topics that you might want to learn about or skills that you might want to build. And to your point, the change in just world in general, technology capability, consumer expectations, how consumers find or discover, gauge with products, what a brand means. All of these things are moving at such a rapid pace that you have to be able to sort of thrive and change instead of that causing you to shut down. So that, that sort of underlying resilience of, you know, just one foot in front of the other is keep actually kind of evolving or learning or adapting to what's happening around you. And really trying to understand that's where I think curiosity point comes and understand what's happening around you so that it's not about being a detractor or a resistor but actually kind of leaning into where does that create opportunity or how does that actually shift things. You know, we're talking about things like search that we assumed were kind of almost game over. This might be the most, most interesting time in that space in decades and no one would have expected that five years ago. And that's just, I think, one example of the kinds of reinvention and transformation that are going to create all kinds of new opportunity. If you think about sort of what's the next platform that's coming, whether it's augmented or audio or anything, there's lots of ways that we might engage with information differently as all this incredible inference capability that starts to really come to life. That could create entirely new platforms that we're going to, you know, get to reinvent how we engage with consumers around. And so it couldn't be a more exciting time but you have to really be ready to lean into the change. In fact, your point about curiosity and resilience is something that we see a lot in terms of learning trends and demand of our learners on our platform as well. So on one side, you definitely have, you know, top courses on deep learning artificial intelligence data science, project management, cyber security, equally popular, man, courses like science of well-being from Lodi Santos at Yale. Yeah, it is. Which is an award-winning work she has done in terms of how you manage through change and, you know, disruption, big stretchers work on purpose driven innovation and a work at, you know, Michigan legend was done. And these are some of the most popular courses on our platforms. It's kind of very redeeming to see that yes, on one side, everybody wants to learn these skills on the on the hardest side and technology side, but they also want to build resilience to deal with this change and have being pathetic and contribute in a meaningful way. I love those sets. That's great to hear. Yeah, absolutely. You should check it out. As we come to the end of the podcast, I want to know what's one way you have had to grow personally as a leader to keep up with all this change and disruption and transformation. I would go back to this curiosity. And I think it's not just trying to understand new technologies and ways of interaction. And, you know, I try to carve out 30 minutes or so a day that I'm actually trying something new, whether it's cursor or new technologies that are coming out or etc. And I think that at least gives me context for grounding to try to understand what we're working through as an organization. But I also think, you know, as we as we age and the culture moves away from us and really trying to understand, you know, what's what's happening with kids or with youth or like what are the challenges that they're doing with you know, I try to really look at how my kids are growing up and evolving and try to kind of understand what, you know, what does the world look like through their eyes? And I think that that's that's as important because it's a little bit of a leading edge indicator kind of where all this is going to go and how things are going to change at every level. Yeah, socially, technologically, from an economic standpoint, you just have to like embrace that they're things are changing faster now than they ever have in our lifetime or maybe in history. And you're just really trying to make the space to understand that has been something that I've tried to make a priority. Thanks a lot, Matt. That was a wonderful conversation. Thank you so much for giving us a ringside view of how you are sort of usharing in the legacy of Yahoo and the story of Yahoo into an AI first world. Thank you so much and I really appreciate it, Convison. Thank you very much for having me. Really enjoy the conversation. Absolutely. If you're looking to build skills in AI digital transformation or leadership, head to Coursera.org. You will find many of the same learning paths that's leading companies like Yahoo are using to invest in the people and obviously you have the wisdom of Matt as well to guide your course. If you enjoyed this episode of the Coursera podcast, please rate, review and share. Until next time, I'm Arunab Senat. Thank you for tuning in. [BLANK_AUDIO]