How Yahoo is Taming the Modern Digital Wilderness with Lee Zen, CTO

(Intro Narrator at 00:00:00) Today, we're talking to Li Zen, CTO at Yahoo, about what Yahoo is up to thirty years into their journey as one of the most trafficked websites. Thanks to Fin for sponsoring this episode. To learn more about the number one AI agent for customer support, go to fin.ai/cto. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:28) I want to know what's going on at Yahoo. Like, everybody knows Yahoo. I think you just celebrated your thirtieth anniversary.

(Li Zen at 00:00:34) Yeah, we did.

(Joel Beasley at 00:00:36) Yeah. So what's going on now? What's the difference? I haven't thought much about Yahoo from being a kid. I'm 38.

(Li Zen at 00:00:43) Yeah.

(Joel Beasley at 00:00:43) So being a kid, seeing it, being one of my first email addresses I've ever had. Fast forward to my mid-twenties, I used it for some financial data. And then fast forward to now, what's been going on?

(Li Zen at 00:00:57) Well, yeah. Well, hey, thanks for congratulating us on the thirty-year anniversary. It's a huge milestone. I think not many companies make it to thirty years. And, you know, it is a very storied internet property. It's a top-five U.S. internet property still. And so even though I think the way you were like, "Oh, you know, haven't really thought about it in a while," most Americans interact with Yahoo every single month. Something like 90% of the U.S. population visits Yahoo at least once a month in some way. And it's funny because I've had people where I took this job were like, "Oh, yeah, I guess I do use Yahoo." And I just think it's so ingrained in the fabric of what people do that they've almost forgotten that they're using Yahoo. Because whether you're using fantasy sports or you're checking a stock quote or you're checking a piece of email or whatever it is, you know, looking at your news, you're interacting with Yahoo at some point, and I think people often tend to forget that, I guess.

(Joel Beasley at 00:01:55) You're exactly right. Because that's me. I mean, I use your stock quote stuff every day, and I didn't even think about it. And then I also see Yahoo News gets piped into different applications that I use. I see a little Yahoo News logo. So those two things are in my life on a weekly basis.

(Li Zen at 00:02:08) Yeah. People interact with our weather app and kind of, "Oh, yeah, did I check the weather today?"

(Joel Beasley at 00:02:11) I do.

(Li Zen at 00:02:11) Check the weather today.

(Joel Beasley at 00:02:12) You do have a good weather app. I do remember that. Yeah. That's crazy.

(Li Zen at 00:02:17) But going back to your question about kind of what's new, I think really the new thing for us is we're taking a very modern take on, you know, the original mission of the company. Right? So the original mission of the company was to help people navigate the digital wilderness. That's still what we're doing today. I think one of the things, I guess, that's unfortunate, let's say, about the way the internet's evolved over the last thirty years is it's actually gotten more wild, let's say. You know, if you think about going back thirty years ago when Yahoo started, that was the mission. It was, "Hey, there's this vast digital wilderness that's unorganized. Let's help people figure out how to navigate this thing, how to find the right resources," et cetera, et cetera. And you would think it's gotten better over thirty years, and Yahoo's obviously done a lot to improve the internet, but actually, as much of the gains as we've made over the last thirty years, the internet has been exponential in becoming more wild. And so, you know, still with the original mission that we're trying to tackle here, but with a more modern twist, let's say, we're using AI for a lot of our things. We're using the cloud. Right? So it's a modern take on that original mission of how do we help people live their lives and navigate that digital wilderness.

(Joel Beasley at 00:03:32) So every day, everyone at Yahoo is waking up with this mantra of taming the digital wilderness?

(Li Zen at 00:03:38) Yeah. We want to be that trusted guide. Right? We want to help people really accomplish what they want to do with life, and that's why we still focus on those core areas I just talked about. Right? People are trying to have a healthy financial life. They're trying to manage their lives through mail and calendar and, you know, make sure they have their life in order. They're trying to be entertained. They want to understand what's going on in the world around them, get context about the news. You know, it's really about helping someone live a good life.

(Joel Beasley at 00:04:08) And what's the big business side? Like, what are the big areas of revenue generation? Is it the news and the financial data? Or?

(Li Zen at 00:04:15) Most of Yahoo's revenue still comes through advertising. Right? We're advertising-funded. We have various subscription options as well. You can subscribe to mail storage or, you know, finance. There's subscription tiers. Fantasy, there's a subscription there as well if you want to have some of the more advanced features and functionality. But by and large, you know, most of our revenue is advertising-based.

(Joel Beasley at 00:04:39) Oh, that's pretty interesting. So that's why you're constantly seeking to bring value through weather, through news, through stock information, and then that value is then monetized through advertising.

(Li Zen at 00:04:50) Yeah, exactly. And it's kind of funny. I think in some sense, you know, if you think about the way the web has evolved, some parts of the web have actually become more closed off, right, where it's gated behind a paywall or it's gated behind some subscription. I think Yahoo has still very much held true to that idea of an open web that can be funded and supported by advertising in a way that is both respectful of consumers. Right? We are very cautious about privacy. We want to make sure the advertising we're sending you is relevant, that it's actually stuff that makes sense for you in that context. But at the same time, you know, make the web still open and free for people to use in a meaningful way.

(Joel Beasley at 00:05:34) And then people using it — do you see any trends with age ranges?

(Li Zen at 00:05:40) Yeah. That's super interesting too. You know, I think right now, roughly half our users are millennial gen Z. And so we are getting this whole set of users who maybe didn't even know about Yahoo as a new thing. Yahoo was new when I was growing up, let's say. It didn't exist, and then all of a sudden existed. There is a whole set of users now who are using Yahoo where it's always been around, which is also kind of interesting to think about for sure.

(Joel Beasley at 00:06:14) Real quick, thank you to Fin for sponsoring this episode. Fin is the number one AI agent for customer service with a 65% average resolution rate. Fin is built to handle complex multi-step queries like returns, exchanges, and disputes, delivering high-quality personalized answers just like your best human agent. Fin is trusted by over 5,000 customer service leaders and top AI companies, including Anthropic and Synthesia. Fin works on any help desk. You don't have to overhaul your current system to get the best AI agent for customer service, priced at only 99 cents per resolution with a ninety-day money-back guarantee. If you're ready to transform your customer service experience, find out how at fin.ai/cto. And then how did you get involved in all of this? How long have you been there? How did you get there?

(Li Zen at 00:07:05) Yeah. I've been at Yahoo for just about a year. Actually just crossed my one-year mark a few weeks ago. And, yeah. Prior to Yahoo, you know, I've worked in a number of different roles, big companies, small companies. But what really attracted me to coming here was really just the opportunity to, A, work at such a storied brand, and, you know, B, to have a really huge impact across the organization. Yahoo happens to be working on a lot of stuff that I'm interested in in terms of platform engineering. We're undergoing this massive shift to cloud and then also AI. And those are two areas I've worked in my whole career. So it's great to be at the intersection of that.

(Joel Beasley at 00:07:46) All right. Let's talk about those. So you're doing a massive shift right now. Tell me about that.

(Li Zen at 00:07:52) Yeah. So, you know, up until the last few years, most of Yahoo has been operating on premise. We've always — we've been running our data centers for the last thirty years, basically, and we had data centers globally. We have data centers in the U.S., in Asia, in Europe, and all of our infrastructure was in those data centers. Over the last few years, we've been shifting more and more of our workloads into the cloud and building in a way that is more cloud-native, let's call it. And that's really unlocked a lot of development velocity, iteration velocity for us. Right? We're able to build on features much more quickly. You're not waiting for hardware to show up to go do that new thing. That's been very beneficial, and we're also able to take advantage of more managed services as a result too.

(Joel Beasley at 00:08:41) You've been at different companies, and now you're at Yahoo, and we can edit anything we want to edit. How does the culture feel? Does it feel like modern Silicon Valley-style company, or does it feel older?

(Li Zen at 00:08:54) You know, it's interesting. It's a combination, I would say, of those things. No company I've worked for has really been the same in any real way. Maybe one way I would talk about it is this: I worked at Amazon for a long time, and one of the things about working at Amazon is a lot of the tooling is all in-house. And so actually, Amazon in many ways doesn't use a lot of what you'd call modern tools. Obviously, everything they're building is relatively cutting-edge and at the forefront of technology, being the technology company it is. But as an engineer there, you don't get to use necessarily a lot of stuff that's out there. You're not using the latest and greatest technologies that are being built as SaaS elsewhere. At Yahoo, we actually do quite a bit of that, in contrast. So, you know, we're using the latest, greatest technologies across the spectrum. And we're not really limited necessarily either by feeling obligated to dogfood our own stuff the way that, you know, during my time at Amazon, I would say we would be. So I think that's actually interesting where that part, I would say, feels very modern. So I was at HelloFresh before I came to Yahoo. HelloFresh was, I would say, relatively bleeding-edge, let's call it, where most of the stack is using all things that were built in the last few years, and that's true here at Yahoo as well. The difference, I would say, when you talk about kind of does it feel modern or not is we have all the new stuff, but we also have all the old stuff because we're in this modernization journey. And so to give you an example in observability, you know, some parts of our system are using OTEL. Right? So it's OpenTelemetry, and we're using various SaaS providers for OpenTelemetry. All that is a very modern stack. It's going into Grafana dashboards. It looks a certain way that you would expect, quote-unquote, in a modern stack. At the same time, you know, we're still running the original metrics stack that was built over a decade ago, which is still really relevant. And, you know, if you think about it, it's actually pretty amazing. This metrics stack was built over a decade ago that was built to scale to, you know, Yahoo top-five internet property scale. It's still working. We're still maintaining it. We still feed tons of metrics through that. But, obviously, it has not kept up in terms of the feature pace as some of the more modern equivalents do. So it's missing some of the same features that you would expect in, say, Prometheus or whatever. But so we have both. Right? That's, I think, probably where it's modern, yet other parts don't feel so modern, let's say.

(Joel Beasley at 00:11:54) But you're enjoying it. This past year has been a fun year for you?

(Li Zen at 00:11:57) Oh, yeah. It's been super fun. I think, A, just the iconic brand. It's been fun to work at a company like that. But I think, moreover, both from a technology perspective — like I said, this transformation journey is super fun — and also, just the cultural journey is also very fun. And, you know, really, I think one of the things I've really been focused on this past year is helping my team really find its own way in being a product-oriented team and really embracing platform engineering. That's been a big focus of mine over the last year.

(Joel Beasley at 00:12:33) Help me understand the scale as far as when you're going in there and working on this big shift. How do you communicate your ideas to people?

(Li Zen at 00:12:42) Yeah. It's through a variety of different forums. Right? And I think communication, I would say, happens in a number of ways. There's obviously what you would expect when you think about communication. So email, Slack, you know, written communication. And there's also obviously various verbal communication channels. And so I think, you know, just talking about those briefly, obviously there's been a lot of emails or documents I've been sending out about how do we want to approach some of these problems? How are we approaching AI, or how are we approaching platform engineering? How are we doing those things? Verbally, there's been lots of various communications. For example, you know, during all-hands or, you know, we have a monthly open house — trying to tell people the story of, "Okay, where are we going as a platform engineering organization? How are we being customer-oriented? How are we thinking about the shift in terms of how we're doing AI?" et cetera. So there's all those forms. But actually, I think one of the things about communication is it's not just those kinds of communication, but it's also, let's call it, the mechanisms and the actions that you do that are also very communicative. Right? And so, you know, for example, one of the things I think that is a form of communication, let's call it, is, you know, we instituted a weekly operations meeting where we've been a lot more hardcore, let's call it, about reviewing postmortems and reviewing operational dashboards and metrics. And that, to me, is one way I was communicating to the team, "Hey, we have to be a lot more rigorous about how we're approaching our operations," without necessarily writing a memo about, "Hey, let's be hardcore about operations."

(Joel Beasley at 00:14:37) So this AI document about how you're approaching AI — tell me about that.

(Li Zen at 00:14:43) Yeah. I mean, one of the presentations I gave in conjunction with our COO a few months ago was really about talking to the organization about, "Hey, we need to really think about how we're working and really approach our work with a different lens." And I think we really break it down into three different buckets. You know, how we work, how we build, and how we serve, and how do we infuse AI into those three different areas? How we serve is really about, you know, how are we serving our customers, and are we using AI responsibly around making sure that it's a great experience for consumers? The how we build portion is really about developers. How are we actually helping our developers utilize AI to go build more effectively, efficiently, with higher quality, et cetera, et cetera. And then on the how we work side, it's really about how are we, just as general employees, all able to take advantage of these technologies and really go from AI being a tool to trying to find ways to use AI as a coworker. Right?

(Li Zen at 00:15:53) How do we actually really offload more of our work to AI versus kind of sitting in this back and forth loop? I think one of the things I really believe—sorry, this is a brief tangent, but I really do think that one of the key shifts we all have to make sure we're enabling is getting the interactivity aspect of AI down for humans. So we're not sitting in front of the prompt or whatever it is, or Cursor, or whatever tool you're using, kind of getting to a point where we actually only have to touch it with human hands once a day.

(Li Zen at 00:16:30) Right? Or once a week or something like that. Versus, I think, today, a lot of times, the feedback loops are on the order of minutes versus on the order of hours or days. And that's kind of what I'm trying to shift this more toward: Okay, we need humans in the loop, but how do we have more AI stuff happening in the background and that we only check in on a less frequent basis?

(Joel Beasley at 00:16:55) I had a conversation a couple weeks ago with a guy who's at a healthcare company. It has about 2,000 people or so in engineering. And he was sharing how they internally ran this contest where he gave parameters of the tools that could be used and he time-boxed it. And then he pulled in, I think, like 50 different leaders and cross-functional people in different teams. And he said, "Here's the toolkit. Use it to make your life easier." And then the people submitted how they were using it, and they took, like, the top three and then productized them and then rolled them out to their departments. And I thought that was a pretty interesting way of deploying. I've heard a lot of different ideas on how to do it, but functionally, this year, that's probably my favorite one.

(Li Zen at 00:17:43) Yeah. We've definitely been doing a lot of kind of—like, for example, in my team, we ran an innovation day a few months ago. We're doing another one this month. So kind of having those opportunities to let people experiment, I think, is super important, especially given how new a lot of this tooling is. I think sharing examples is actually one of the most important things teams can be doing, and I think that's what's also been great about—I'm super grateful for anyone on the internet who's published anything about what they're doing because I think seeing how you can apply it is super helpful and kind of sharing those ideas about, okay, this works, this doesn't. You know, here's a workflow I've constructed. You know, here's how I'm using this, here's how I'm approaching that, here's how I'm thinking about context, et cetera, et cetera.

(Li Zen at 00:18:20) And that's also been super helpful. I think one thing I would say that's hard is getting out of the mode of "what am I good at" versus "what can I offload?" And what I mean by that is I think there are just so many things we as humans and as workers are accustomed to and kind of like, "Well, that's up my alley, so I'm gonna go do it." And so I think it's very easy to default into just doing something versus pausing for a second and thinking, "Wait. Maybe I can just not do this."

(Li Zen at 00:19:14) A super silly example that I give to my own team a bunch is: there's a—it was about half a year ago. I went to some conference, and I wanted to—I missed a bunch of the talks. I couldn't go to every talk I wanted to go to. And so I was like, okay, let me just download a bunch of these talks, and I'll watch them on the plane, and that'll be great.

(Li Zen at 00:19:40) And I had favorited a bunch of them from the agenda list. And I ended up writing a script to go through, take the favorites, and run them through this downloader because—I guess maybe I shouldn't say "listen to your podcast," but they were on YouTube and, you know, there's no easy download button on YouTube. But I remember, like, "Oh, I'm an engineer. I know how to do that." And so it's like, "Oh, I'll brew install whatever the YouTube downloader—"

(Joel Beasley at 00:20:07) I know what you're talking about. I just did this, like, a couple weeks ago.

(Li Zen at 00:20:10) Yeah. And then, let me brew install this YouTube downloader tool. Then I'll just build the script. I'll loop through my favorite things from that page and then have it download them, and then I'll watch them on the plane. Perfect. It probably only took me, I don't know, twenty minutes, thirty minutes to put together. Right? It's kind of like, "Yeah. I know exactly what to do. I know what I'm doing. I know the things I need to do to make it happen," and it all worked. And I was happily watching the videos on the plane. But kind of after the fact, I realized maybe this would have taken, like, three or four minutes of me just prompting an LLM, like, "Hey. This is what I wanna do. Here's the page. Do all this stuff for me," and I would've saved myself twenty minutes, thirty minutes. I didn't do that because I'm just so accustomed to, "Yeah, I know exactly what to do." Right? So to get out of that mind space of "Yeah, I know exactly what to do"—it doesn't matter if I know exactly what to do. What matters is, can I offload this? And I think that's actually a hard mindset shift, I would say. Even however many times I've repeated this, I still have times where I'm, like, after the fact realizing, "Oh, yeah. Actually, I could've offloaded that."

(Joel Beasley at 00:21:24) Well, the reason why you're hesitant is because you have experience, and the experience will tell you that if you know how to do something a very specific way, that's usually the shortest path.

(Li Zen at 00:21:33) Right.

(Joel Beasley at 00:21:33) Because you could try to implement—like, take it back to you're gonna do a rewrite. You're not gonna learn a new framework doing a rewrite. You know what I'm saying? You got it. You don't need to be learning the new syntax and the new method calls and all that stuff. What you need to do right now is achieve business results. Right? And so if you know that there is a rabbit hole that it could turn into, like, you trying to figure out how to get an AI to do that could end up taking you two weeks.

(Li Zen at 00:21:59) That's true too.

(Joel Beasley at 00:22:00) Yeah. Yeah. So then you lean toward—this is the same thing that I wrestle with all the time too. But there are some points when you have major shifts in technology when you have to just take a look, and usually it's really loud. A lot of people are saying it. I just recently got into, like, Cursor and using the LLMs to, you know, figure out exactly how I wanna write projects and, like, how to use it. And I at first, I was, like, not that impressed. And then my best friend sent me a video of what he was doing. And I said, "Oh, I'm just using the tool wrong." And then I learned from him, and that just made all the difference in the world.

(Li Zen at 00:22:38) Yeah. That goes back to, I think, just the examples. Right? I think kind of being able to see how far some people are able to take this tooling and kind of what you can achieve, I think kind of forces us to really think about, okay, what can I really offload here?

(Joel Beasley at 00:22:54) Yeah. I wanna talk a little bit about Artifact. So you guys acquired—you guys acquired Artifact. They're an AI news aggregator, doubled your news app usage. Were you involved with that transaction at all?

(Li Zen at 00:23:07) No. I wasn't. That happened before I joined the company. So I wasn't involved in the transaction.

(Joel Beasley at 00:23:14) How does it come up in your day-to-day life now?

(Li Zen at 00:23:17) Yeah. I mean, you know, a lot of technology from that acquisition was incorporated into the recommendation systems we are building and are continuing to improve on for news. So that's where it would show up the most, I would say.

(Joel Beasley at 00:23:31) Nice. Your CEO said, "If we don't innovate, we die." What are you—like, do you feel pressure to innovate?

(Li Zen at 00:23:40) No pressure or anything. No. I mean, I wholeheartedly agree. I mean, I think that's the world we live in in terms of just how products work and what consumers expect. Right? I think if we are not constantly innovating and we're not able to continually meet new needs and find ways to make things better and easier for customers, we won't survive. And that is why, you know, like, if you go—if you look at the Yahoo homepage and you look at across all our properties, almost every pixel is new over the last couple years. Like, we really have been spending a lot of our effort on—you know, when I talk about modernization, it's not just, "Hey. Let's change the way the underlying stack works. And let's just redo that," all you know, to your point about refactoring. Like, we are looking for business results. It's not just, "Okay. Yeah. Refactor for the sake of refactoring." Right? We are trying to find ways to innovate faster, to let teams build faster, and so that, ultimately, they can build new capabilities, new features to continue to please our customers.

(Joel Beasley at 00:24:48) When you joined, what's something that surprised you? Because Yahoo is this storied brand. As you've said, you have all of the stuff in your mind about what it could be or what it should be, and then you join, you start working there. What's something that surprised you?

(Li Zen at 00:25:02) Maybe one kind of funny thing is—is not so much from a technology perspective, but from just an organizational perspective—is, you know, obviously the company that is Yahoo today came out of Verizon Media, Oath, right, from before. And depending on who you talk to, depending on where they came from, they actually kind of talk about it slightly differently, I guess. So a lot of the folks who were around when—because part of the company is AOL. Right? And so for the folks who were originally on the AOL part of the business, they often refer to it as "when we bought Yahoo," which is correct. Like, I guess the way, you know, historically, it worked was, I think, Verizon acquired AOL, and then that entity acquired Yahoo. And so but just kind of given the brand recognition and the way that, you know, we are obviously Yahoo now, it's always—it always gives me a second pause where it's like, "Oh, wait. Oh, yeah. You're right. When we acquired Yahoo," it kind of said this kind of, like, funny, you know, inverse relationship thing you don't quite expect.

(Joel Beasley at 00:26:16) So did just Yahoo spin out, or did AOL stay at Verizon Media, or did the whole thing?

(Li Zen at 00:26:21) It's one entity. Yeah. We're just combined as—yeah.

(Joel Beasley at 00:26:23) So you are AOL and Yahoo?

(Li Zen at 00:26:25) Yes.

(Joel Beasley at 00:26:26) Oh, that is so cool. Do you guys still have AIM?

(Li Zen at 00:26:29) We don't have AIM anymore. No.

(Joel Beasley at 00:26:31) Oh, no. Oh, man. That was the first bot I ever talked to—this bot called SmarterChild.

(Li Zen at 00:26:37) I remember that. Yes. SmarterChild.

(Joel Beasley at 00:26:39) Yes.

(Li Zen at 00:26:39) I remember me interacting with SmarterChild as well. Yeah. It was—yeah. Good times. Like, SmarterChild was a better version, I guess, of—I mean, I remember going all the way back to, like, you know, playing with ELIZA. Right? So, yeah, lots of bots that I've played with, but I do remember IMing with SmarterChild. Yes.

(Joel Beasley at 00:26:59) Yes. That was middle school days for me. So, uh, what's the coolest feature? I'm sure there's all sorts of stuff you can't tell us about, which is exciting and maybe off the podcast. But on the podcast, what's the coolest AI thing that you can tell us about?

(Li Zen at 00:27:17) It's hard to pick. I think I personally—I would say I would either pick the TLDR features in mail because I love having those summaries. I think it's super useful to just, you know, get a quick snapshot of, "Okay, here's what's going on in my email." It's either that or the AI features in Fantasy because I often spend—well, still spend—it's kind of, like, unfair to the AI that I do this, but I'll, like, you know, I'll still spend as much time researching, you know, what I wanna do with my lineups and stuff like that. And, really, I should just rely on the AI GM because it does just know better. But—

(Joel Beasley at 00:28:01) Have you seen that study with doctors that they did? Okay. So my brother and mom are physicians, so we always have these good technology conversations, and they hate when I text them certain things. So there was a study done where they took AI alone, AI with doctors, and doctors alone. And the group that did the worst—tell me which group did the worst. They took—and I'll give you some more background. They took known patient data with the known outcomes, like verified outcomes, and then they only gave it the first part. And it's, like, this controlled test. Which group did the worst, do you think?

(Li Zen at 00:28:39) I'm gonna guess doctors alone did the worst.

(Joel Beasley at 00:28:43) Nope. Doctors with AI did the worst. Okay. Interesting.

(Li Zen at 00:28:47) I thought it was gonna be that they did the worst than with the AI, they would have done the same, let's call it, or something like that.

(Joel Beasley at 00:28:52) That's what I thought too. Honestly, I thought the same as you. I was very surprised to find out. And so the underlying takeaway was that they don't trust the AIs. So the AI gave them the right answer.

(Li Zen at 00:29:08) That's really funny.

(Joel Beasley at 00:29:09) Yeah. But they didn't trust it and they didn't—and but doctors alone just coming to the conclusion was good. Doctors with the AI, not as good. And then the AIs were better than all three. That's the rank. The AI was the best, and then doctors, and then doctors with the AI was the worst.

(Li Zen at 00:29:24) That's really funny.

(Joel Beasley at 00:29:25) Have you had a chance to use AI for, like, any medical stuff, like reading your lab results or anything like that?

(Li Zen at 00:29:34) I can't say that I have. I have consulted for kind of, like, you know, I've been doing some, like, physical therapy type stuff just because I had, like, some knee stuff. And I've used it for that. Like, "Hey. Help me construct some physical therapy exercises to kind of keep on going and then kind of improving my knee stability." But beyond that, I don't think I've really used it for, like, "Yeah. Yeah. Go read this chart or something like that."

(Joel Beasley at 00:30:04) Oh, I have. I get blood tests, like, quarterly. And, uh, because my parents—they own, like, a medical clinic for health stuff, their doctors. And so they just have me, you know, do tests and things like that. And I put the blood test data into the AI and then had a—and then I called the doctor, and I was like, "Alright. Read me these results," and then I could check the doctor against—it was so interesting. It's a lot of fun for me because that's very empowering. I mean, a lot of the platforms are trying to shut it down as much as possible for liability reasons, which I totally get. But there are unfiltered AIs like Venice AI. Have you come across that?

(Li Zen at 00:30:47) I haven't. No.

(Joel Beasley at 00:30:47) Yeah. It's pretty cool. If you want, like, a private AI, Venice AI is where to go.

(Li Zen at 00:30:54) Interesting. No. I've definitely seen, you know, the various examples on, you know, social media where people have, you know, something that they could just never figure out, and then they spend a few sessions chatting back and forth with an LLM, and then it's like, "Oh, yep. It figured it out. Like, it was, you know, whatever." Just no one had ever connected all the dots.

(Joel Beasley at 00:31:15) Have you used it for parenting at all?

(Li Zen at 00:31:17) I have. I've used it for—I've used it for a lot of various other purposes. Physician, just not the medical charts. But, yeah, parenting, you know, even, like, one thing I like using LLMs for is really around kind of challenge circles. Right?

(Joel Beasley at 00:31:34) Oh, I don't know if that's our—

(Li Zen at 00:31:35) You know, it's the idea of like, hey, you know, I'm thinking this. Be contrarian to me. Like, challenge me on that idea. So be the devil's advocate kind of person that I can talk to. Let's go back and forth on this and really explore if this is gonna be good or not.

(Li Zen at 00:31:51) And I use that actually all the time from a work perspective. Right? It's like, okay, here's an idea I have around, you know, changing maybe the way a mechanism works or changing the way something might work. And I'll definitely consult an AI and have it act as a contrarian to try to poke holes in it. You know, hey, maybe this isn't gonna work out so well, and tell me what is gonna go wrong.

(Joel Beasley at 00:32:17) No, I like that. So when I'm prompting it, I'm doing a very similar thing. But in my mind, instead of challenge circles, I ask it for perspective iterations. I'm like, iterate through multiple perspectives and let me know how different people might see it. And that tends to work out pretty well for me.

(Li Zen at 00:32:33) Nice.

(Joel Beasley at 00:32:34) Yeah, that's so cool. So the coolest AI feature you could tell us about, what was that again?

(Li Zen at 00:32:41) Oh, I was saying, maybe I'll have to pick one. I'll go with the—go to chatyahoo.com. You should be able to interact with—so it's part of our search. But if you go to chatyahoo.com, you can interact with our chat interface.

(Joel Beasley at 00:32:55) What's the name?

(Li Zen at 00:32:57) Oh, it's beautiful.

(Joel Beasley at 00:32:58) Yeah. Does it have a name?

(Li Zen at 00:33:00) It does not have—it's not like—

(Joel Beasley at 00:33:03) You didn't personify it.

(Li Zen at 00:33:05) Cortana or something like that.

(Joel Beasley at 00:33:06) Yeah. Aww, you guys gotta have fun. Have you ever had—and see, I don't know Yahoo's history. Maybe I could ask the chat. But have they ever had any type of mascot?

(Li Zen at 00:33:19) You know, I don't know the answer to that.

(Joel Beasley at 00:33:22) Ever? Maybe we're gonna do this right now. Yeah, you should know this. Had any type of mascot?

(Li Zen at 00:33:29) I mean, obviously, the thing that Yahoo is most associated with is our yodel. And I think, you know, that's—

(Joel Beasley at 00:33:36) If that's in my head.

(Li Zen at 00:33:36) Anything, that's the mascot, I guess, is, you know—

(Joel Beasley at 00:33:40) Do you ever do that in your meetings? You just go like, "Yahoo!" Do you ever do that?

(Li Zen at 00:33:44) I am not someone you want to sing, so I can't do that. I'm still waiting, actually, for—we have a company store where you can buy company swag and cool stuff like that.

(Joel Beasley at 00:33:57) Okay.

(Li Zen at 00:33:58) And one of the items that was sold out last time I checked is there's a little button that you have on your desk, and it plays the yodel when you hit the button.

(Joel Beasley at 00:34:09) Okay.

(Li Zen at 00:34:09) So I'm waiting for that to be back in stock and, you know, then I can just slap the button and get the yodel to come out.

(Joel Beasley at 00:34:17) Okay. I have an answer here from Yahoo AI chat. "Yes, Yahoo has had a mascot in the past. The most notable one was the Jumping Y Guy, which was introduced in 1995, designed as a stylized yellow figure that resembled a person jumping for joy, forming the letter Y with its body."

(Li Zen at 00:34:37) Oh, okay. Interesting. That's good lore to know.

(Joel Beasley at 00:34:40) Jumping Y Guy. Yeah. There we go. I also have some more lore for you. The original Yahoo, when they put the founders together, they had gone to Guy Kawasaki to run, to be a part of that, essentially to lead it, and he turned it down. And he said it only cost him a few billion dollars. I just got to talk with him. It was perfect timing because I did an interview with him like two or three weeks ago, and he just randomly shared that Yahoo story, and I was like, we've got Lee coming on the show in like two or three weeks.

(Joel Beasley at 00:35:17) Yeah.

(Li Zen at 00:35:18) Yeah. Well, I have a kind of personal connection to Yahoo in some sense as well. I guess maybe like, you know, let's complete the circle kind of thing, let's say. Because, you know, obviously Jerry Yang—actually, by the way, he went to Stanford, and I went to Stanford. And, you know, actually got to take a lot of the same—I would say exactly the same path as him, but obviously had an opportunity to kind of be in the same place as he was on campus, albeit in different decades. But kind of cool to think about, you know, hey, like, this was started by Jerry and David, Stanford guys. And, you know, as a Stanford alum, getting to kind of tread in their footsteps. And still, you know, like I said, the brand is just amazing. Such an amazing company.

(Joel Beasley at 00:36:07) Well, I wanna wrap up with a couple different leadership questions if you're cool with that.

(Li Zen at 00:36:11) Sure. Yeah, of course.

(Joel Beasley at 00:36:12) Okay. Authenticity as a leader, how important is that?

(Li Zen at 00:36:16) That's probably one of the most important things in my book. I think I really care about bringing your authentic self, let's say, to work. I think it's really important to lead in a way that is true to your own values. So I think authenticity is super important to me. And I think that the two things I care most about are authenticity and transparency, like honesty and transparency.

(Joel Beasley at 00:36:42) Do you reflect that pretty well as a parent?

(Li Zen at 00:36:46) I try to. I try to. I think it's a little different with kids because I think they don't have the life experience necessarily in all cases to—that you can trust that they'll do the right thing with all the context, let's call it. Whereas at work, you know, in my mind, my goal is to give the best context I can to every employee so that they can autonomously go and kick butt. Right? So that they have all the information they need to do a great job and really empowering people to have the autonomy to do that. And to do that, that requires them to have that context. And the only way to do that, in my opinion, is to be as transparent as you can about, you know, what's going on.

(Joel Beasley at 00:37:35) I love that. Yeah. Because for me, I think we've got a lot of similarities because when I interact with people, my job is to knowledge transfer directly the most important detail so they can do their thing and then point me in the right direction. I found that that's one of the most effective and fastest ways to get the result that I'm looking for. And so what I have also found is sometimes people aren't familiar with that level of direct communication. And my little secret, I guess, is I'll just ask for permission to speak like directly and unfiltered and just knowledge dump. And I used to do that before and it would rub some people the wrong way. But once I realized, like, if I ask for permission, then it's okay. That's helped me a lot.

(Li Zen at 00:38:26) I like that. I don't know if I always ask for permission necessarily, I think, to just, yeah, give people context. So maybe I'll try that. Yeah. Because I'm not the smartest person, right, in the room, nor can I know everything or do everything. And so the only way I know how to make sure that as an organization we do the best we can is if we are all empowered with knowledge. Right? Because then, you know, every individual person who is the best in their field and knows the most about their field and their domain has the business context, the technology context, the organizational context to go do what's best for the company.

(Joel Beasley at 00:39:06) And this is something that you look for when hiring too onto your team. Is that correct?

(Li Zen at 00:39:10) 100%.

(Joel Beasley at 00:39:11) Yeah, for sure. How do you kind of fish that out in an interview? Authenticity and transparency.

(Li Zen at 00:39:18) It's hard. I think, you know, most of my interviewing relies on behavioral interviewing techniques. Right? It's kind of just, you know, I always talk through past stories and try to talk about, you know, specifically talk through stories where leaders will have had to do hard things and to kind of understand how they approach those tough situations. Right? And I like to think a lot of times, you know, kind of how someone approaches situations comes through in an interview. And I think you really can get a decent sense at least of, you know, what someone's character is gonna be like.

(Joel Beasley at 00:39:57) So you'll share this story of how a leader handled a difficult situation and then kind of see how they respond?

(Li Zen at 00:40:02) Oh, no. I'll ask them about kind of like, hey, tell me, like, you know, tell me about a situation in your career where you had to, you know, X, Y, and Z. Right? And kind of we talk through that.

(Joel Beasley at 00:40:13) Oh, wow. Okay. Cool. You get a lot of interesting stuff. I'm sure you've got a lot of good stories.

(Li Zen at 00:40:18) Yeah. I think it's a great—I love that kind of interview style just talking through, you know, people's war stories, so to speak. Right? Like, hey, let's talk about a time you had this huge deliverable that you said you were gonna ship and then turns out you couldn't make it. Like, what'd you do? And I think you—a, you hear really interesting stories. And I just love hearing—I love stories, actually. I think I like storytelling, and I like hearing other people's stories. And it's—I think it's a great way to understand how someone approaches problems and how they think about leadership.

(Joel Beasley at 00:40:51) Would you be open to sharing the hardest story of your professional career?

(Li Zen at 00:40:57) I mean, I'm definitely open to it. I'm trying to think like what I would consider the hardest. Maybe be more specific, I guess, and we can try there. That'll help maybe narrow down—

(Joel Beasley at 00:41:08) The things that you most didn't want to do, but you knew you had to do it.

(Li Zen at 00:41:14) Yeah. I mean, I think I could talk about that. It's probably more of like a personal story, actually.

(Joel Beasley at 00:41:21) That's okay.

(Li Zen at 00:41:22) It's yeah. So it's probably less to do with the organization, but kind of more about a personal choice where, you know, this goes back to my role at Pulumi, where I was the VP of engineering there. And I think the thing that was really hard was deciding to leave the company, actually, because I thought, you know, if I think about—try to pull myself back in that mindset—we had such—and Pulumi still is such a great product. I love the product. I love the team. Everything was great, actually. Like, there's kind of like, why would I leave? But I think one of the things I realized over the two years working there was that the way that I philosophically thought about certain things was not the same philosophical approach as the rest of the leadership team. And there's nothing wrong with that. It wasn't like they're wrong and I'm right or I'm wrong and they're right or something like that. It was just very fundamentally different approaches to how we thought about really, really key aspects of how you run a business and kind of how we think about how our organization should run, how you think about planning, all these different topics. I think we just had very fundamental differences in beliefs about. And, again, I don't think there's anything right or wrong about those philosophies. It's just mismatched. And that was really tough because then I realized, okay, well, you know, do I really wanna spend the effort to try to get that alignment, or am I kind of not the right person for this company if that's how the rest of the executive team thinks? And ultimately decided that, okay, this probably doesn't make a ton of sense and made the decision to leave.

(Joel Beasley at 00:43:43) And I totally understand what you're talking about too because if 80% of the company sees things one way and you're in the 20%, you're either gonna fight a long, near impossible battle with just very difficult—or you could just go find people who are closer to what you think as far as philosophy goes. Yeah. And a lot of people struggle with that. I noticed when I was doing—I did a lot of talks in public, like at conferences and stuff before, and we'd do Q&A portions after the talk. And a lot of people would ask me questions, and I'd find the underlying root thing is that they're in that type of situation. You know, they don't really wanna know how to lead their team better. I mean, they'll—that's the question, but then when you start digging and pulling the layer—well, why? Why? Why? Be more specific. They'll get down to that they just see it entirely different. So that's actually—there's a lot of people who don't leave, who do stay. Right? And they have very difficult time.

(Li Zen at 00:44:38) Yeah. But it's tough. Right? Because like I said, really, really believe in the product and, you know, love the team that I built, but—

(Joel Beasley at 00:44:46) But you also have to watch how you spend your time. Right? You want your life to be enjoyable. And if you find people who do have, let's just use the word compatible philosophies, or at least not identical, but like compatible, then you can take off and you can all have a real good time and you can build each other up and it can go. But if you're at philosophies that are at odds with each other—like, if you're pro debt and you're trying to work at Dave Ramsey's company, it's just—you're gonna hate your life.

(Li Zen at 00:45:15) Yeah. Exactly. Exactly.

(Joel Beasley at 00:45:16) It's not gonna be good. No, that was a really good story. Thank you so much for sharing that. Wow. Calls to action. I mean, most people, 90% of people here, they're going to yahoo.com. They're interacting with Yahoo already on a very regular basis. So I think that wraps up. Is there anything else that you wanna add there before we wrap it up?

(Li Zen at 00:45:40) I guess if I get to make a plug, you know, I would—maybe the plug I would make is all of these services in a lot of our websites, but actually our applications did. So if folks wanna try out—you know, if they haven't tried the Yahoo Mail app, or if they haven't used, you know, any of our mobile applications—download those and give them a try. Like, there's a really cool feature in the Yahoo Mail app around managing your inbox. And it's almost like a gamified thing to kind of get to that, you know, zero inbox situation. I think it's kind of fun. So like I said, if your listeners haven't given those apps a try in a while, definitely give them another look.

(Joel Beasley at 00:46:24) This is a good excuse to log in to your Yahoo account and check out the new mail interface. Oh, wait. Olivia said something. She said, "Love this plug. In addition, Lee, would you want to share a little about your sports broadcasting background? That's unusual for a CTO."

(Li Zen at 00:46:42) Sure. I can cover that briefly. Yeah. So at Stanford, I joined the club. Well, actually, this ties in with your singing thing too, by the way. So ever since my voice changed as a teenager, I've had so many people be like, oh, you should sing. And I actually did try singing in high school, by the way. And the choir teacher was like, no way. Like, just don't—she's kind of like, let's, you know, maybe do something else. So I was like, okay, guys, maybe not for me. When I got to Stanford, you know, a bunch of other people, like, oh, you should sing. And I was like, no, I don't think so. Like, oh, but there's so many a cappella groups, blah, blah, blah. But there's a club for radio. KZSU radio at Stanford. Actually, it's really cool that this university runs its own radio station. It's open to both students and the local community. And the original radio equipment was built by an electrical engineering student who later became the chief head of the radio station. So he built all the equipment originally, and, you know, is still using that equipment today. Anyways, so I said, okay. Yeah. Let me check out this sports broadcasting thing.

(Li Zen at 00:48:03) And went on one of their guest segment things where it was one of those every week yelling at each other sports segments where everyone has a different opinion about what's gonna happen next week or whatever. And everyone's, you know, it's one of those talking heads, just everyone kinda giving their opinion shows. And that segued into I eventually became the play-by-play and color announcer for Stanford women's basketball. So I covered women's basketball for two years, two full seasons.

(Li Zen at 00:48:33) And then in my final year as a senior, I covered baseball, mainly because by that point I didn't wanna travel anymore. And so I split the schedule with someone else where I said, okay, for the baseball Tuesday night home games, I'll do those because the baseball schedule is so long. So I'll do those, and then you all do the rest. But the women's basketball years, those two years were incredible because I got to travel with the team, visit tons of interesting places, you know, small university towns you may not ever go to, except for the fact that you're playing women's basketball.

(Joel Beasley at 00:49:11) That is the coolest thing I have ever heard. I'm gonna be the announcer so I get to travel with the women's sports team.

(Li Zen at 00:49:17) Yeah, it was super fun. Like, I mean, I don't think I would... Well, should I name these cities? Yeah.

(Li Zen at 00:49:25) I guess that's okay. Like, well, no. I mean, like, is there a reason to go to Corvallis, Oregon or Lubbock, Texas?

(Joel Beasley at 00:49:33) Hey, Lubbock. That's where my dad's from.

(Li Zen at 00:49:35) Right. Except for, like, okay. I mean, yeah, Lubbock. I think it's half the population is Texas Tech at least, right?

(Li Zen at 00:49:43) And so I don't know if I would ever get to go to Lubbock except that I—

(Joel Beasley at 00:49:47) There's not much there. Yeah.

(Li Zen at 00:49:49) Except that I got to go, you know, broadcast this game and got to explore Lubbock. That was cool. So it was, yeah, it was super fun to get to travel with the team, see all these different college towns, and learn a lot about basketball, honestly. That was—I love sports. Also another good reason to work at Yahoo.

(Li Zen at 00:50:08) And yeah, I really got to connect with the team and the coaches and the players, and that was super fun.

(Joel Beasley at 00:50:15) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email: [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.