A (3:22)

Wow, that's amazing. Now when I go to do a Google search, there is an AI prompt that comes up, or you can see it searching, and then it gives me this wonderful answer. Is that AI Frontiers or is that something different?

C (3:35)

So I assume what you're talking about is a product that we call AI overviews. Okay, Dr. Nall, that was the first product that my team and I built out, and now we'll talk about the next one, AI Mode.

A (3:49)

Okay.

C (3:50)

AI Overviews is a product that today is available to over 2 billion monthly active users since we launched it to the general public.

A (4:01)

And then what's AI Frontiers exactly? And how is it different than what we're used to?

C (4:07)

AI Frontiers is ultimately responsible for thinking about how we can build new capabilities in our product experiences to satisfy ways and people in ways they couldn't before. So, for example, I was saying, you know, AI Overviews, when we started that, that was when we really said, let's bring generative AI to search under the assumption that this is going to allow people to talk to us in much more natural, fluid ways. And since we've launched that, we've just seen that organically, people are asking us more complex, harder questions that we can now do a really good job at answering. AI Mode was us pushing the next frontier of that and saying, okay, that's great, now let's enable a surface that's powered by our best models and that is conversational. And the result of that has been that people are asking us again, even more complex question, more nuanced questions, but in a conversational engagement where they can back and forth with search engine. And it's really changing the way people interact with search. And not only that, it's actually changing people's satisfaction with search. We're seeing them actually be a lot happier given these new AI products that we're building. And I'm sure we'll talk about this some more. Dr. Null. We're just at the beginning of this journey.

A (5:27)

Right?

C (5:27)

Right.

A (5:28)

We're at the Frontiers. That's right, that's right, exactly. So let's talk about what I like to call the holy smokes moment. You know, our listeners, they like case examples, they like numbers, they like a description of a real person and what they experience. So talk to us about when a user or a tester actually experience what you're doing with AI Frontiers and what was that like? What was that jaw dropping moment where it changed everything for them?

C (6:04)

So I think one of the best parts of my job is actually to sit in on user research and listen to people, watch them use the product and talk about how they've been using it after they've had access to it for a while. And I'll remember a moment where we were sitting with a bunch of college students and they were talking about their experiences using our AI products in search. And it really stood out to me how one of them talked about the fact that they're about to go graduate and they're trying to figure out what their next career is. And they had so many questions in their minds of what can I do with the degree that I have? What are the options? What are the different variables to consider? And so hearing them go back and forth using our AI products to actually reason, that was when it really validated to me what we were hoping we would do for people, which is really make it easy to think through some of the hardest questions. And then, Dr. Hall, every day I take my kid to school, I'm at a birthday party, and the jaw dropping is actually around the fact that I have parents of my son's classmates who come up to me and say, that box on the top of Search AI mode. It has changed the way that I understand things, that I plan my family vacation. And so it's that delight that we are helping people in their everyday lives and get things done faster, easier, that really gives me the confidence that we are in the right direction, the right track in what we're building.

A (7:39)

Yeah, well, I mean, what a gratifying position you're in. I mean, to be able to see it and to talk to your sons, you know, friends, parents, and to see how it's making a difference like that, it's got to be gratifying. And also our listeners, they want to know how it works. So if we were to open up the hood and look at the brain of how this works, this next generation of AI intelligent complex surgeon, how does it work?

C (8:09)

This is what we call, and I love to talk about it, Dr. Nall, a query, fan out technique. It's pretty amazing actually, when you look under the hood and say, what's going on? When you come to Google, search now and ask anything that's on your mind, what's happening is that these generative models take your question, understand the facets of that question that are really important, and then break that into subtopics. And when they break it into subtopics. They then go and do the research that's necessary to come back and compose a thoughtful response for you. It's equal to thinking the way a person thinks logically. So for example, I was saying, hey, a parent came up to me and said this is much easier for travel planning. Imagine now that you can express to search. I'm looking for a hotel. I need it to be kid friendly. But also my husband and I are exhausted so please make sure it's not too large and loud. And by the way, this is the budget considerations. That's what's all on our minds. So now with this fan out technique, the model is capable of taking all of those nuances and considerations and going and running multiple parallel searches, coming back with the best sets of options and considerations for you and then connecting you in turn to the best of the web. Because it's gone out and taken a look at all those different facets and brought that forward in the response. So it's pretty amazing actually.

A (9:32)

Yeah, no, that is amazing. Now this is the question that I have and it's actually kind of an area of research for me. Are you finding that mo that there are people that are using the Google search function to then have a conversation related to some companionship? Are you saying that at all?

C (9:49)

You know, I think that at the heart of it, people look to search to really ask a lot of their core information needs questions that is ultimately how people think about search. It's their trusted source of information. So all our focus innovation leans more there, I guess.

A (10:09)

Yeah, yeah, go ahead. Okay. No, no, perfect. That was good. I was always curious about that. So it is for the search to kind of help with my, my everyday, what I'm doing in life, even questions that I might have about the universe and why it exists and why things happen. Exactly. Etc.

C (10:26)

That's exactly right. That's exactly right. That is the heart of what, what we are and we're making that so much easier to anything that you're curious about to get a response to.

A (10:35)

That is great. Now I like to also talk about the real world magic and you gave us some examples of talking to college students that were about to graduate, talking to your network within your son's school. I would love to hear an example where you as director of product management you saw some, some next iteration, some development and it gave you chills like I can't believe that I did this and my team did this and this is where we are. Take us there. If there was a victory dance you could do that too. But take us to that moment in the room or wherever you were. It was like, wow, I cannot believe this.

C (11:17)

You know, it's so funny. Dr. Al, the rate of innovation of this technology and therefore the possibilities is moving so quickly that without a doubt I have that sentiment on a regular basis right now. And that is why I often say I've got the best job in the world. Because that feeling of, oh my gosh, we can do that. And then dreaming of the next iteration is happening so quickly and continuously and we are pushing innovation as fast as we can. We announced our image generation capability that just launched. Well, Dr. Nall, my sons are both October babies and I use AI mode in my day to day life to actually look at different themes. I browse some visual intents. Then I asked the model to generate an image of what would be an ideal birthday cake for his personality. And then I use that to actually then iterate and build on the thing. I mean, it was incredible. And so that's just one example where I'm saying, look at all this technology. But then we also launched Agentic booking and so you can now use it to say, go find me a dinner reservation for two for my husband on Friday night so we can go for dinner. It's amazing. And this is just a handful of things that we have and will continue to bring to the forefront.

A (12:36)

So no, that's really exciting. So you take the picture of the cake to the bakery and they're like, are you kidding me? I have to create this.

C (12:44)

Can I just say they did a great job. Both, both the image generation did an excellent job, as did the baker. And ultimately the most important, my son was very happy.

A (12:54)

Oh, that's amazing. Because I've been, you know, on social media and you see the, this is what it was supposed to look like and this is the actual cake. And some of them are really hilarious. So I'm glad it turned out for his birthday. Now let's go to a very important topic that I have to ask every guest and that's around ethics. Talk to us about the ethical guardrails, your ethical thought process as you are developing these products. Literally by the second.

C (13:25)

Yeah. So while I talk a lot about speed of innovation and how thrilled and excited I am about that, Dr. Nal Trust in giving accurate information. It's at the core of what matters to search. It is what we expect to meet and our users expect of us. And so with that, all of our AI products are anchored in our core search quality systems. In 25 years of US fine tuning it to give you the most accurate factual information that we can. And then on top of that, you've probably seen Dr. Null, that AI overviews and AI mode. Our responses are all anchored and grounded in the web, both to ensure that we're giving you the best, most accurate information, but also to enable people to dig deeper and dive in for themselves as well. We hold the same policies for our AI products as we do for a variety of our search features out there and the same bar of trust and quality.

A (14:30)

Amazing. That is great. I always like to ask that question and I love it when people say they started with it from the beginning. It wasn't an afterthought. It was not an afterthought.

C (14:39)

No. It has to be a core part of it. That's right. That's exactly right.

A (14:42)

Absolutely. So let's talk about the big future. I used to say, okay, what does the future look like five years from now? But my gosh, the future in AI could be a minute from now. So using whatever time frame you want to talk to us about the big future with Google search and AI frontiers, what can we expect?

C (15:07)

So I love this question, I love to think about this, I love to talk about this. You know, I'll hit some themes that you will see us push on and I've already seen us talk about that, give you a glimpse into of the areas we're really going to innovate on. The three of them are agentic, multimodal and personalization. And I think these three are really going to change the way people think about seeking information beyond what you're the innovation you're already seeing on agentic. I talked about the fact that we enable dining reservations where the model can do it on your behalf. That is the start of the fact that these models can take actions on your behalf when you request them to do so. And you can imagine how that can open up a whole range of things that we can do on your behalf to save you time. The second is multimodal. These models are natively multimodal. And I think there's so much opportunity to push the innovation both on the kinds of questions you ask us and how you share your question, whether it's through camera or video or audio, but also how we answer you, how we can bring to life new formats that are composition of audio and visual in order to answer you in whole new ways. And the third is personalization. And this is an incredible space of innovation. I think we'll see that will continue to happen which is, as you use Google Search can understand what matters to you more and more, what your preferences are, what your interests are, what things matter to you when you make a decision. And ultimately we can then help get you to the right answer much faster.

A (17:02)

No, I love those three areas, particularly personalization. That's the one that's, that's most important. I like to know that it remembers my searches. It knows what I like. It knows that I don't like green peas and cooked carrots. You know, like, I don't even suggest that I have an egg allergy. I like that, that it remembers that and I don't know, repeat yourself.

C (17:23)

That's right, exactly. That's right, exactly.

A (17:27)

So call to action for those individuals who want to really, really experiment. I mean, everybody does Google searching. We don't have to worry about that. But this next frontier, this next level with Google Search, what can they do today to really, really get a sense for what it can do?

C (17:47)

I mean, besides the, the obvious, which is like, dig into AI mode, try it out, ask your questions. I would say there's two new capabilities that tie back to one of the themes I talked about, multimodal, that just came out that I would encourage people to try. The first is search Live. And it's this idea that you can ask a question about anything that you're seeing in your world and have a conversation with AI mode. And the second that I was saying, we just recently launched as well, Dr. Nall, is visual aim, where we're trying to push the boundaries of how you do visual inspirational search, how you shop in a conversational interface. Both of these are new capabilities that we've announced and something I would encourage people to go try.

A (18:39)

Suvi. That's amazing. That's absolutely amazing. So tell us a little bit more because our listeners love actual examples. So tell us a little bit more about SearchLive and some examples and then visual AI mode in some examples.

C (18:54)

So Search Live, Dr. Null is a technology that allows us to have you turn on your camera and ask any questions about what you're seeing. So in our household, yeah, we use it all the time to debug appliances. Imagine you got a new television, right? We see people use it this way. You can now open up SearchLive and say, Listen, I got this television, I've got all these chords. Where do they go? And you can have a conversation with AI mode where it will help guide you exactly which chord goes where. So you can imagine how much effort that removes from people being able to do that. And Just talk back and forth with AI mode to get the task done.

A (19:37)

Oh my gosh, Sufi, that's amazing because there are sometimes I buy products and I get the manual and I get so frustrated. And so now I can use Search Live and it will tell me exactly what I need to do and how I need to. Where I need to put this cord and that cord. Oh my gosh, that's amazing. Bad for thumbtack because that's normally where I go to, but good for me. That is great. Okay. Gray some. Okay, so we always. And I love this question. It's called From One genius to Another. So it's from my previous guest. And Sufi, your question is when future generations see your work, what question do you hope they'll still be asking?

C (20:25)

I love this question. You know, I am so excited by. I think the rate of innovation is going to be exponential for future generations. And so I hope they ask the question that I frequently ask myself when I come to work, which is what's the next paradigm shift? How does that enable me to meet my mission in entirely new ways? Really Constantly asking themselves what next? What are the possibilities and bringing that to life?

A (21:00)

I love that. I love that. And it's so innovative. Like what's next? How can we stretch our mind and our thinking? That's amazing. Okay, great. So bonus rapid fire. I'm going to ask you a question real quickly and then you'll respond whatever first comes to mind. So most, most overrated tech trend, the.

C (21:22)

Use of the word AI you all the time, everywhere, for everything. In many instances, I think it is overused.

A (21:32)

Okay, great. What is the most under hyped or underused or under thought about a breakthrough?

C (21:40)

I think that multimodal. The multimodal capabilities models. We are just starting to really think about how to apply it and bring it to life in everyday life.

A (21:50)

Okay, awesome. What's a book we should all read?

C (21:54)

Innovators Dilemma. Because in this world where things are evolving so quickly, I think that is a great read for everybody to think about how to keep innovating.

A (22:06)

Got it.

C (22:06)

Love it.

A (22:07)

And okay, wow us. What is the biggest, boldest prediction?

C (22:15)

I think this references something I said earlier, which is couple folds. The rate of innovation I think is going to be exponential. I mean, you said it earlier, we can't even think five years from now because we're looking around the corner and already thinking what next? And things are changing so quickly. Right. And so as a result, this actually is what gets me really excited when we think about the next generation. I think you're going to see so many people become entrepreneurs that historically have not been because these capabilities, the technology, the products are going to let people express their creativity in completely new ways, be able to engage and really lower the barrier to engaging with these technology advancements. So we're going to see a whole range of entrepreneurs that we historically wouldn't have seen.

A (23:11)

I love that. And I know that you said that the word AI is overused, but one thing that I do love about AI is that it does level the playing field. Right. It gives people opportunities that did not have innovation that they might have can now, you know, come to the forefront. More accessibility. And that is what's so exciting to me. You can have that small child that's in some village somewhere that can be just as competitive as somebody with, you know, more resources, et cetera. So that's what excites me about AI.

C (23:48)

Exactly. That's right. 100% agree. We can have a lot of builders.

A (23:52)

Yes, a lot of builders. I love that. So, Suvi, I've enjoyed this conversation. So if we want to stay on top of what you and your team are doing and all the innovation that you have, how can we do that? How can we get in touch with you? What's the best way to follow you and your work?

C (24:12)

I would say, listen, we are innovating and moving so quickly. Dr. Nall that the best way is to keep abreast of all the changes we're talking about in search and also go to google.com AI where you can access our latest and greatest at any time.

A (24:28)

Okay, I'm writing that down. Y' all heard it here too, google.com AI so I can stay abreast as well. Well, Supi, I so enjoyed having you as a guest. Your energy is contagious. It's exciting. You know, everything that you're working on. Thank you for being at the lead of, you know, all of this AI innovation, particularly with AI frontiers, with Google search and everything that you're doing. So thank you so much. Can't wait to stay abreast of the new developments by first going to google.com AI and then continue my search because I use it for everything. But I, by the way.

C (25:06)

Right, that's right. And I'm so happy to hear that. And thank you.

A (25:11)

Absolutely.

C (25:12)

This was such a pleasure.

A (25:13)

It was. And thank you for being here. And everyone, until next time, lead with AI.

B (25:20)

Thanks for tuning in to Lead with AI. I'll see you next time. As we continue exploring the cutting edge innovations shaping AI across across the public and private sectors. Until then, keep leading with AI.