Big Technology Podcast - Google's Search VP on Generative AI in Search, User Notes on Search Results, and SEOs — With Cathy Edwards

Episode Date: November 15, 2023

Cathy Edwards is the VP of Search at Google. Edwards joins Big Technology Podcast to break some news about Google's two newest search features, including a more personalized experience called Follow and a labs product called Notes that allows users to leave feedback on search results. We also cover the state of Google's Search Generative Experience and how the company feels about SEOs after the latest controversy. Tune in for a deep, fun conversation about the cutting edge of Google's most important product, with insights on its latest features. --- From the interview, here's quick note of clarification: Google's new Notes experiment does not currently impact rankings, and the company doesn't have imminent plans to change that. --- You can subscribe to Big Technology Premium for 25% off at https://bit.ly/bigtechnology Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice. For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/ Questions? Feedback? Write to: [e-mail]

Google Search VP Kathy Edwards joins us to break some news about some rather interesting changes to Google Search and to discuss generative AI's role in their work. And what's with those SEOs? All that and more coming up right after this. LinkedIn Presents. Welcome to Big Technology Podcast, a show for cool-headed, nuanced conversation of the tech world and beyond. Today we're joined by Kathy Edwards, a VP of Search. at Google. We've talked before. It's great to have you on the show so we can actually continue this conversation to break some of your news and also follow up on some of the conversations that we've had before. Kathy, welcome to the show. Thanks, Alex. It's great to be here.

Great to have you. The last time we spoke, you and I were having like a little bit of a disagreement about generative search experiences. I was saying that I had some worry about as a publisher. Would traffic ever come to publishers anymore if everybody's just going to get the answer is at the top of search. And you had mentioned that this would spark people's interest in going to visit a number of different links. And so it's really the main question about this product is, is it going to keep people within Google or still make room for the open web? You're a number of months into the labs experiment right now where people search. They get a conversational answer at the top through the search generative experience. I have that

turned on in my labs. And you have the data. So what are you seeing right now? and where my fear is overblown. Yeah, look, I think your fears are totally reasonable, and we definitely have designed SGE specifically as a jumping off point for the open web. We really want to make sure that as we're bringing generative AI into search, that we're really thinking about the health of the open web. And we just, as we talked about last time, we just really genuinely believe that people want to hear information from other people,

from other sources, not just from Google or from a chatbot. Do you know what I mean? So we're really excited to say that we're seeing that not only are we showing more links on the page, but that we're seeing like a new diversity of clicks coming off the page and that that's really encouraging early data. It's still really early. I just want to repeat that. We're still early on in this whole experiment, but I'm really encouraged by the data that we're seeing.

Interesting. So when you say a new diversity of clicks, that could be taken in so many different ways. I guess what people really are curious about is, does that mean people are clicking more out to the open web or less? Well, it's sort of hard to say at this point. The data that we have is because it's an opted in experience, you can't just like A, B, test it like we would if we were just rolling it out. But we are seeing more links and a range of different clicks going out. So we're pretty encouraged by the way that I think generative AI can sort of help people explore the diversity of content on the web. Okay. I'm going to try to read behind the lines here and you can tell me if I'm being ridiculous or if this is not something that you want to get into. To hear that there's a range of different links and a diversity of clicks to me,

It leads me to believe that people are actually clicking out less to the open web. Otherwise, it would be easy to say we're seeing more clicks. How am I, am I misinterpreting that or what am I getting wrong here? I can understand that you're like, hey, Google, you have all the data. Like, of course, why wouldn't you just be able to give me a simple yes or no question? It's actually like it's driving our statisticians mad right now. Because it's an opt-in experience, we actually don't really know. We haven't, you know, run an A-B test at scale yet.

So it's actually something we genuinely, like, don't fully know the answer to, but we know that we do still see people clicking. And we're pretty excited about the way that users are finding even more long-tail content as part of this experience. And long-tail content means stuff that might have been on the second or third or fourth page of search. And actually, when you're typing in a query and search generative experience is answering, it's going to look much deeper and say, hey, this might be relevant in search. surface it to you. Am I getting that right? Yeah, that's right. Well, that's great news because big technology loves to be on the second, third, and fourth page. So, very happy about that. One of the reasons why you're releasing this in labs is obviously to kind of see how things are going and see whether people are using it, adopting it, and whether it's something that makes sense to roll out to the

broader product. So how are you evaluating the performance of these generative search answers within Google. And, you know, I know the data is early, but you must be evaluating on some metrics. How does it look? And do you think this is going to be something that we're going to end up seeing rolled out to the broader population? Yeah, we're really excited by what we're seeing. We're seeing, you know, we track things on a few different fronts. We look at our user research where we actually survey users, as well as obviously our metrics that we track from a business perspective. And we're seeing that users are really excited about this experience, like sentiment is higher. We're also seeing users do more complex queries, which is great.

It's really, to me, one of the strongest pieces of evidence that this is helping because it's unlocking these things that users maybe didn't feel like they could ask before. And so we're pretty jazzed about that. Now, having said that, it's early. And we want to be really careful here. You know, Sunda has said repeatedly bold and responsible is the way forward. And so we're really taking our time to make sure we feel like we've nailed it before we graduate. And there's definitely, I would imagine a bit more time because Bing, which initially seemed like a bigger threat, doesn't seem to be making much headway at all in its battle against Google. I'm curious.

I mean, I open the floor for you to comment on that if you want. want. But let me just ask another question in case you don't want to touch that. What does advertising look like when it comes to these generative experiences? Because, you know, you write the question, it pops down a big page of text. And I don't know if I've seen any ads there. Is the monetization side different from the typical Google search? Is that developed yet? And is that going to hold back this thing from prime time? If you decide from a product end, we like this, let's roll it out. Yeah. So first, as you might expect. We're really focused on our users and the health of the web ecosystem. And that's what we're really focused on as we're building this experience. To the ads point,

you know, I'm sure you've heard we keep ads in organic very separate at Google. And so you'd have to talk to the ads team if you want to know how ads are doing. So that's not even playing in at all to your experiment right now with this product? We're really focused on building a really great product for users and making sure that the web, system is really healthy. Let me put in my one feature request. Sure. I think it's just slow. Like I, and I understand why, because it takes a much larger level of compute resources to do it. But when I Google something and I have this generative search on, I'm already scrolling down to the results by the time this thing decides what it wants to tell me. Look, I totally hear you on that. It's something we're working really hard on. And we've actually

made some pretty amazing progress, even in the past six months, in terms of speeding it up, but it's still clearly not fast enough. And that, to be frank, is one of the reasons why we don't put it in in everyone's hands, and it's still an opt-in thing at this stage. Now, you know, if I have confidence in anything, it's Google's ability to continue to improve its compute efficiency. And so I'm pretty confident. We're going to see it continue to come down. So substack has a toggle for me, that. I can decide whether I want generative models to train on the data that I produce data, basically my articles on big technology.

And I don't get any money for it, but I have it switched on because I kind of actually thinking about your product. Like if I'm on that third page and I have it switched on and others have it switched off, then there's a chance that I actually end up getting way more readership. That being said, like it is helping to power. you know, this training is helping to power billion-dollar products across the board. You know, maybe it's with Google, maybe it's with Open AI. I don't even know.

There's no transparency there. What do you think I should do? Do you think I should keep that on? Do you think I should, like, call substack and be like, pay me for this? Or should we, should us as publishers be speaking more directly with companies like yours and finding out the right value exchange for the content that is trained on? I mean, recency, right? We do news.

If people are asking what's happening today, that's actually quite important for a model. So I'm curious what your perspective is on that. So firstly, I can't tell you what you should do, but I can tell you. What would you do if you were me? Let's, you know, speculate, have some fun. I think it's really critically important that publishers like yourself have control. And they have the choices. And so I'm really pleased that we announced this Googlebot extended control for

robots.comtext, for those of you who don't know, this is a way at which a site owner can instruct the various bots that are out crawling the web, whether or not those bots can use your content that's on that website or not. And so search has had generative models in it for years and years, even prior to SGE. So that's still controlled by GoogleBot, which has been our bot that's existed for 25 years plus. But now with GoogleBot extended,

you can choose to opt out of BOD and our vertex APIs, which are Cloud, Google Cloud, Generative AI APIs. And I'm really encouraged to see that other players in the ecosystem are being responsible and also publishing their control. So the publishers have the ability to choose what they want to do and what's right for their business. Exactly. But it's tough.

It's tough because if you say no, okay, listen, if you're doing, no one's going to opt out in the right mind will opt out of search because, I mean, I'm sure some do, but I won't because people will see the link and they'll come to the product, right? And I guess, yeah, search is a billion dollar, multi-billion dollar, multi-trillion dollar business, right? When it comes to these, allowing these models to train on publisher material,

we are basically building the training data that is being used to create these billion-dollar companies to the point, where in Drescent Horowitz, recently in a submission to the copyright agency in the U.S., was like, if we were forced to pay for the copyright for the material that we trained on, are billions of dollars of investment in AI companies will be worth a lot more. Do we need to get to a point where there's a fair value exchange here between the publishers

and the tech companies that are using this to build very lucrative business? It has to be a little bit more than like opt in, opt out, don't you think? So look, I think we're really early. You're going to hear me say that a lot. I do genuinely think we're really early. And I think it's really, I mean, I will say you've said, you know, multi-billion dollar businesses. I don't think there's a multi-billion-dollar generative AI business in terms of revenue yet. I mean, valuations are one thing.

Right. It will get there. I mean, Open AI itself is 1.3 billion run rate this year. Sure. I think, and you'd have to ask them about, like, what they think in terms of a value exchange. But in terms, for us, we definitely are focused on how do we make sure that the web is healthy as we go through all of this? Obviously, like, the web changed humanity for the better. you know, really important for the world, I think. And so we're really focused on being good stewards of

that. I think having publisher control, so you can choose to be in or out is the right thing. And then I think over time, we'll figure out if there are other business models needed. Okay. It'll be interesting to watch it play out. And yeah, I know, no, I think generative, there's definitely some use cases that I'm using generative search for. Honestly, it's become like a crucial. And I know Google was doing this beforehand, but thesaurus for me and like almost like a mini Wikipedia in some cases. But yeah, it'll be very interesting to watch it evolve and I'm very curious about the business situation as it plays out. Let's talk now about the news that you came here to share. I mean, this is cool. We're going to be breaking some news as this podcast goes out. So apologies

we gave the competition 12 minute heads up here. But look, you have these two new new features coming out, one that's rolling out more broadly, one that's very interesting coming out in labs. The first is called follow, and I'm just going to quote right off the blog post that you've sent over before. So it says, we're introducing the ability to follow exactly what you're interested in from your favorite sports team to something more specific like vegan cooking. With follow, you'll see more content related to what you care about across search with less effort. So does this mean I'm going to be able to like follow newspapers, sports teams, the latest in like, you know, stir fry recipes, which is basically the only thing I can cook. I mean, what is it going to look like in practice?

Yeah, so the concept is, you know, what's so amazing about search is that you can type in any query, right? And no matter how weird or specific long tail it is, you can type in any query and get some results. And so the concept is that you can come in and follow those search queries. And so whether that search query is your favorite sports team, or whether that query is something really long tail. Like maybe it's not even stir fry recipes. It's like stir fry recipes containing broccoli that are gluten-free, right? Like you can follow something that specific.

And then in Discover the homepage of the Google app, and on that search results page, if you do the query again, you'll see content that's new for you, including perspectives from across the web. So it's a really great way of sort of staying on top of that interest that you have. And it's fascinating. And so when is that going to roll out?

So we're starting to roll that out. As of the day that this podcast goes live, we're going to take it a little slow, but it'll definitely be out fully to every single person by December. Very interesting. So what I'm sensing is like there's stuff that shows up on the web that you might want to know about, but the only way to find out about it is to just frequently

Google, try to find it. Right. And so this is just a way to be like, all right, you want to know what, I mean, it's amazing, how specific you're getting also. You want to know what's going on in these areas you talk about. We're actually going to push it to you. Exactly. And I think that's, you know, this is just, Google understands information so amazingly.

How can we help you stay on top of that without having to, you know, require you to come back and do a query all the time? Can you talk a little bit about, I mean, this is, it's interesting. interesting because, of course, it's a helpful feature. It also feels like something that's so intuitive and it's amazing that we're here in year 20 plus of Google and we're just starting to see this. So what was the compelling event to create something like this inside the company? It's a good question. I think we just really have been very focused on understanding what was going on with users. And we saw, you know, we did this thing maybe last year. I can't actually even

remember the timing that we in that where we put these widgets at the top of discover that are connected to sort of different things you might want to stay on top of so there's a weather widget there's a stock widget there's a cricket widget which like the cricket world cup is going on right now which i'm very interested in you are probably not but believe me i'm very interested in it i don't have the patience for cricket it's a very good game you should you should give another go try a t20 game but Anyway, what we saw is that these people love these widgets, and they just keep coming back and like, what's going on? Like, I now wake up every morning and I'm like, what's going on with the cricket?

And I check these widgets. And we really had this insight that sometimes the query, it's amazing that you can come to Google and express anything you want, right? But sometimes to just continue to express over and over is a little tiresome, right? So this idea, this really proved to us that there was this idea that users just want to stay on top of their information without having to do the queries. And we really wanted to extend that. So it wasn't just like the few widgets that we came up with, but that users could really do anything and express their interests, how they already are expressing their information needs to Google today, which is the search query. Absolutely.

So I want to go back to our AI or generative AI conversation. And, you know, for a long time, we've been hearing Google talk about how this company is going to become not just a place where you search, but something that is assistive, right? Something that kind of knows what you're interested in or you help shape what it knows that you're interested in. And it will push that to you. And I'm hearing this. And it's like, interesting because it's like, of course, this is going to happen in the main search area. But it also feels like this is something that you might want to tune like your Google assistant for. We don't really hear a lot about the Google assistant anymore.

So is this kind of something that as we end up, I mean, forming more personalized relationships with Google seems like another way, another way that that happens is through an assistive agent. So is this sort of the groundwork for that? Can you see this playing in there? Tell us a little bit about your thoughts on that front. Yeah. I mean, so firstly I'll say the assistant just announced some very exciting news just recently. So you should go check that out and definitely more to come here. but um wait wait what's the updates on the assistant uh so we announced that we're bringing

uh barred into the assistant uh which will be launching soon as it's uh pretty exciting um i don't know a lot more of the details you'll have to talk to them about that but there's a bunch of very exciting uh stuff happening on that front i think for us we're really focused on like what is the core mission of google search it is to organize the world's information and make it you know universally accessible, right? And so that focus is where we are really like, how can we just improve that over time? And I think the question's about, you know, an assistant you ideally want to do more for you. You also wanted to read your email and maybe book a restaurant for you and those sorts of things. We're really focused around this concept of just how do you get your information,

how do you answer the questions that you have, and how do you stay on top of your interests? Right. And there's also some news with the perspectives side of thing, if I'm not mistaken, that when you're searching on Google mobile right now, there's a perspective tab that will include things like tweets and blog posts and stuff like that. So you're going to roll that out more broadly also. Yeah, we're bringing it to desktop. This is something that we're really super excited about. We really have heard loud and clear that the users really want the perspectives of other people when they're searching for information. And it's been doing. so great on mobile. So we're really excited that it's that it's finally coming to desktop too. Yeah, I personally enjoy it. For me, it's also a way to sort of stay on top of what's happening on Twitter or X without like really having to log in very much. Like when I take my mandated time away from the app, when I feel my mental health going down the toilet, I still want to check on what's what's going on in the world. And this is actually like a pretty cool module.

So, so on that note, okay, there's there was a moment of, where you're, where we had social media, right, where, like, we would follow our interests and people would push our interests to us through feeds. Now, that still exists just kind of in weakened forums on places like Facebook, but also X and threads in blue sky or true social, whatever you use. And, you know, it's also like I'm seeing Google come in to talk a little bit more about what's the latest that happened. And it feels like, you know, intentional or not, this is an opportunity to take advantage of that vacuum that's being left by social media going to its next evolution. Is that something you think about? And is that an opportunity for you? I'm not saying build Google

Plus again, but I'm wondering, you know, how that factors. Yeah, I mean, we're definitely, this is not a social network we're building. We're really focused on the idea of people following their interests and staying on top of their information needs. It's not about. Which is, but that's exactly like what you would do on Twitter. I'm not saying you're going to, you're building a new Twitter, but it does seem like the decline there has left an opening for you. Well, I don't know. I don't really see it that way. I think we're really focused on, you know, and I think honestly, social media is doing pretty great more generally. I mean, I think like TikTok's going fantastically, Instagram's going fantastically. So I'm not

particularly worried about the health of social media overall as a segment. Obviously, that's something we rely on for our perspectives filter. So, we're really invested in making sure that all works really well. I think that we're just really focused on how do we help people with their information needs? And it just seemed this really great opportunity to sort of allow people to stay on top of that more proactively. I just think this data in a big technology's newsletter, it's past week. But it's actually worth highlighting because you're right, Instagram is doing incredibly. It's engaged hours per day, 1.1, billion.

billion. And its growth year over year is up 29%. So that's wild. The reason why I say you have and this is per sensor tower, which is an app at analytics company. The reason why I say there's an opportunity to fill a gap here is because Twitter's engaged hours per day is down 8%. So, but anyway, I understand that it's not like you don't want to be in that category and it's probably not something that you're focusing all too much. Just looking at it from the outside, though, it does seem like an opportunity. All right. Let's go to a quick break. Kathy Edwards is here. search rep for at Google. You might know Google.

It's the place where you search. Everyone says search is boring and static. I disagree every time Kathy and I talk. We have so many interesting things to discuss. Speaking of which, on the other side of this break, we're going to talk about notes. It's a new product and it makes Google a little bit more, I don't know, interactive or community focused.

Anyway, I'm going to just leave that tease and hope you stay with us after the ad so we can talk more about notes. Hey, everyone. Let me tell you about the Hustle Daily Show. a podcast filled with business, tech news, and original stories to keep you in the loop on what's trending. More than 2 million professionals read The Hustle's daily email for its irreverent and informative takes on business and tech news. Now, they have a daily podcast called The Hustle Daily Show, where their team of writers break down the biggest business headlines in 15 minutes or less and explain why you should care about them. So, search for The Hustle Daily Show and your favorite podcast app, like the one you're using right now.

And we're back here on Big Technology podcast with Kathy Edwards, the VP of Search at Google. All right, Kathy. So notes. This is from your blog post. Notes will let people share their knowledge right on search, helping others find the best answers for their weirdest or most wonderfully unique questions and discover what's most useful for them on the web. It's very interesting. It's like giving people a voice right. in the search results? Tell us more about this.

Yeah, I'm just so excited about this launch. So we're launching in labs. We're launching in the U.S. and then in India. And we're really bringing this layer to search results where people can provide their perspectives and, you know, advice to other users. And as you said, and in your team, I thought it was great. It's like, this is really about trying to make search a bit more interactive and a more human, which is what we're hearing, you know, I use this really want. So what's this going to look like in practice? Like, I'm searching Google and what do I see

in these results with these notes involved? Yeah, so you'll see the same results page that you normally see. Under each result, you'll see the opportunity to either add a note or see the notes that other people have left. You'll see how many people have left notes and you'll be able to click through and see the additional like information that other people have added onto that onto that link. And so can you give us like a practical example of, you know, I'm scrolling through my feed, I'm searching for X and I might see why. Yeah. I mean, here's a really great example, which I think highlights why this is so great for both users, but also actually for publishers. So let's say your recipe writer and you've written this recipe for like lasagna, right? And it's

got meat in it. If I'm a vegan, I could come along and say, hey, you know, add a note to this recipe and say to this recipe, you know, you are all saying, hey, I made this vegan by subbing out the fish sauce and, you know, with coconut aminos and used impossible meat instead. And suddenly, actually, as a user, I now get a range of different recipes. I don't just get recipes that are written, you know, specifically for vegan people. And I also, as a public, that that now enables that to rank for vegan lasagna recipes right so it sort of allows other people to come in and give their helpful tips in a way that I think will be really really helpful for uses that's very interesting so it could actually be like more of like a constructive

like almost like a genius like annotation yeah exactly the old rap genius style that's interesting But, you know, it's kind of, I was almost thinking of, like, people, like, rating, like, saying this was, like, a good recipe. This was a bad recipe. But you're imagining people being even more constructive here. Yeah, totally. Like, coming in and say, you know, if you're like, which hike to go on in the Bay Area and you get a result that has, like, 10 hikes, you know, somebody could come in. And let's say that result is a little bit out of date now, right? Like, somebody could come in and leave a note and say, actually, number nine is now shut down. You know, like, you can't go on a hike anymore. So it's a way of bringing. that human perspective, which will include sort of a fresh perspective to the web. People are going to say this is Google's way of making sure you spend even more time on Google. Like, do I need to even click the link to see all these other alternative hikes or updates? What do you think about that? I mean, we're really trying.

I feel like I've said it a lot already today. I'm going to say it again. We really care about the health web ecosystem. And a lot of this is actually. actually, we think a way of helping keep the web healthy by allowing people to share those human perspectives against it and, you know, things like the freshness, things like bringing in those different ways of looking at things. And so we're really focused on making it easy for users to, you know, come in and interact on search. And in making sure, like, this is

specifically designed to link through to the content. Right. So we're still focused on on making sure you're all able to get your voice out there. Right. Okay, that's good. When people come and they give negative feedback, they say this website is out of date. You know, I was looking for information on, I think you have a marathon example in your blog post. I was looking for information on the 2023 marathon. This is the 2022 marathon.

And by the way, you can no longer do like a proxy pickup for your race number. Is that going to impact the way that these pages rank in your search results? So it's very early. I could imagine over time, maybe, but right now, no. We're really very focused on learning. That's why we're launching in labs. This is obviously a pretty new thing for us to do. As you say, it's been 25 years.

And we're very interested to see what happens. and we're also, you know, very interested in engaging with publishers through this process and learning about what matters to you. You know, in the case of that marathon example, you gave, I actually think that's, like, incredibly helpful. And I wouldn't see that as a negative thing from a publisher perspective. But you could imagine other nodes being published that maybe publishers aren't so thrilled about.

So we're thinking about weight can integrate into search console and connecting publishers with insights from these things. Yeah, it's almost like taking the comment section and making it more prominent. And it's like, I don't know, from a publisher, it's like, man, you used to make me like click in before you punch me in the face and now you're going to do it right in search results. Like, do you have like some concerns about how nasty people can get in the comments and how what are the, what do you think about that? Like there's also this balance between like letting people say what they need to say and then, you know, trying to maintain some sense of civility on the most important, you know, page. that Google owns search result page. Totally. And yes, this is something we have been thinking very, very deeply about, as you can imagine.

Thankfully, for us, we have 25 years of ranking web content, which can also be pretty dicey as well. So we've learned a lot through that. We also have been spending a lot of time with our YouTube colleagues. I was about to say, you must be hanging out with YouTube people. We're best friends now. I mean, we've our best friends before. But we spent a lot of time.

We've learned a lot from their processes. We have a whole aligned policy. You know, we have a bunch of policies. We're definitely taking an approach that combines an algorithmic approach with a human moderation approach because we really do want to get this right. And this is, frankly, one of the reasons that we launched in labs is because we want to make sure that we have the time to sort of test and learn. Are people going to be able to, like, vote up and down the comments?

comments, whether they're helpful or not? Yeah, you'll be able to thumbs up and thumbs down. Okay, that's great. What is the key? Because not too long ago, I'm talking like three or four years ago. YouTube comments were one of the nastiest places on the internet. And now they are cleaned up. Like, you can actually go in and read lots of constructive things.

What I've heard is that YouTube effectively went in with a large list of words that just could not show up there, a lot of negative sentiment. and try to induce positive, constructive conversations. So is that the secret to YouTube? And if that's the case, can you really apply that to search? So I know the person very well who is responsible for cleaning up YouTube comments. She's amazing. What's her name? Shavon.

And I can tell you that she is, there's far more than what you just said in terms of like a block list of bad words. We got time. A ton of different things to a lot of. actually they've had a ton of different levers to really make progress there. And there are a bunch of them that are applicable in the search space as well. So we've, we've really dug in there and learned a lot from them. Yeah. I mean, it really is. It's wild the shift that's happened on YouTube. And it's interesting because I know this is like tough. It sort of goes to like how much content

moderation do you actually want to have. And you know, I've definitely, I hear the people who are like, you know, be careful to slippery slope. On the other hand, like, I think properties have a right to make sure that the stuff that they're incentivizing isn't, you know, just disgusting bile. And obviously, it's been a very strong arm approached by YouTube, which would, like, probably send, like, a lot of folks who've, like, followed the Twitter takeover into a tizzy. However, like, you know, the benefit is that the product is actually pleasant to be on now. Yeah, and I think that matters. I think it matters to uses that, that you have. I would agree.

Having said that, I think there are unique, like, the fact that this is an information product, as opposed to sort of more of an entertainment product, brings different challenges. And this is something very hard about. It's a really hard problem, I will say. And we've done, I mean, I can't tell you the amount of work we've done to proactively prepare for all the things we could think could go wrong. But this is one, as I say, another reason why we're launching in life. is we're going to see what happens and we're going to learn and we're going to really rapidly respond if we see things happening that we think of the wrong thing. I mean, I'm just like thinking through, like, imagine like a New York Times article surfaces

in search about the Israel Hamas war and just like the battle on social media there is like extremely intense and now you're inviting it effectively into the search results where there's going to even more at stake. You feel comfortable doing that? Well, so we will have policies in place and notes won't appear on every query. So the notes won't appear on that specific query because we're not going to, we want to make sure that we're not inviting. You know, if we think that a space is risky, really risky, we're not going to go there just yet. But they will be on the majority of queries.

Yeah, looking back at a Twitter example, they have this community notes product. Are you familiar with it? Yeah, I love that. It's actually, so did your team take any inspiration from it? Because I actually think it's the best, one of the best features that Twitter has put out there, which is like, yes, the tweets will stand. They're not going to take tweets down. But if they are outright lying or missing important context and get a large, you know,

percentage of you, they're going to look to the community to then go ahead and share the perspective. And maybe that's something that could also be applied here. Yeah, look, I think that product is fantastically well done. And I think it has added a lot to, you know, my experience on Twitter or X. It's definitely not exactly what we're trying to do with notes. We're really focused on, you know, that product is really like a fact-checking product, whereas we're trying to provide like a more general social, like human layer in search. But, yeah, we'll see how to evolves over time.

And I think it's definitely something we're watching closely and learning from. Okay. Great. So one area that this could come into play, and this is going to just bring us into our next topic is people see sites that have been like engineered for search engine optimization and they get frustrated. There's really no, you know, feedback mechanism there. So is this in some way a response are going to help Google fight against some of like the, the baity tactics, shall we say, that some professional and Black Hat SEO is used to try to rank higher? within Google? So firstly, I mean, just to talk about the SEO question for, you know, directly, look, we have a long history of really focusing on, on fighting spam in search. I'll also say that, like, SEO is not a bad thing. I mean, people advising less technically savvy people about how to structure their content to be helpful to users is not something that we say shouldn't happen. We, we do hear and have been focused on that, look, search is never a solve problem. That's one of the reasons that, like, I love working on search is because, you know, for all

of the amazing things it does, because it is this infinite product where people can do any query and your content corpus is totally changing, like, on a daily basis, you just have new problems to solve all the time. And so we have a very focused team working on this. you know, made for search engine content, and we actually did a launch recently that had a pretty big impact on that front. It's something we're continuing to work on. We hear what users are saying about this and we want to fix it. I do think the idea of like notes provides, you know, notes, you'll be able to create notes in the Google app, right? You need a Google account to do it.

So it provides a different level of layer of content that isn't, you know, really like these, these people who are trying to like spam search with this content can't come in really create it that way. And so it provides an interesting, like more authentic human layer, you know. Right. But the feedback layer, like, you know, I'm just imagining searching for that recipe. Let's say I'm following, what was a recipe that you had, the vegan? Vegan lasagna, yeah. vegan lasagna. I'm on vegan lasagna and I had to scroll for like 20 minutes to find the actual recipe. I could actually, which is I think SEO tactic. I'm not sure. They want to optimize for time

spent. Anyway, you could have people in the notes being like, hey, this is ridiculous. They kept me going, you know, my finger hurts from scrolling. And then I finally got the, you know, I mean, that could actually be really good feedback, both for Google and also for people clicking in. Like, imagine people like, it might then incentivize people to just give the right away and then get the good good notes saying hey this delivered exactly what I wanted and I didn't have to scroll and it was right there in the top of the window I mean I think I think it's you know anytime you build an ecosystem you don't like things sort of happen that emerge out and one of the reasons that we're launching in labs is to sort of see whether this sort of behavior

happens it's um right not I think it could go in a number of different ways and yeah I mean And suddenly, if this becomes a tool to, like, help users connect with the information that they really want more easily and incentivize website owners to create the content that users really want, I think that'll be great. Okay, awesome. I think it's cool. I'm going to turn it on in labs. I mean, I like turning everything on in labs.

Like, let's just go wild here. It's being a really great tester for us, Alex. Of course. Well, we'll see about that. But this should be fun. So just we talked about the SEO. The reason why this is top of mind is because the Verge, we're not going to litigate the whole SEO industry in the time we have left.

But it's worth bringing up that The Verge had this story by Amanda, Chicago Lewis. Actually, one of my former colleagues at BuzzFeed, she's to cover weed, which is interesting. I saw this. I was like, oh, she's gone from marijuana to tech reporting. And the title is the people who ruined the internet talking about SEO, people who are working in professional SEO. I felt the title was a little less. less charitable than it should have been.

It seemed a little bit sensational. But anyway, there was a few interesting things that I thought were worth highlighting here. The first, and I'd just like to share them with you and see if you have a response to. The first was Amanda spoke about how, I'm just going to read, Googling anything can now feel like looking up sneaker in the dictionary and finding a definition that sounds both incorrect and suspiciously as though it were written by someone promoting Nike. So the definition she puts is footwear that allows you to just do it.

Perhaps this is why nearly everyone hates SEO and the people who do it for a living. The practice seems to have successfully destroyed the illusion that the internet was ever about anything else than selling stuff. What do you think about that? Look, I think that we have a ton of data that suggests that the, Overwhelming majority of users are really happy with the search results that we return, as I said, there are always queries we can do better on. I mean, let me tell you, as somebody who works on a search engine,

you are constantly aware of the ways that you can improve. And I think that we definitely do need to continue to improve on this front. But I think we're, I just think that like it's categorically not true that that that's, It's people's typical data experience of using Google search. Yeah. The other thing that was interesting to me was this SEO professional name is Darren Babin. And I was talking about what this wave of generative AI is going to do, right, to what the web looks like and turn what search looks like.

So this is a quote from Babin that Amanda relays, all the assholes that are out there paying shitty link building companies to build shitty articles, now can go use a free version of GPT. Soon, he said, Google results will be even worse, dominated entirely by AI-generated crap design to please the algorithms, produced and published at volumes far beyond anything humans could create, far beyond anything we'd seen before. I mean, is this something you're seeing now? Are you anticipating more of this? How do you fight back against these GPT-created articles? So, firstly, I'll say, even as he points out, like, this isn't you, we've had years and years of people creating a bunch of content that's really not for other humans, right? And that's the guidance that we give is create content for humans, not for search engines, right?

And we have a very long history. We have a wonderful spam team who have a very long history of fighting off this content. Now, do I think that there's going to be a lot of AI-generated content on the internet as we move into the future? Absolutely. That I don't think we've, and we've issued guidelines, actually, that say that just because content is AI-generated doesn't make it bad, right? I think that generative AI will assist a lot of creators in creating content over time. What we do want is for it to be helpful to human beings and not creative for such engines. and we fought it before and we'll find it again in the future.

So while the internet might contain a lot of this stuff, we actually don't, I know that, you know, search engines index the internet. We actually throw out a lot of content and it doesn't even go into our index because it is spam. And we'll continue doing that on this front. So there's fear that people create these websites and then the models train on them

and then they sort of, you know, they chew their own cud, so to speak, they throw it up, they ingest it, they chew it up again, and they puke it out again. I was actually way more visceral than I expected, but it's actually pretty, like if you've been to a farm, this stuff happens, people. It does. Okay. Anyway, that for yours is a little bit overblown, you think?

I mean, I'm obviously not the team that builds a large language models, but I think they're pretty intentional about this stuff. And it's not just going to be everywhere in the search results. Yeah, we're very intentional. intentional about what goes into the search index. We're very intentional about how we rank, and we think we're pretty good at fighting this sort of spam. Okay, great.

Let's go full circle here as we come to a close. When ChatsyPT initially came out and then Bing very quickly integrated it, there was a lot of concern from industry watchers and Google shareholders, the whole crew, right? That, like, search was over and now everything was going to happen through the chat. Bat bots. You're sitting as a VP at search at Google. I'm curious, what has it been like from your perspective in terms of the magnitude of this challenge? I mean, obviously, Google has its own answer now and Bard, and you have generative search experiences. How has this played out over the past, you were looking at a year since chat GPT came out? What's the magnitude of what you've

actually felt inside Google? I mean, honestly, and this is a personal answer. Like, this is the most exciting time in my whole career. It's so great to the possibilities that I can see about really helping people connect to information. It's super exciting to do what I do and to feel like we're creating the future together. It's amazing. And I'm really, you know, we just have these really clear guiding principles as we go through this. And one of those fundamental guiding principles is that while we think that technology can really help enable creation in really amazing, powerful ways, I think that fundamentally people want to hear about information from other people. And so figuring out how we build something that combines the best of those,

and that's why you see things like follow and notes and perspectives that we're launching today that are not about generative AI at all, to me, the combination of these things is what's going to be the really amazing future. And so we're really exciting to build, excited to build that. Okay, but outside of having such a good time doing this, do you think the format of search is going to change? I mean, is it going to fundamentally change or is it, go ahead. Look, I think if you go back to search in like 2000, it looks pretty different today, actually.

It's still a page with links, though. Like, people are saying it's now going to be a chatbot. I think it's still going to be a page of links. I really believe that. Kathy Edwards, thank you so much for joining. Thank you, Alex. Great way to end it. Thank you.

It's always great to speak with you. Appreciate you being here and going through all this stuff with me and no questions, off limits or out of balance, which we always appreciate here. So thank you so much. All right, everybody. Thank you so much for listening, everyone. Great to have you here.

Thank you, Nate Gwattany for handling the audio, LinkedIn, for having me as part of your podcast network. And all you, the listeners, if you enjoyed today's show, please give us a rating of five stars on Apple Podcasts or Spotify. And maybe one day you can do it on Google, with notes. Fingers crossed on that front. And if you're a first time listener, please hit follow or subscribe. We do these once a week on Wednesday and then we break down the news every Friday. All right, that'll do it for us. We'll see you next time on Big Technology Podcast.