In what's become a bit of a Decoder tradition, I spoke with Google CEO Sundar Pichai in person after I/O. The conference this year was all about AI, particularly a slew of actual AI products, not just models and capabilities.
To Sundar, this marks the beginning of a new era for search and the web overall. So I had to ask: what happens to the web when AI tools and eventually agents do most of the browsing for us? It was a very Decoder conversation.
To Sundar, this marks the beginning of a new era for search and the web overall. So I had to ask: what happens to the web when AI tools and eventually agents do most of the browsing for us? It was a very Decoder conversation.
Category
🤖
TechTranscript
00:00Sundar Parchai, CEO of Alphabet and Google. Welcome back to Decoder.
00:06Good to be back. It feels like a nice tradition post-IO to be talking to you, so good to be back.
00:11I think this is the third year we've done this after IO. I'm excited. Thank you for keeping the tradition alive.
00:17Lots to talk about. You announced a lot of things yesterday during the keynote.
00:21There's AI mode rolling out for US users, big updates to Gemini, there's VO3 and Imagine,
00:26the generators. You solved Pokemon with robots, which is very exciting. My takeaway yesterday was
00:32that Google feels very confident now. There's a real confidence about the technology coming to life
00:38and the products. A lot of things are shipping imminently. What's the one piece that gave you
00:44that confidence? Is it just the volume of things that are shipping? Is it one technology that
00:48clicked into being ready for consumers? Where is it coming from?
00:52Look, I think it comes from the depth and breadth of the AI frontier we are pushing in a more
01:01fundamental and foundational way. We spoke a lot about this theme called Research Becomes Reality,
01:07but it is always felt we are a deep computer science company and we've been AI first for a while.
01:13Putting all that together and bringing it to our products at the depth and breadth is what I think
01:20it's really pleasing to see. For example, people may not have noticed it much. It was so quick.
01:27We spoke about text diffusion models in the middle of the whole thing, but we're pushing the frontier on
01:32all dimensions. Demis spoke about world models. I think, to me, that's the exciting part, how deep
01:40we are pushing this frontier and then bringing it to users. Maybe that's what makes it feel that way.
01:46You mentioned research into reality several times. Obviously, a lot of these projects have been
01:50cooking in labs for a long time. You've said many times you think AI will be as profound as electricity
01:56over the past many, many years that you and I have talked about it. But you said something yesterday that
02:00I think adds to that, which is that we're in a new phase of the platform shift. People have talked about
02:07AI being a platform shift for quite a while. That always has meant to me that there's a user interface
02:11platform shift coming. We're going to interact with computers in natural language in more natural
02:15ways and they will interact with us back in that same way and everything will change.
02:19Is that the platform shift?
02:20Yeah, you are right. Each of these platform shifts changes many things on the I.O. front.
02:29Nothing to do with Google I.O., just I.O. in the traditional computer science sense.
02:33You could feel it. Yesterday when I watched the Android XR, I've used them and played around with
02:38them. But watching it, two people talking in different languages, you can envision the future
02:44one day where it will actually be seamless. In a way, you couldn't have done it with phones.
02:47You couldn't have done it without AI, because there's nothing in your way. You're looking at the
02:51other person and talking. That is an element of platform shift, but there are many more elements.
02:58This is the only platform where I think the actual platform is, over time, capable of
03:06creating and self-improving and so on. In a way, we could have never talked about any other platform
03:11before. That's why I think it's much more profound than the other platform shifts. It will allow people
03:17to create new things in a way, because at each layer of the stack, there's going to be profound
03:23improvements. I think that virtuous cycle you get in terms of how you can unleash this creative power
03:31to all of society, be it software engineers, be it creators, I think that is going to happen in a
03:38much more multiplicative way. When I say it's a next phase, I'm talking about that part too.
03:43Let me just make that more concrete for people. I think the last platform shift we all
03:48understand is the shift to mobile. That's right.
03:50And that was, we're going to have multi-touch, we're going to have faster cell connections,
03:55we're going to increase processing power that can go with you everywhere. And then there was a layer
03:59of applications that was enabled by all of those things. You can push a button and a Toyota Camry
04:04will show up wherever you are in the world. It's a very powerful thing that required all of those
04:08ideas. How would you describe the phase we're in now compared to that?
04:12The phase of this, because the first phase of AI was the Transformers work and the models work,
04:17and we can all see this capability. The second phase, what is it to you?
04:21Just imagine, when the internet came, blogging became a thing. Pre-internet, very few people had
04:34a means by which they could put out their thoughts out to the world. When the internet came a new medium,
04:40it allowed people to create and express themselves in a new way. With mobile cameras and you could
04:45shoot and you could create videos, look at what's happened with YouTube. For me, the similar part of
04:53this is, we're all talking about things like wipe coding. Yesterday, you saw VO3. We are now in that
05:01phase. I think people are going to be able to create AI applications, you can call it.
05:05But wipe coding, there are many names to it, but that power is yet to be unleashed. You're barely
05:13scratching the surface. These models are now, they aren't quite there. You can do one-shot coding,
05:22but you really need to be a programmer to go iterate and create something with polish. That frontier is
05:28evolving pretty rapidly. I think you're going to see a new wave of things just like mobile did,
05:33just the internet did. I came to Google at the time when AJAX was the revolution. The fact that
05:42the web became dynamic, you had things like Google Maps, Flickr, Gmail, all that suddenly came into
05:48existence. I think AI is going to turbocharge in a way we haven't seen before.
05:53It feels to me like what you're describing is we're in the phase where the products are developed.
05:58The capabilities were the first phase, and now we're going to make some actual products.
06:01And more people can build products than ever before. That's the multiplicative part I'm talking
06:06about. Not just this platform helps you create more products. The process of creating, developing,
06:11et cetera, is going to be accessible to a much wider swath of humanity than ever before.
06:16I'm wondering, when you look at the landscape of products that exist now, most people experience AI in
06:21Gemini or ChatGPT as a chatbot, the general purpose interface to a bunch of knowledge that we'll talk to
06:28you. What products do you see that will have the same kind of impact as the Web 2 products
06:32you were talking about, besides the general purpose chatbot?
06:36Well, obviously, you've seen a wave with coding IDEs. That entire landscape is, I can't even keep
06:44track of how many new companies have come into, and people are using a lot of it. Yesterday, we showed a
06:52bunch of partners with whom we are working. That's an area where, because coding is where maybe AI is
06:58making the most progress, you're seeing the application later, at least in terms of code editors,
07:06really come into vogue. We've had success with Notebook LM. We're launching products like Flow
07:12yesterday. Flow is a new product which allows you to create and imagine. Those are all applications we
07:17are doing. I think others are beginning to do. People are working on legal assistance. There are
07:23all kinds of startups. I was recently in a doctor's office. They do have AI to transcribe the whole
07:32thing, put it all in reports, and so on. That's an enterprise application layer. It completely works
07:37different than when I went to that visit two years ago. All that change is happening across the board,
07:44but I think we are just in the early stages. You will see it play out over the next three to five
07:48years in a big way. Did you ask your doctor what model their transcription software is running out?
07:53No, I didn't. No, I didn't. Yeah. One of the reasons I'm asking this and I'm pushing on this is
07:58the amount of huge investment in the capability from Google and others has to pay off in some products
08:04that return on that investment. Notebook LM is great. I don't think it's going to fully return on
08:09Google's data center investment, let alone the investment in pure AI research. Do you see a
08:15product that can return on that investment at scale? Look, do you think in 2004, if you had looked
08:20at Gmail, which was a 20% project which people were internally using internally as an email service,
08:29how would we be able to think about Gmail as what led us to do workspace, get into the enterprise?
08:36I made a big bet on Google Cloud, which is tens of billions of dollars in revenue today. My point is
08:47things build out over time. Think about the journey we have been with Waymo. I think one of the mistakes
08:56people make often in a period of rapid innovation is think about what is that next big business versus
09:02looking at the underlying innovation and say, can you build something and put out something which
09:06people love and use, and out of which you do the next thing and create value out of it.
09:14So, when I look at it, AI is such a horizontal piece of technology across our entire business. It's
09:21why it impacts not just Google Search, YouTube, Cloud, all of Android. You saw XR, etc., Google Play,
09:32things like Waymo, isomorphic, which is based on AlphaFold. I've never seen one piece of technology
09:38which can impact and help us create so many businesses. AI is going to be so useful as an
09:44assistant. I think that people will be willing to pay for it. We are introducing subscription plans.
09:49There's a lot of headroom ahead, I think. Obviously, that's why we are investing because we see that
09:56opportunity. Some of it will take time. It may not be always immediately obvious. I gave the Waymo
10:03example. In fact, three years ago, the sentiment on Waymo was quite negative three years ago.
10:10As a company, we increased our investment into Waymo at that time because you're betting on the
10:15underlying technology and you're seeing the progress of where it's going. These are good
10:20questions. In some ways, if you don't realize the opportunities, that may constrain the pace of
10:27investment in this area, but I'm optimistic we'll be able to unlock new opportunities.
10:33One reason I wanted to start here as the foundation of the conversation is you showed off Android XR
10:38yesterday. You showed off some prototype glasses. You have some partners making glasses.
10:41A lot of people think augmented reality glasses powered by AI will be the realization of the full
10:48platform shift. You will have an always-on assistant that can look at the world around you. You showed
10:52some of those demos yesterday. The form factor will change. The interface will change. This will be a
10:56market as big as smartphones were. How close do you think we are to that as a mainstream product?
11:00It was a nice reflective moment all the way back from Google Glass. Wearing the product, I think there's
11:10a difference between goggles and glasses. Everyone at Verge understands this as well, but obviously we are
11:18also shipping goggles. We have announced products with Samsung to come later this year. On the XR side,
11:26I think I'm excited about our partnership with Gentle Monster and Warby Parker. We'll have products in the
11:32hands of developers this year. But I think those products will be pretty close to what people will
11:39eventually see as final products. I'm excited. I think the pace is actually pretty palpable.
11:47I'd be shocked if you and I were sitting next year. I wasn't wearing one of that when I'm doing the…
11:54Do you think that will be a mainstream iPhone-level replacement product? Because there's a lot of
12:00hardware that needs to get developed along the way to pull that off.
12:02You know, there are… Look, you're wearing something on your face. People are like,
12:05I have prescription, right? The bar is higher, I think, in terms of making the experience seamless
12:13enough that you're willing to wear it in your face and enjoy it for all. I don't think next year
12:19it's as mainstream as what you're talking about. But would millions of people be trying it?
12:25I think so. Yeah. Both are true, I think.
12:27So I have to ask you, just before we sat down, OpenAI announced that Johnny Ive was selling a
12:33company he had started called I.O. to the company and Ive and his design consultancy love from would
12:38take over all design. They didn't announce the product, but they said it's the future of computing,
12:42and it's coming next year. Do you anticipate more of that competition that your competitors
12:48who don't have a smartphone operating system will go even harder in this direction?
12:51I'm looking forward to an OpenAIo announcement ahead of Google I.O. the night before.
12:58First of all, look, stepping back, I mean, Johnny Ive is, you know, one of a kind. You know,
13:04you look at this track record over the years, I've definitely… I've met him only once or twice,
13:09but, you know, I've admired his work, obviously, like so many of us. So I think it's exciting. Shows,
13:17you know, this is why I feel like it's such… there's so much innovation ahead. And I think
13:22people tend to underestimate this moment. In some ways, people tend to… I always like to point out,
13:29when the internet happened, Google didn't even exist. Right? I think what people… So when you think
13:35about… I think AI is going to be bigger than the internet. There are going to be companies
13:39products, categories created, which we aren't aware of today. So I think the future looks
13:45exciting. I think there's a lot of opportunity to innovate around hardware form factors at this
13:52moment with this platform shift. So I'm looking forward to seeing what they do. You know, we are
13:57going to be doing a lot as well. And I think, you know, it's an exciting time to be a consumer.
14:03It's an exciting time to be a developer. So I think looking forward to it.
14:06I've, in that video, described the phone and the laptop as legacy platforms,
14:10which is very interesting considering his own history. Are you all the way there that the phone
14:15and laptop are legacy platforms? Look, I think these things, if anything,
14:22I found through this AI moment using the web a lot more, right? Because I'm like,
14:27it's easier to create a VO3 video on my browser and a big screen, right? And so I think the way I've
14:38internalized this is computing will be available and you don't have to make these hard choices.
14:43You know, computing will become so essential to you. You're going to have it in multiple ways
14:50around you when you need it, right? Like I use a phone, a tablet, a laptop, and I have my workstation,
14:58right? And so I have the breadth of it. But over time, it makes sense to me,
15:07at some point in the future, consuming content by pulling out this black glass display rectangle
15:13in front of you and looking at it is not the most intuitive way to do it. But I think it's going to take
15:17some time. I feel like we could do a full hour just on Android tablets and where they could go.
15:21We're going to come back for that. Yeah.
15:23A big part of what you're describing implicates search in really big ways, right? We're going to
15:28be surrounded by information. Search or Gemini or some future Google product will organize that
15:33information, take action for you across the web in some way, and you will have a companion. And maybe
15:39you only pull out your tablet to watch a video or something. A lot of what's going on with search has
15:44downstream effects on the web, downstream effects on information providers broadly.
15:49Last year, we spent a lot of time talking about those effects. Are you seeing that play out the
15:53way that you thought it would? Look, it depends. I do think people are consuming a lot more
16:03information. And web is one specific format. So we should talk about the web, but zooming back out,
16:10you know, there are new platforms like YouTube and others too. So I think people are just consuming a
16:15lot more information, right? So it feels like an expansionary moment. I think there are more
16:21creators, people are putting out more content, you know, and so people are generally doing a lot more.
16:28Maybe people have a little extra time in their hands. And so it's a combination of all that.
16:32On the web, look, things which have been interesting, and you know, we've had these conversations for a
16:38while, you know, obviously in 2015, there was this famous, the web is dead. You know, I always have
16:44it somewhere around, you know, which I look at it once in a while. Predictions, it's existed for a
16:49while. I think the web is evolving pretty profoundly. I think that is true. When we crawl, when we look at
16:56the number of web pages available to us, that number has gone up by 45% in the last two years
17:02alone. Right? So that's a staggering thing to think about. Can you detect if that number, if that
17:07volume increase is because more pages are generated by AI or not? This is the thing I may be worried about
17:12the most, right? It's a good question. We generally have many techniques in search to try and understand
17:20the quality of a page, including whether it was machine generated, etc. This trend, that doesn't
17:26explain the trend we are seeing, right? So generally, there are more web pages, right? So, you know,
17:33at an underlying level, so I think that's an interesting phenomenon. I think everybody as a
17:39creator, like you do at Verge, I think today, if you're doing stuff, you have to do it in a
17:44cross-platform, cross-format way. I look at the quality of video content you put out. It's very
17:51sophisticated, right? You know, and very different from how Verge used to be maybe
17:56five to ten years ago, right? It's profoundly changed. I think things are becoming more dynamic,
18:02cross-format. I think another thing people are underestimating with AI is,
18:09AI will make it zero friction to move from one format to another, right? Because our models are
18:14natively multimodal. We kind of tease people's imagination with audio overviews in Notebook Ella.
18:21The fact you can throw a bunch of documents at it, and you have a podcast, and you can join and
18:25learn from it. So I think this notion, the static moment of like, you produce content by format,
18:33whereas I think machines can help translate it from, it's almost like different languages,
18:38and they can go seamlessly between, I think is one of the incredible opportunities to be unlocked
18:44right ahead. And so, but maybe I didn't want to drift from the question we were having. But look,
18:51I think people are producing a lot of content, and I see consumers consuming a lot of content.
18:57And, you know, it's, we see it in our products, others are seeing it too. So that's, that's how
19:05I would probably answer at the highest level. The way I see it currently is that the web is at
19:09an all-time high as an application platform, right? The fact that Figma exists and is as successful as
19:15it is in its primary interfaces as a web app is, I think, remarkable. A lot of the products you are
19:20talking about are expressed as web apps. Even some of the most interesting search results you showed
19:24yesterday are, you know, Google would generate a custom web app for you and display it in the
19:29search result to do some data visualization. I think that's all looking incredible. I think the
19:33web as a transaction platform is reaching new highs, especially with rulings that mean smartphone
19:38makers have to let people push transactions to the web. There's something very interesting happening
19:42there. As a media platform, it feels like it's at an all-time low, right? If I was starting The
19:47Verge. You mean the web as a media platform? The web as a media platform, as an information platform.
19:50If I was starting The Verge today with 11 of my co-founder and friends, we would start a TikTok
19:56channel. We might start a YouTube channel. We would definitely not start a website with the
20:00dependencies we have as a website today. And that's the dynamic that it feels like AI is pushing on
20:06even harder. I'm not fully sure I agree, right? I think, you know, I think if you were to go and
20:11restart Verge again, I bet you would start a, you would have an extraordinary web presence.
20:16At best, no, I've thought about this a lot. I think at best our web presence would look like
20:20a substack or a ghost or something, right? Maybe. Look, I, you know, I'm not, you know,
20:25I'm not fully sold on that. Like, you know, but you know, you know the space. I acknowledge you
20:30know that space better than I do. So I don't mean to be, you know, I don't have that intuition,
20:36which you do here. But look, I, I see, in fact, you say the web application platform is an all-time
20:43high. But I've looked, you know, I was wipe coding with Replit a few weeks ago. You know, create,
20:51I mean, the power of 40 are going to be able to create on the web. We haven't given that power to
20:57developers in 25 years. So that is going to come ahead. So, you know, it's not exactly clear to me.
21:08You know, maybe today you're looking at and say, I won't put all the investment in because it looks
21:12like a lot of investment to do that. But that may not be true two years out, right? Like, you know,
21:20if you feel like you would create a TikTok channel, then, you know, maybe with like two
21:25percent extra effort, if you could have a robust web presence, why won't you, right? And so,
21:30so I'm not fully, you know, I'm not fully sold on it, but I think it's a good question to ask.
21:35But, you know, you have to somehow reconcile that with the fact that
21:39overall, that traffic seems to be growing. We see more web pages in our, somewhere we have to
21:44explain all of that too. You know, the publishers, as they often do,
21:48responded to Google ION announcements. So the News Media Alliance, after AI Mode was announced
21:53yesterday, I would say they're very upset. Here's a statement from the president of the News Media
21:59Alliance. Links were the last redeeming quality of search that gave publishers traffic and revenue.
22:04Now Google takes content by force and uses it with no return, no economic return. That's the
22:08definition of theft. And they go on to say the DOJ lawsuits must address it. That's pretty furious.
22:13That's not a negotiation, right? That's a, we just want this to stop. How do you respond to that
22:18very loud set of people who say, yeah, okay, maybe it's growing somewhere, but for us,
22:23it's crushing our businesses? Look, I, you know, first of all,
22:29through all the products, I mean, AI Mode is going to have sources and, you know,
22:36we're very committed as a direction, as a product direction to make sure. I think part of why people
22:41come to Google is to experience that breadth of the web and go in the direction they want to,
22:48right? So I view us as giving more context. Yes, there are certain questions which may get answers,
22:54but overall, and that's the pattern we see today, right? And if anything over the last year,
22:59it's clear to us the breadth of where we are sending people to is increasing. And so I expect that to be
23:05true with AI Mode as well. But if it was increasing, wouldn't they be less angry with you?
23:10Look, look, I, more than, you're, you're always going to have areas where people are robustly
23:17debating value exchanges, et cetera, right? Like app developers and platforms that's not on the web,
23:22et cetera, right? It's inherently, you know, you know, there's always going to be, when you're running
23:27a platform, these debates, I would challenge, I think, more than any other company. We think about,
23:36we prioritize sending traffic to the web. No one sends traffic to the web in the way we do. I look
23:43at other companies, newer, emerging companies, they openly talk about it as something they are not
23:48going to do, right? We are the only ones which make it a high priority, agonize, and so on. Look, we'll
23:54engage. And, you know, we've always engaged. There are going to be debates through it all,
23:59but we are committed to, you know, I've said this before, everything we do across all, you will see
24:06us five years from now sending a lot of traffic out to the web. I think that's the product direction
24:11we are committed to. I think it's what users look for when they come to Google, right? And the nature
24:17of it will evolve. But, you know, I, you know, I'm confident that that's the direction we will continue
24:23taking. Is there public data that shows that AI overviews and AI mode actually send more traffic
24:28out than the previous search engine results page? Look, it's, you know, the way we look at it is,
24:33I mean, obviously, we take a lot of, we're definitely sending traffic to a wider range of
24:39sources, publishers. And, and big calls, just like we've done over 25 years, we've been through the
24:44same with featured snippets. The, the quality of, you know, it's higher quality referral traffic,
24:50too. So, right. And we can observe it because the time when people spend as one metric and there
24:56are other ways by which we measure quality of our outbound traffic is also increasing.
25:00So, and overall through this transition, I think generally AI overviews are also growing.
25:05And, you know, the growth compounds over time. So whenever we work through these transitions,
25:10it ends up posted. That's how Google has worked for 25 years. And, you know, and we end up sending
25:17more traffic over time. So that's, that's how I would expect all this to play out.
25:20So why do you think that there is so much general economic turmoil on that side of the house,
25:25right? If you're sending more traffic and the goal over time is to make sure that that works out,
25:31we're a year into it, right? And it, it doesn't seem to have gotten better over there.
25:35No, look, we are sending traffic to a broader source of people. People may be,
25:40you know, surfacing more content, looking at more content. So somebody individually may see less,
25:45I mean, there are all kinds of, at the end of the day, we are reflecting what users want,
25:50right? You know, if you do the wrong thing, users won't use their product, go somewhere else,
25:56right? And, and, and so you have, you know, you have all these dynamics underway. And I think we have
26:02genuinely, you know, we took a long time designing it, AI overviews, and we are constantly iterating
26:09in a way that we prioritize, you know, sources and sending traffic to the web.
26:14I mean, my, my criticism of this industry, just to be clear, is that everyone's addicted to Google,
26:18and it would be better if they weren't, but they're addicted to Google, right? And they're,
26:21they're feeling it. And then on top of that, you see, you've mentioned several times, like overall
26:26queries are increasing on Google surfaces, but they're, they're changing, right? They're getting
26:30longer, they're getting more complicated. AI mode might walk you through several steps.
26:36Maybe some people are searching on TikTok now. Eddie Q on the stand in the trial the other day said,
26:42search in Safari for the last month dropped for the first time in 22 years. That's a big stat.
26:47Obviously your stock price was affected by it. There was a statement. Is that trend bearing out
26:52that the standard Google search is dropping from devices and different kinds of searches are increasing?
26:57No, look, we've been very clear. We are seeing overall query growth in search.
27:01But have you actually seen the drop in Safari?
27:07Look, we have a comprehensive view of how we look at data across the board. There's a lot of,
27:13there can be a lot of noise in search data. But everything we see tells us we are seeing query
27:19growth, including across Apple's devices and platforms. Specifically, I think we quantified the
27:25query growth from AI overviews. And what's, what's healthy is that the query growth is continuing to
27:32grow over time. This is what I've said before. It feels very far from a zero sum game to me. I said
27:38this last year, it's interesting. We spoke about TikTok, right? Think about like how profound a new
27:43product TikTok was. How has YouTube done since TikTok has come? Right? You could ask all these questions
27:50there. Like why is it that TikTok comes and YouTube has grown? I think what we always underestimate in
27:57these moments is people are engaging more, doing more with it. We are improving our products. And,
28:02and so, you know, so that's how I would, I would think about these moments.
28:06Let me just broaden that out to agents, right? I watched Demis Asabis yesterday. He was on stage with
28:11Sergey Brin and Alex Kanchowitz asked him, what does a web look like in 10 years? And Demis said,
28:19I don't know that an agent first web looks anything like the web that we have today. I don't
28:22know that we have to render web pages for agents the way that we have to see them. That kind of
28:27implies that the web will turn into a series of databases for various agents to go and ask questions
28:32to, and then return those answers. And I, I've been thinking about this in the context of services like
28:36Uber and DoorDash and Airbnb. Why would they want to participate in that and be abstracted away for
28:43agents to use the services they've spent a lot of time and money building?
28:46Two things, though. First, there's a lot to unpack. It's a fascinating topic.
28:51The web is a series of databases, et cetera. We build a UI on top of it for all of us to consume.
28:56This is exactly what I wanted is the web is a series of databases.
28:59It is. But I think, I think, I listened to the Demis Sergey conversation yesterday. I enjoyed it. I
29:05think he's saying for an agent first web, like, so, you know, for a web which is interacting with
29:10agents, you would think about how to make that process more efficient. Today, you're running a
29:17restaurant. People are coming, dining, and eating, and people are ordering takeout and delivery.
29:23Obviously, for you to service that takeout, you would think about it different than all the tables
29:29and the clothing and the furniture. Both are important to you. You could be a restaurant which
29:35decides, I'm not going to participate in the takeout business. I'm only going to focus on the dining
29:42experience. You're going to have some people vice versa. I'm going to say, I'm going to go all in on
29:47this and run a different experience. To your question on agents, think of agents as
29:54a new powerful format. I do think it will happen in enterprises faster than consumer. In the context
30:03of an enterprise, you have a CIO who is able to go and say, I really don't know why these two things
30:08don't talk to each other. I'm not going to buy more of this unless you interoperate with this.
30:14I think partly why you see on the enterprise side a lot more authentic experiences. On the consumer
30:21side, I think what you're saying is a good point. People have to think about and say,
30:25what is the value for me to participate in this world? It could come in many ways. It could be
30:31because I participate in it and overall, my business grows. Some people may feel that it's
30:38disintermediating and it doesn't make sense. I think all of that can happen.
30:46But users may vote with their feet. You may find some people are supporting the agent experience
30:55and your life is better because of it. You're going to have all these dynamics and I think
31:01they're going to try and find an equilibrium somewhere. That's how everything evolves.
31:05I think the idea that the web is a series of databases and we change the internet. First of
31:08all, this is the most decoder conversation that we've ever had. I'm very happy with it.
31:11I had Dara from Uber on the show. I asked him this question from his perspective. His answer
31:17attracts yours broadly. He said, first we'll do it because it's cool and we'll see if there's value
31:22there. If there is, he's going to charge a big fee for the agent to come and use Uber. Because
31:27losing the customer for him, losing the ability to upsell or sell a subscription, none of that is
31:33great. It's the same is true for Airbnb. I keep calling it the DoorDash problem. DoorDash should not be
31:39a dumb pipe for sandwiches. They're trying to actually run a business and they want the customer
31:44relationship. If the agents are going across the web and they're looking at all these databases
31:49and saying, okay, this is where I get food from and this is where I get cars from and this is where I
31:53book. I think the demo was booking a vacation home in Spanish. I'm going to connect you to that
31:58agent, that travel agent. Is it just going to be tolls that everyone pays to let the agents work?
32:05Because the price, I still don't, the CIO gets to just spend money to solve the problem. He says,
32:10I want this capability from you. I'm just going to pay you to do it. The market, the consumer market
32:15doesn't have that capability, right? All kinds of models may emerge, right?
32:22I can literally see, envision 20 different ways this could work. Consumers could pay a subscription for
32:28agents and the agents could rev share back. That is a CIO-like use case you're talking about. That's
32:36possible. We can't rule that out. Rule that out. I don't think we should underestimate. People may
32:41actually see more value participating in it. I think this is tough to predict, but I do think
32:49over time, if you're removing friction and improving user experience, it's tough to bet against those
32:59in the long run. I think if in general, you're lowering friction for it and people are enjoying
33:07using it, somebody is going to want to participate in it and grow their business. Would brands want to
33:16be in retailers? Why don't they sell directly today? Why won't they do that? Because retailers
33:23provide value in the middle. Why do merchants take credit cards? I'm just saying. There are many
33:35parts. You find equilibrium because merchants take credit cards because they see more business as part
33:40of taking credit cards than not, which justifies the increased cost of taking credit cards.
33:46It may not be the perfect analogy, but I think there are all these kinds of effects going around.
33:51What you're saying is true. Some of this will slow progress in agents just because we all are
33:57excited about A2A and MCP. We think, no, some of it will slow progress, but I think it'll be very
34:07dynamic. There's other pressures on Google. There's antitrust pressures. The government would like you to
34:12sell Chrome. Can you do all the things you want to do if you're made to sell Chrome?
34:17Look, I don't want to comment on ... Look, we are in a legal process.
34:24I look at having directly been involved in building Chrome. I look at ... I think there are very few
34:32companies which would have ... We not only improved our product, we improved the state of the web by
34:39building Chrome. We open sourced it. We provide it as Chromium. Everyone else has access to the
34:45browser. I think the amount of R&D, the amount of innovation we put into it, the investments in
34:51security, et cetera, we do.
34:52But if you're made to sell it, can you do all the things that you want to do?
34:57Look, I don't think that's the scenario we're looking at, but stepping back as a company,
35:04look, I think as a company, I think of ourselves as a deeply foundational technology company,
35:11which then translates it into products that touch billions of people. We do it across many,
35:17many things. Of course, I think, look, as a company, we are going to continue investing and
35:21doing our best to innovate and build a successful business in all scenarios. This is how I would
35:26answer it. The Trump administration is extremely transactional, I would say. The tech industry has
35:32a new relationship with Trump in a second term. You were at the inauguration. Have you had conversations
35:37about what a settlement might look like and what the Trump administration might demand to make these
35:41problems go away? No, no. Look, obviously, we've engaged with the DOJ like we do over the years
35:49on in the context of all the cases we have. That's how we normally do these conversations.
35:56Trump has very publicly said he doesn't like his search ranking and he wants it changed in some way.
36:00Would you ever adjust the search ranking for Donald Trump?
36:03No. Today, the way Google search works is no person in Google can influence the ranking algorithm.
36:14AI mode is different, right? We've seen system prompts adjusted in very chaotic ways from some
36:19of your competitors. Is that something that you would be open to in a world where you're serving
36:23the full answer? Would you adjust the AI mode responses in response to political pressure?
36:28No.
36:29No. Because we've seen, certainly in Grok and others, the system prompts change the answers
36:35in dramatic ways. The way we do ranking is sacrosanct to us. We've done it over 25 years.
36:45We make a lot of ranking signals. We take into account and stuff. If there's broad feedback from
36:52people that something isn't working, we will look at it systematically and try and make changes.
36:57But we don't look at individual cases and ever change ranking.
37:01When you think about those sources of information, one of the things that I have been thinking
37:04about a lot is the CDC webpages have changed a lot recently. Diversity, equity, inclusion language
37:11has been removed from pages across the government. Those used to be very high-ranking sources in
37:16Google search. We just implicitly trusted the CDC's webpages in some ways. Are you reevaluating that?
37:22There might be misinformation on some of these pages that then get synthesized into AI results?
37:27No. It's a misunderstanding of how search works. We don't individually evaluate the
37:31authoritativeness of a page. It's what our signals does. In PageRank, obviously, our signals are
37:42multiple orders of magnitude more complicated than PageRank today. But to use PageRank as an example,
37:46we weren't the ones determining how authoritative a page is. It's how other pages were linking to it,
37:52like an academic citation, etc. So we are not making those decisions today, and so I don't see that
37:59changing. As you synthesize more of the answers, do you think you're going to have to take more
38:02responsibility for the results? Look, we are giving context around it, but we are still anchoring it in
38:08the sources we find. But we've always felt a high bar in Google. Last year, when we launched AI Overviews,
38:19I think people are adversely querying to find errors, and the error rate was one in seven million
38:26for adversarial queries. But that's the bar we've always operated as a company. I think,
38:33to me, nothing has changed. Google operates under a very high bar. That's the bar we strive to meet,
38:39and our search page results are there for everyone to see. With that comes that natural
38:46accountability, and we have to constantly earn people's trust. So that's how I would approach it.
38:51Yeah. What do you think the marker is for the next phase of the platform shift after this one? We
38:54open by talking about we're in a second phase. What's the marker for the final phase, or the third phase?
39:00Of the platform shift, you mean? Yeah. Of the AI platform? What are you looking for as the next
39:03marker? Oh, look, I think the real thing about AI, which I think why I've always called it more profound,
39:12is self-improving technology. Having watched AlphaGo start from scratch, not knowing how to play Go,
39:23and within a couple hours or four hours, be better than top-level human players. In eight hours,
39:31no human can ever ask for it to play against it. That's the essence of the technology, obviously,
39:39in a deep way. Look, I think there's so much ahead on the opportunity side. I'm blown away by
39:49the ability to discover new drugs, completely change how we treat diseases like cancer over
39:56time, et cetera. The opportunity is there. The creative power which I talked about,
40:04which we're putting in everyone's hands. The next generation of kids, everyone can program and
40:10if you think of something, you can create it. I don't think we have comprehended what that means,
40:15but that's going to be true. The part which the next phase of the shift which is going to be really
40:20meaningful is when this translates into the physical world through robotics. That aha moment of robotics,
40:28I think when it happens, that's going to be the next big thing we will all grapple with. Today,
40:33they are all online and you're doing things with it, but on one hand, today, I think of Waymo as a
40:41robot. We are driving around a robot, but I'm talking more general purpose robot. When AI creates
40:50that magical moment with robotics, I think that'll be a big platform shift as well.
40:54Yeah, I'm looking forward to it. Next year, we're going to do this with glasses
40:57and robots. It's going to be great. We'll give it a shot.
40:59Thank you so much, Sundar. All right. Thanks, Nila. I appreciate it.
41:01Thank you so much, Nila. Thank you so much, Nila, thank you so much, Nila, thank you so much.