Sundar Pichai On A.I. Backlash, The Future Of Work And Google’s Next Era

Sundar Pichai, welcome back to Hard Fork. Thanks for having me. Great to be here. So the last time we had you on the show was in 2023. BARD, RIP, had just come out. And I think at the time, the perception was that Google was catching up in AI. How are you feeling about your position in the race these days? Well, that brings back memories. So it feels like eons ago, you know, those three years feel like a long time ago now. But I think it's staggering to see how much that both the technology is making progress. We've made progress as a company. And I think it's a very dynamic moment in the industry. I think our models are at the frontier in some areas. You know, and there are areas where we are behind the frontier. And so it's a combination. I think if you look at, you know, overall capabilities, including text, multimodality, voice or audio, reasoning in general, or intelligence, I think we are very capable. When it comes to agent decoding with tool use and instruction following long horizon tasks, I think we are a bit behind at this moment. But we are hard at work. And the space is so dynamic. You know, all of the leading labs have their own pre-training cycles. So you have these cadences and they may not exactly match up. I think the moment is intense enough that if you're slightly off, you know, three months ago, people are like, we are ahead and no one could catch up with us. And then, you know, the conversation flips. But that's part of the territory of being at the frontier. I think we are the only large company which is actually at that frontier, right? So one way to think about it is there are, you know, in this moment, there are a couple of startups which have made extraordinary progress. And, you know, we've been deeply working on this for a long time. I think we took a big step forward with 3.5 Flash. It does address some of the areas we've been behind. And I think, obviously, getting it out in the real world and iterating with that data coming back is going to really help us. I think coding was the area where getting access to the data flows was important. I think we maybe quite didn't have the surface, like cloud code, as an example, or what Anthropic maybe had with Cursor too. And so getting anti-gravity with 2.0, we've been using it internally at Google for a while. I shared the token usage at Google IO. I've never seen anything like it internally, right? We are doubling every week and people are really putting the models to work. And so that is helping us hill climb quite a bit. But, you know, the frontier is very dynamic, but I'm very, very optimistic and confident we'll push through there. It sounds like if there's any place where you feel maybe not quite at the very lead where you actually want to be, it is coding. Is that right? Is that where you're sort of putting the pressure? Look, I think coding ends up being very foundational in everything we do. So I think it's an important frontier to be on. There are areas in coding where we've been very good. You know, we've been very, very good at creating single-shot web frontends, everything. But in terms of this long-running, you know, task where serious developers are working on complicated code bases, I think we are making progress. It is just that there is a gap to the frontier where others are. But we are working, you know, we are well aware of it and, you know, making progress. 3.5 Flash has been out for a day. I do think it typically takes a few days to really put these models through their paces. We have seen some complaints, though, about pricing, model quality. Curious, like what you've made of the reception so far. You know, it's, I'm looking forward to being done with my interviews and so on so that I can spend more time with the teams. Yeah, wrap it up. I'm going to meet the teams right after this. I think we are definitely, you know, it'll take a day or two to settle in. I think it's a new model. And in a new area where we've made some progress, there could be some regressions, but we will be able to quickly address them through our post-training. Pretty quickly, I think. There are some artifacts and behaviors we are seeing, which I think are easy to address. So we will. I do think given it was a day after us putting out a lot of things, I think we have tightened usage limits to avoid outages. But you will see us make progress on usage limits very soon. That is rightfully a source of frustration when you encounter. I feel the same. But those are areas we will address pretty soon and make progress. It seems like one thing that some of the AI companies are succeeding at is focus. Just, you know, Anthropic and OpenAI have this sort of relentless focus on coding. And OpenAI was criticized last year for sort of spreading their bets too thin, trying to do too many things all at once. They've now sort of tightened that. Do you feel like Google is appropriately focused on coding? Or are all the other bets you're making taking away resources and time and focus from the main push? I think all of us saw some time around, you know, there was an inflection point in coding. And I think we are all responding to it. I think we all have, you know, pretty serious strikes around this area. And so I don't see it as an issue. We are a large company and we have scale. So we will be able to focus on a few multiple things at the same time. I don't see it as any fundamental issues as much as we are making progress. We're going to make progress. I think we are in a moment in time in this field where 30 to 60 days look like five years. That's all it is. Yeah. Another thing that got a lot of attention this week was the changes that you all made to the search bar and the sort of front door of Google, the biggest change in 25 years. I think a lot of people expect that at some point, the kind of normal Google sort of classic search interface goes away. The 10 blue links maybe go away and you just kind of have this AI mode as the default. But you haven't sort of done that yet. There's a lot of integration, but it's still, you still can get the 10 blue links if you want them. Do you think that goes away at any point to rip the bandaid off and just go full AI mode? You know, I think it's important to bring users along the journey as well as making sure the product is working for their expectations. So, you know, I try not to get ahead of that. I think it is very clear as we have evolved through these changes, people are responding positively. We can see it in the long-term metrics of the product in such a clear way, right? And so I think we understand that. But, you know, people want search to be fast. I do think through search, people are looking to connect with what's out there on the web. So that's important to us. It's all of that. So I think you're seeing us evolve the product. And I think you'll continue to see it be methodical. But, you know, we didn't have AI mode a year ago. But now a lot of people are experiencing it. I think we've made it more seamless to go there than before. And so it's a continuum. But I don't see sources and links will always be there as part of it. He was telling me on the ride down that he feels like he basically has not done a traditional Google search in the past year. When you hear that, are you like, cool, this is the kind of user that I want right now? Or does it send you a little chill? Because, you know, the sort of traditional search ads business is pretty good for you. Well, you know, I think we will, if anything, in the AI mode, you know, in an agentic way, these things are going to do a lot more for you than what we were able to do for users 10 years ago. I think the economic value is always a function of the total value you're giving users. I think all of us would say over time, the value we are providing users increases, there's more competition, there are more choices. So I feel comfortable between a combination of subscription and ads that the right models will continue to be there. Adam Smith's rules don't change in this new world. Let's talk about public perception. The New York Times did a poll this week, found that only about 16% of people say that AI is mostly good, about 35% say it's mostly bad. What do you make of the AI backlash that we're seeing right now? And how much leverage do you think Google has to change that perception? I mean, AI is, you know, I've always viewed it as the most profound technology humanity will ever work on. It's progressing at an extraordinary pace. You know, humans aren't evolved to process that much change. And I think people rightfully so are anxious about, you know, what is the future that this technology will bring? So to me, you know, I understand it. I think it feels to me natural with like such a profound technology shift. We've had far simpler technology shifts where there has been anxiety around those shifts. This is of a scale unlike anything we've seen before. I think we as an industry have to do a lot more to continue driving and showing the benefits that's possible with the technology. So that is something in our control. We have more work to do to make sure when we are scaling up the infrastructure, investments, et cetera, you know, what are the things we can do to make some of that work better. But I think people's concerns are a bit more fundamental around the shift than all of that. I think a natural part of this is people are anxious about their economic future in this world. And, you know, you have a lot of conversation where people are saying, you know, jobs are going to radically change. Some of it will go away, et cetera. I happen to think, you know, the outlook is better than some of those dire predictions. But, you know, as a society, if you're hearing, it's natural. You know, I would be surprised if people weren't more anxious about it. I do think it's important because the change is happening so fast. You need people to, in democracies, you need citizens to be engaged, be aware that this is happening and, you know, and make their preferences known. That's what causes action in society. So I think there's something healthy about this dialogue, too, which is happening. And given the pace at which the technology is moving forward, it seems right to me, both the concerns and the fact that we need to take it seriously. You're giving the commencement speech at Stanford next month. I'm sure you've noticed or heard that a bunch of commencement speakers have been booed recently by college students who are worried about AI. What are you planning to tell the graduates about AI? And do you have your, like, boo strategy in place? Anytime we have driven technology progress, I think it helps drive progress in the world. And in some ways, these graduates are actually both going to be a big part of driving that progress and also dealing with the impact of that technology. So I think we have to be very mindful of that. And, you know, I've always been extraordinarily optimistic about the next generation. I think we all always have this view in the world, you know, we are anxious and we worry about the next generation. But I think the next generation rises to the challenge and, you know, builds a better world. And so I view it as no different from those moments. And, you know, my goal would be to share, you know, my experiences and share that with them. And that's what I'm looking to do. You can just pretend they're saying Google. It's close enough, you know. I'd be curious to hear a little bit more about your case that, like, that you think that sure, jobs may change a lot, but, you know, you entry-level graduate, the economic future is still bright for you. Like, what is that case in your mind? You know, at a basic level, I do think we are, you know, there is a new level of capability all of us are going to have to be able to do things. And, you know, I wasn't there, like, when spreadsheets rolled out to people. Like, you know, I wouldn't know how you did financial analysis before that. Like, you know, how did people do it? I'll say it. I didn't do it. I had no idea how to do it. But, you know, spreadsheets change that. And so there's an aspect of this. I think it just is going to change the starting point for many, many people. Just even coding. I think if you fast-forward the progress we are seeing there, so many more people are going to be able to code in the world, right? And, you know, I've heard you two might be examples of that and, like, you know, in that journey. But I think you're just at the leading edge of what is going to happen more and more. So I think those are the new serendipitous ways this will all work out that we underestimate. I think people are going to be more productive. They will have more time for leisure. All of that will simultaneously be true. There are so many areas where today people's work, you know, involves a lot, you know, doctors have, you know, high burnout rates. It's because they train and, you know, their calling is to spend time with patients, taking care of patients. But, you know, most doctors will tell you if you actually watch their time, the percentage of time they spend with patients is less, right? So I think AI will actually help them do more of that, right? And I think those are examples of it. The radiologist analogy has been fascinating. It's been now a decade running. I look at myself and I say, well, I've gotten a lot more scans in my life than my dad ever did. And each of the scans have like 10x the amount of information than his scans had because they were constrained by a printing film versus us being digital. And I think that number is going to be 10x in 10 years. So where is that projection going into, you know, you are actually going to need AI to keep up, right, with that demand coming. So I think it's non-linear how the impact of all this will be. This is not, I don't want to be, I don't want in any way to minimize every technology shift brings disruption with it. And, you know, I think, you know, there will be disruption and we as a society need to be super serious about it and engage. And so some of the conversations I think are rightful in like, you know, thinking through that. But I do think there are many positive dimensions to it, which are maybe not being talked about. And I think, and also there is overly deterministic, dire scenarios, which I quite don't agree with as part of it. Let's talk about agents. Because I feel like agents actually tie sort of into this question of, well, what is going to make us more productive in the future and how will it change our job? Later this summer, you're releasing Spark, which seems sort of intended to be an agent for the regular person. And I'm so curious to know, like, could you walk us through something this agent is doing for you personally? I've used it a lot more in my professional context, more to the context because it was mainly available in my Corp account, right, as part of it. In that context, I've definitely used it as, it's super easy to use it to prepare for any meeting. I wish I had brought the prompt slash the output for, like, I just, as a test case, used it for hard fork. Honestly, if you emailed us, we would flash it on the screen. Yeah, well, yeah. It had some things about the two of you, so I don't think I can project it. No, that's what we want. We want that. We want to know how Jablai dragged us. I'm not sure I love that too. No, I'm just kidding. Partially kidding. Yeah, you should see Casey's browser history. Partially kidding. But I have had it in my personal account more recently. So again, here's a simple task I did. I just asked it to just look ahead at my meetings and color code it in categories so that I can keep sense of how I'm spending my time. I think, you know, it's extraordinary to watch it. It came back with, like, you know, suggestions of two color coding schemes, and I just had to choose one. And it's actually like sci-fi. It's just like, you know, changes the color in the calendar. Personal meetings, health-related meetings, you know, time I'm spending at work, etc. That's an example of a personal query I just did just to see what's happening, right? And but with agents, I think you have to give people a sense of, I think about this as what allowed us to get someone to sit in the backseat of a self-driving car, right? We did it in steps. And I think there's an element of that where with agents, if something unexpected happens, I think, you know, people back off from this. And, you know, so part of it is earning their trust and so giving them a sense of control transparency. But more importantly, from a security standpoint, these systems, you know, can be hacked. And so we want to make sure we are not ahead of the frontier in a wrong way. Speaking of meetings and your calendar, we hear that you're headed to the White House for some kind of AI executive order signing. What should the government be doing right now to regulate AI? Do you like this idea of a kind of pre-release strategy where the government gets to sort of see models before they're released and sign off on them? Is that a good idea? Is that potentially dangerous if it gives them the ability to censor or jawbone companies into releasing different kinds of models? What's your take on that? Look, we'll have to wait and see the details of the full executive order, but they've engaged with the industry in a very robust way. And I think the approach really balances, you know, innovation and oversight. We'll have to wait for the details to come out. But, you know, there are a few areas which are coming up. You know, we will need more cross-industry, cross-industry government coordination. So it makes sense to me. Cyber is a great example of that. I think we all have to work together. It makes full sense to me if you found an exploit, which could impact a governmental agency, you know, the government needs to be prepared for it. So there is validity, but of course, doing it in a moment with this important technology, where it's important as a country to be at the frontier too, not doing it in a way where you're overly slowing things down. And maybe that balance has to shift as we reach more advanced levels of technology. But I think to me, this seems like a prudent approach. In general, you know, part of what we are doing with, you know, building Synth ID, open sourcing it, and then sharing it with others, and more importantly, building a consortium together, I think is an example of in a different area. These things only work together if you can come together as an industry. So, you know, I'm glad they're approaching it in that way. Another sort of safety-related question. All of the big labs are racing toward what you call recursive self-improvement. So building AI systems that improve themselves rapidly. Do you think that can be done safely? And do you feel like you have a line of sight to it right now? You know, these models are getting better at, you know, coding and agentic workflows. And so at some point, you know, we did, you know, we did, you can see in anti-gravity today, you know, in over 12 hours, it can build a simple OS from scratch, right? And, you know, that is genuinely those are multiple thousands of hours for somebody to do, right? So you are seeing some of that in work today. We are all in our products in some version or the other have agents and sub-agents and the orchestration of those agents building things together. It's a continuum, right? And I think, I think we are all definitely making some progress, but in the way people describe RSI, I don't think we are there yet. And, you know, and I, you know, I, that would represent the next level of acceleration. And I think, you know, and I think would have a lot of implications, but we aren't quite there yet. Is there like a plan for, like, I mean, great news, like we just hit RSI, like, is there a sort of, do we, do we break glass or what happens then? It's a great example of like, look, I think all responsible labs, I think, I think if you're approaching moments like that, you would be, you know, consulting with, you know, it shouldn't be an internal conversation at that point. I think it has to be a much broader conversation, conversation than that. And I think we all have to avoid race conditions at those stages of AGI. Right now, all the labs are racing to get more compute. There seems to be bottomless demand for compute. They're hoarding it wherever they can, striking deals, you know, building their own data centers. Google is still selling access to TPUs to rivals and other companies in the race. Why? Why aren't you just keeping that for yourselves and your own models? I think each is not a constraint on the other, right? So as long as we can make enough chips, it's not a constraint. The right way to think about it is we have GDM and our first party services. If you can think about that as a company business cash flows, you're planning for that. And then you have Google cloud, which is a business and which has revenue cash flows, and you're doing long-term plan and you're planning for that. So if you didn't have cloud and we weren't providing, we wouldn't be planning those chips anyway, right? And so that's at the simplest level, obviously, you know, it's a bit more complex than that, but there are a lot of advantages of providing TPUs to others. The fact that researchers at Anthropic are using TPUs is what will allow us to make, in addition to us, allows us to make the best hardware in terms of next generation. And by the way, we use NVIDIA chips too, and the next generation chips are incredible. And so we use that and we work too. When you're running platforms, I mean, you have a platform side of the business and, you know, I've always have worked on many platforms in my life and be it Chrome or Android or Google cloud, you know, why would you ever open source something or why do you provide this technology? I mean, all that makes sense on its own merits. I do think there are advantages, like I mentioned, it allows us to stay at the frontier, you know, economies of scale helps in various ways. And so it makes a lot of sense that way. The last time we had you on, we asked you about AGI and your feelings about the term. And at the time you responded that it didn't really matter whether you've reached AGI or not, because the systems are going to be very, very capable and Google strategy should be the same. I noticed that you did not say AGI in your keynote, Demis did, but you did not. What's your relationship with the term AGI today and sort of the idea that all of this progress is building towards something singular and world changing? Oh, we are, you know, there's inevitable progress towards AGI that's happening. I've long understood it and, you know, otherwise I wouldn't have pivoted the company 10 years ago to like put that technology at the heart and center of the company. All I meant by that statement was, even in this scenario, AGI is going to take 10 years. The technology, which three years out will be so much more powerful than what we have today that I don't want people to think because AGI is 10 years out, you don't need to act or prepare. That is all my statement means in those contexts. Are you AGI-pilled? Well, I absolutely, you know, I'm sure that the technology is making foundational progress towards AGI. I'm less able to predict with certainty, whether it's in the three to five year time frame or five to 10 year time frame. The rate of progress over the last one to two years has made me feel it's on the closer side than not. And, you know, I just don't, you know, in your role running one of the largest companies in the world, which has a responsibility to society, the language I choose to use around it might be different than other people. But I think, you know, as a company in terms of, you know, 10 years ago, the IEO stage, I announced TPUs and AI-first data centers. You know, yes, clearly understood where this technology is headed. Maybe as a last question, one of the more memorable phrases, I think, from the keynote this year did come from Demis when he said that we're in the foothills of the singularity. Can you tell us, like, concretely what that means from Google's perspective? And should people be excited about that? Or afraid? Or both? Look, I've had many conversations with Demis, obviously, on this topic. And I think in this context, I think he's defining singularity as the advent of AGI. I think he's talking in that context. And, you know, I think if you think, regardless, if I remember, I think he had kind of articulated by 2030 or so. I think if you believe that, it makes sense to you. That's what you're articulating. And, you know, it's as simple as that. I think for him, that's how we define singularity. And I think Demis, myself, many others, we all feel it's important to, if that's what you believe, it's important to articulate that because we are all at the frontier building this technology. And, you know, hopefully people are listening. And I think it's important to, as a society, we are internalizing that and getting ready for it. Sundar Pichai, thanks so much for coming. Thank you, sir. Thanks. Great to talk to you. Appreciate it. Take care.