Subscribe
Share
Share
Embed
Behind the current AI boom is a developer community that both uses and builds powerful tools. Google recently released a wave of new developer-focused AI tools at Google I/O, but one person has gone further: fostering an entire online community.
Google's Logan Kilpatrick joined Sabrina Ortiz on The Deep View Conversations podcast straight from Google I/O, not only to share more insights on the new products but also to draw on his rich quilt of experiences to address the broader AI space, the future of software engineering, the developer community and more.
While Logan's official role is as a member of technical staff at Google DeepMind, he has become a well-respected voice in the developer community, accumulating over 320K followers on X and constantly engaging with users. Prior to joining Google, Logan worked for NASA, OpenAI, Apple, and other AI startups, making him a builder at his core.
Topics covered in this episode:
+ 'Vibe coding' compared to agentic engineering
+ The capabilities of Google's new Gemini 3.5 Flash
+ The internal "flywheel" at Google, where teams use AI to accelerate the development of products
+ The differences between AI Studio and Project Antigravity
+ The need for developers to regularly "reset their level of ambition”
If you want to understand the distinction between rapid prototyping and managing million-line production codebases with AI, this conversation will leave you much more knowledgeable about the principles of agentic engineering.
Subscribe to Deep View Conversations for interviews with the leaders shaping the future of AI, business, and technology.
And don't forget to sign up for The Deep View daily newsletter. We don’t just cover AI, we decode it. In a world flooded with hype, we deliver sharp, no-nonsense insights to keep you ahead of the curve and help you put AI to work every day: subscribe.thedeepview.com
Creators and Guests
What is The Deep View: Conversations?
From frontier labs and enterprise platforms to emerging startups reshaping entire industries, The Deep View: Conversations podcast interviews the brightest minds and the most influential leaders in AI.
Sabrina Ortiz: Thank you so much for joining me on the podcast today. I, as soon as I sat down, I joke that I feel like I see your presence on Twitter is so massive that I'm just so used to every day. Whenever I open up my Twitter feed, you're like all over it every day and you're a real person.
Logan Kilpatrick: I'm a real person. Unfortunately, I don't, I don't just exist on the internet. Life would be, life would be probably easier if that was the case, but.
Sabrina Ortiz: How about you tell our audience, even though they're going to see you, you're going to be super familiar with you. Tell us a little bit about what you do here at Google.
Logan Kilpatrick: Yeah, yeah. That's a great question. At DeepMind, I lead the developer platform team focused on the Gemini API, AI Studio, and another product called Kaggle. So really focused on, you know, sort of not in the tongue in cheek way, sort of bringing the GDM mission to life, which is like both building AI responsibly and then also actually bring it to our customers, to our product. So mostly focused on the developer side of the product house and then helping make sure we have, we have great Gemini models.
Sabrina Ortiz: Obviously we're going to hear at IO to talk about the announcements today and broader developer space, but I think we need to touch a bit about your background. I think it's super fascinating. NASA, OpenAI, like, did you ever rest? You've always just been super, I guess, into, again, the AI space and then building, I guess, space. So tell us a bit about that.
Logan Kilpatrick: Yeah, yeah. That's a, I've gotten lucky that I can do a bunch of things that are really fun. I get paid to do things that I enjoy. It's the best job in the world. And yeah, before Google was at OpenAI, join was a small 200-person startup, you know, nonprofit research lab, building AI models and sort of seeing, you know, the company that it is today is really cool to sort of see the evolution that they've gone through before that was, was that another startup doing machine learning and deep learning for digital pathology before that was a, was a software engineer and a machine learning engineer at Apple. And then I started my career at NASA doing like early reinforcement learning models for the NASA Viper lunar mission, which was super cool. So got to, got to work on all that. And it is like in hindsight, yeah, I've been doing AI stuff, but not super obvious. I think they, you know, disjointed sort of journey always when you're in the seat, and then it looks more coherent when you sort of step back from a 10,000 foot view 10 years later. And it definitely does.
Sabrina Ortiz: Do you think you ever had a moment in your journey where you were like, oh, actually AI is what I'm the most interested in and what I'm going to like kind of focus my attention to?
Logan Kilpatrick: Yeah, it's a good question. I think earlier in my career, it was always just like a, it was a piece of technology. And so I think the framing was like, oh, you could be sort of, uh, you know, you were, you were maybe an engineer and you were doing AI stuff, or you were an engineer and you were doing like back in engineering stuff, or you were an engineer and you were doing front engineering stuff. So it was sort of like a, it was like a engineering discipline, almost. And obviously a research discipline as well. And I was looking at it more from a, from an engineering perspective. And it always just, it seemed like the most interesting. So I feel like I just naturally gravitated towards doing that. And then it didn't feel like a conscious effort, though. It didn't feel, even, even when I joined OpenAI, I wasn't like, oh, you know, AI is clearly the thing. I should go and go and do that because I still think at that point, it hadn't, the AI moment had not sort of like encompassed the world in the way that it has now. And so it was still like sort of AI 2.0, where, um, you know, it was just a bunch of like tools behind the scenes, powering product experiences that we were all using. And, um, to see what it's become now, I think is, uh, it's very fascinating.
Sabrina Ortiz: Do you think that any of that like initial using AI more of a tool, cause it wasn't like, okay, I have to use AI. It's just like super trendy thing. And apparently I'm going to get a ton of ROI. Let me learn this thing or let me like invest in this thing. You just started generally using it as a way to, to get results in your role. Do you think that's kind of impacted what you do now? Cause I feel like that's exactly what your position kind of is.
Logan Kilpatrick: I definitely think it keeps me more grounded because I have the perspective of like the stuff that it used to work. And I was like training computer vision models, uh, for, you know, the Apple store eight years ago and sort of like the level of complexity and just like making, I think it gives me a lot more empathy actually, because you like it was so hard to train models to just do even one thing and have it work reliably well. Um, and I think now if you look at like what are developers able to do, like it's mind blowing that you could just like zero shot an app and like actually go and solve to like a reasonably high degree of production quality. So many problems that it like my lived experience, it would have taken like me and an entire team and weeks and months of engineering work and like model iterations, just to solve this like very narrow problem. And the models just like do that out of the box. And I think actually the, the thing that I also keep coming back to is like how empowering that is, because like you think that you look at the world and the world doesn't have, you know, every company is not a large company with tons of resources and lots of money and researchers and all that stuff. And so now they're like random business owner who wants to solve some problem, can just like go to these models and ask them to solve the problem. And it works, which is like really, really cool. And I don't think something that I try to reflect back on that as much as I can, because like it is, it is the point of, of what we're, why we're doing this.
Sabrina Ortiz: Yeah. I love that you brought that up. Cause I feel like I kind of got that vibe from you even before actually meeting you from just your presence online. You're, I've seen people just be like, Hey, this isn't working and like tag you. And then somehow you have enough bandwidth to answer. Basically, I feel like almost an endless amount of requests online. And I feel like is that purposely done? You prioritize that to build that sense of community and help as many people, I guess make sense of this as you can.
Logan Kilpatrick: I think I just have too much of this like DNA of being a customer service agent inside of me. I think that, I think that really is like part of it. I think the other angle of it is you, it just like helps me do my, my, you know, my, my job job is like help make sure that we're building the right product for, for all of our developer stuff inside of D-Mind. And I think the way to do that is by seeing what people's reaction to our technology is, what, what the rough edges are. And I also think it's partially an intentional strategy because like Google is a very large place. There's lots of people in Google. I think sometimes it's, you know, you can, you can hit the rough edges and I've hit this myself as a Google user sort of like it's a big place. There's many people who have responsibility. And so I think being able to just cut through that to be like user has a problem, fix the problem for the user has just a massive amount of alpha, I think for, for us as Google and for our team specifically as well. So try to do as much as possible. The challenge is I don't scale infinitely. The volume of people who have things that they want us to make better. I think is very high and I love the feedback and comments, but trying to find balance of all the other things that we have to do.
Sabrina Ortiz: No, but I think it's working in a really positive way. So I think you've developed a really, really strong and healthy developer community. And obviously some of that is just Google's doing. Google's had that. But also you really do help feed that. And even today at the keynote, I love coming to Google I/O because it'll be like the nerdiest announcement ever. And like you'll just hear like, wow, we're talking about this backstage page announced,
Logan Kilpatrick: you know, being able to build native Android apps in AI Studio. And she was like, and it's using Kotlin. And I have my, my sort of, you know, feedback to the script, which is, I think for a lot of developers who don't use Android, they don't know why. Kotlin's incredible because you can actually like natively target all of these different like devices of you, actually the same code that you use for your Android app can also be used for Android XR, Glasses application or a watch application or whatever. So it's like really powerful to use Kotlin. Clearly there was someone in the audience who knew what, why they were excited about Kotlin. There was the loudest cheer. And I was talking to the Sam from the Play Store team and he was poking fun at that as well. 40%. I think of developers at IO or Android developers. So super strong community. I'm glad we landed native Android apps for them. I think it's going to be tons of fun tonight. But like, again, Paige and I were talking like neither of us had actually ever built an Android app in AI Studio until this experience. So like just being able to like actually deliver on this mission of like letting more people build stuff in the Google ecosystem who like wouldn't have, like I was probably never going to build an Android app unless something happened. And I don't think Paige ever was. And so it's been fun to like actually see that democratization happen. Now it's time for a word from this week's sponsor, Deal. Have you ever loved your HR, IT or payroll software? Yeah, didn't think so. Me neither. Most platforms are stitched together. So when they add AI, there's not much it can do.
Sabrina Ortiz: Deal rebuilt the entire global workforce stack from the foundation up. One AI native platform for HR, IT and payroll. Built in house, not white labeled. And because it's one system, AI inside Deel can actually do real work. It manages onboarding, payroll, compliance and approvals following your policies, workflows and risk thresholds. You can hire, manage and pay anyone, anywhere without juggling tools, logins or manual handoffs. AI agents take care of the repetitive work while over 2000 local experts in 130 countries keep everything compliant from five people to 50,000 Deel scales as you grow. It just works and you'll actually like using it. We use it here at the deep view so we can confirm how good it is. See how global work runs when AI is built into the system. Book a demo at Deel.com/deepview. That's Deel.com/deepview. We thank Deel for their support of the deep view. And now back to the show. Yeah, let's talk a bit about I guess those announcements from today. Some of the ones that sit out the most to me again, update, upgrade to AI studio and also Antigravity expansion. So how about you could say it better than I can. You tell me a bit about them.
Logan Kilpatrick: Yeah, I mean, there's so much there's so much news to cover. We can spend we can spend an hour just talking about all the news.
Sabrina Ortiz: But that's why I narrowed it down to those two buckets and possibly the highlights of even those two buckets.
Logan Kilpatrick: I think I think we would be remiss not to mention three point five flash. The best model we've ever launched as Google is the most capable AI model available to developers at a price. And I think like latency that makes it really competitive and sort of puts puts it in a really I think a class of its own is how I was describing it to someone earlier today. So that's really exciting. And then Gemini Omni. So being able to bring together a bunch of our sort of native multimodality input and output capabilities, like truly having a model that can take in anything and create any output is really, really exciting. I think starting with it being actually really good at video editing. I think is the hero use case. So Nano Banana for video, I think is the way that the team is describing sort of the first iteration of the Gemini Omni model. So that's on the model side. And I say I have to talk about that because I think the model is the product in many ways. Like I think there's the developer ecosystem matters, but it is sort of scaffolding around having a great model.
Sabrina Ortiz: I like that you say that because I feel like a lot of times when it comes to like any big model announcement and again, I feel like I'm covering a new one from any from all the players like every week, multiple times a week. But I feel like they're usually geared towards like or trying to be marketed towards just regular consumers. So oftentimes they could get lost in translation like, OK, like now my chatbot is smarter and faster, but it's like, no, actually for developers or people who are actually using these models more than beyond the chatbot. It actually makes significant leaps in differences, right?
Logan Kilpatrick: It does. Yeah. And I think the most beautiful thing is for the developer ecosystem, there's some, you know, there's some great quote that I saw on on X, which was talking about sort of every time. I think it was from Gary Tan or somebody from from YC. Every time there's a new model launch, there's a whole class of these startups that weren't possible before that become possible. And I think that that is like the most exciting thing to me is like all of these new businesses and these new problems that people can solve like are now actually in scope of what can be solved. So yeah, I think it's fun. Like hopefully Flash does that for people. I think Omni does it for people. We saw it with Neto Banana. Like Neto Banana just like helped all these businesses that weren't and otherwise like I've talked to maybe like 20 virtual try on companies who are doing something in fashion or something like that. And like they just, you know, it was a really difficult business before and now they're using Neto Banana and their users love it. And I think it's the best experience ever. And I think we're going to see something very similar for Omni, which is, which is super exciting.
Sabrina Ortiz: Again, I think that's cool because I, when a user is, and I think there's just so much AI every day coming out, everyone, I guess a general consumer could be like, okay, I guess like what? Fine. This is cool. Like then it's another image model. But to your point, people are using them. That's just like the foundation. Most people are building on them to make way more specific or advanced or tailored things that are actually extremely helpful for people.
Logan Kilpatrick: Yeah. 100%. And I think to answer the other part of your question, speaking of building on top of things, I think the Antigravity team, I think has been doing a ton of work behind the scenes over the last probably like six months in order to sort of set up this IO. And I think for folks who haven't watched the live stream yet, the sort of theme of IO is it was agents, agents, agents. I think it was agents everywhere, agents powering so many new product experiences across Google and sort of agents in Google search and agents in the Gemini app with Project with Gemini Spark and so much other stuff. The highlights from an Antigravity perspective, I think is two-fold A to your point, the ecosystem. I think they've landed a bunch of new product experiences across the entire developer stack. So if you want a managed agent interface, Antigravity has it. If you want to sort of continue to use the IDE experience, they have it. If you want to use Antigravity in the CLI, they have it. If you want an SDK to sort of spin your own agents experience, they have it. If you want an API with managed agent service in the Gemini API, we have the Antigravity agent there and so many other places. So Antigravity everywhere, I think is like one theme of this. The other theme is, and it's related to that Antigravity everywhere story, is Antigravity actually powering much more than just their own product. And I think it's really interesting to see this. And I had this conversation with Sundar probably like a year and a half ago, which was just around like, just the observation, the reflection about historically Google hasn't had this like through line of all of our products. And Gemini became that through line. Literally like every Google product has Gemini baked in and sort of infused in some way, which is really exciting. And I think it's cool to see actually Antigravity sort of become another one of those through lines, like as all of these products turn the corner to become natively agentic products, you see search and you see workspace and you see the Gemini app and you see our developer platform sort of all building around the Antigravity agent harness, which is really cool. And it means these teams can move out ultimately from like an end user perspective. Google builds better products because we sort of have this foundational layer to build on top of, which is awesome. So it was fun to see that.
Sabrina Ortiz: And like I guess, a table set with people listening, Antigravity is a really cool platform that we're going to water it down a ton and you can actually leverage, but basically anybody can create an application in, you could talk to it. You could type, you could, and it could happen super quickly. I love watching like videos online on YouTube of like, I'm going to create an app. You think Antigravity in like five minutes and it's really neat. Yeah. That's, you would describe it that way too. I did an okay job.
Logan Kilpatrick: Yeah. I'm too close to the sun on this because I see, we, we spend a lot of time talking about how to, how to make sure we position these correctly to folks. I think, I think that's a great description. I think they're like tag, they have a bunch of different tag lines. I think one of them is like mission control for agents, which I really like. So like as you try to manage sort of agents doing work on your behalf, and actually it doesn't even have to before, for coding work. If you want to do knowledge work, even it's connected to a bunch of the Google ecosystem and can help you do that stuff. And we're seeing like people, you know, take meeting notes, put it in Antigravity, summarize the feedback and then like kick off, you know, agents to go and do X, Y and Z on their behalf. So, yeah, I like, I personally like mission control for, for agents. I feel like it's a great tagline.
Sabrina Ortiz: I'm curious because I, I love how I'm always fascinated about the Google developer ecosystem because when I hear something like that and maybe I'm just too protective of my own job, but I feel like my first thought would be that there would be backlash from developers or it'd be like, this is almost, anybody could, could build an app now or anybody could use this. Or it's like almost making it too accessible, you know, like what, like then what, what's the point of me? But it's the exact opposite. I feel like developers are the most excited about it. Yeah. Uh, we'd love to know from you, from like, yeah, why you think that is. Is it just because they're also able to move just that much faster too? Like I feel like there's been such an overwhelmingly positive response. At least that's my perception of it.
Logan Kilpatrick: No, no, I think you're spot on. It is interesting. I do think the perception has been super positive from the developer ecosystem, not, not, I think, you know, obviously there's been some fumbles over the last few years. So it's, it's been a, it's been a path to get to this point. But, um, I actually think Andre Carpathia had a really great framing of this, which is vibe code. And it was sort of describing the difference between vibe coding and sort of a genetic engineering, which I think is sort of what developers do. Um, vibe coding is sort of bringing the floor up so that everybody has the ability to build something. Um, agentic engineering is sort of like when you are working with like production grade code bases and like you're creating writing software, maintaining developing software for a living, it's your full-time job. That's sort of where agentic engineering comes in. And I actually think they're like distinctly different. It is the same underlying capability that's enabling them, but sort of the way the audience, the way you go after solving those problems, I think it's actually distinctly different. Um, and I, and I think actually one good example of this back to the sort of definitions of products, I think it is like AI Studio versus Antigravity. And I think Antigravity is sort of going after this, um, this sort of like agentic engineering, like you should be able to go into a million line code base with Antigravity and like go and make a bunch of intelligent updates. And they have sort of the guardrails to make sure you do that in a way that's like safe for your code, et cetera, et cetera, hasn't caused production outages. Um, and I think in the other end of the spectrum, AI Studio, I think we're trying to help you sort of go from zero to like a fully provisioned app built inside the Google ecosystem without ever actually having to see any code. Um, and it's very vibe coding. Like you wouldn't sort of take that thing and then like, you know, build your entire, you know, trillion dollar business on top of it, at least initially. Um, and I think that the, the tension I think for the ecosystem is in the sort of fluidity between those different personas. Um, and I think there's lots of like rough edges and like one of the things again that we're trying to do to sort of remove those rough edges, like very simply one example is like letting people move between products. And so you could start building something in AI Studio, realize that, um, you know, maybe this, maybe I do need to be more serious about this and like sort of bring the agentic engineering principles in order to like keep scaling this up. And one click exported into Antigravity and sort of like continue that development locally, um, which is, which is exciting. And hopefully lots of developers will do.
Sabrina Ortiz: Yeah. I feel like you're in the weeds yourself too. Which is why I enjoy, or why I was looking forward to, um, this conversation, why enjoy listening to your takes on this too? Cause I feel like you've always kept a relatively chilled demeanor when I cut, whenever you're a pro with your ask questions, like, will it take over jobs? Like, well, developers still be needed. Like where, what's the future of software engineering or you usually have a, at least my take is a generally like calm approach to that. And we'd love to know, uh, what kind of fuels that from, from your end?
Logan Kilpatrick: It's a good question. Um, I think being, to your point about like being close, um, there's, there's many layers to this. I think one is like not living in the Bay Area. Historically, I lived in the Bay Area. And I think it's, I think it's sort of easy to, there's just so much, there's like, it's the, it's the eye of the hurricane is what I describe it as. Um, and there's just so much happening and there's so much exciting stuff happening, but I think it is easy to sort of get lost in the fact that, you know, most of the world, most developers, if you sort of go other places, like maybe haven't used this technology before, maybe aren't fully bought in. And sort of, I think it's, it's helpful to have that perspective. Um, so try to maintain that perspective as much as possible and spend time with people who are building in all different types of ecosystem, especially now, because I do live in the Bay Area, uh, new, new Bay Area resident. Um, I think the other piece of it too is like, I do think Google has the, I take, there's many things about Google's mission that I take super seriously. And I think D-Mind and Google have this sort of stewardship position, um, in the ecosystem and, uh, making sure people understand the technology can like actually build on top of it or, um, are excited about the direction that we're going, our participants in the conversation. Like, I think we have such an important obligation to, uh, to make sure that that happens. And so I, I try to, I try to take that as seriously as possible because it, it feels, um, yeah, it feels important to me.
Sabrina Ortiz: Yeah. Uh, again, I feel like with the way that you interact with people and the, in the way that again, even people are reacting to these products, it, that, that importance is felt throughout the development community in a really neat way. Um, what do you think comes next? Like, where do you even go from now? Right? Like it's like, okay, so you could build an app from using your voice. And again, I love watching those clips and it's like, oh, in five minutes, we have like this super interactive, cool website kind of thing going on. Oh, what, like, is there something? Is it just quicker? Is it speed? Like, I'm like, what else? Where else do we scale from here?
Logan Kilpatrick: Yeah. That's a great question. I think about this all the time because it feels there's this double edge sort of like we've made so much progress and yet there's so much more to do. Um, and this IO feels like that, which is like, obviously tons of progress, a bunch of amazing launches, but it really is like, um, it's actually like the start of a product roadmap. It's like not the, it's not the end of this roadmap for us. Um, and as far as like people building apps, I think the exciting thing is like, and there's a spectrum of this, but, um, we want to let you actually go from like, you know, the sort of internal half a joke, half serious, but it's like prompt to profitable company. Like we want to let you like actually go the next set of, set of milestones and building something. I think if you've, if folks have like tried to build something before, actually the software bit is obviously important and has historically been the gate to make it so that you could build something if you're trying to build software products. Um, but that, that will be solved. Um, so like the ability to create software, I think like as the models continue to get better, like will be something that is like fully within, within everyone's ability, like hopefully within the next like 12 to 18 months is all the products and models continue to develop. Um, but there are many challenges to go and actually build a business that are sort of have nothing to do with the creation of software. It's like, how do you talk to users? How do you sort of make sure you're forecasting the right stuff? How do you, you know, do customer support? What's the right pricing? How do you do marketing? There's like, and other things that I'm, I'm sure I haven't even mentioned. Um, and what's most exciting to me is I think Google is so well positioned to help remove as much of that friction and burden as possible for people who are like trying to go into this next era of building. And I think about AI Studio's responsibility in this, similar to like what YouTube did for creators. I think AI Studio can do for software to like let everyone, um, have the means to not just build software, but actually like distributed to the world and monetize it and like build a business around it. Um, and I think that's, I don't, I don't think that's like fully in the zeitgeist zeitgeist yet that like that's going to be possible and that's what the world is going to look like. But I very much believe that we will very, again, very similar to like creators and YouTube and how the platform has enabled so many folks to like have a voice at the gas studio. Hopefully, um, if we do all the things right, we'll sort of enable folks to do that with software.
Sabrina Ortiz: I really like that analogy because I feel like it really highlights what we're talking about earlier. Like, why do you not feel like super threatened about jobs and stuff? It's like, yes, everybody has a camera in their pocket and anybody can, you know, set up things and posts and it's easier than ever, but you don't see everybody becoming like, you know, a YouTuber of millions of followers and having a successful business from it.
Logan Kilpatrick: But some will, which is also important. I think I love YouTube because you have this or it's a great example. And obviously YouTube is like the most successful version of this analogy, but like you have the ends of the spectrum. You know, you have the Mr. Beast and then you have the person who's like, I'm just uploading a video to YouTube so I can send it to my mom so she can really see it or something like that. You know, it's like a, like a Google drive video. And I think that was maybe more common in the original heyday or like not heyday of YouTube, but like early days of YouTube when it was, when there weren't other video platforms that you could do, like drive sharing, like things with, but yeah, it's exciting. I feel like if we can, if we can pull that off, it will like create so much opportunity for folks who want to be able to create businesses around software who again, otherwise wouldn't have it able to.
Sabrina Ortiz: So can I ask you like when you're like, when I'm imagining everybody sitting in like a round table in an exact room, like, why are we building the X, which obviously doesn't happen that way. But let's just say when you all just like sit around in your brain store, like to your point, there's AI could be applied, used, tailored to basically every software part of the pipeline, building, executing, you know, all that. But also every industry, every role, do you kind of just start tackling like the foundation and kind of build that from there? Or like, do you, yeah, like, do you have to think about all these specific things? Because there's too much there.
Logan Kilpatrick: There is too much. I mean, I think this is like the sort of, in some sense, rephrasing what you said is like the paralysis of all the opportunity, which I think is a really interesting thing that I think folks are grappling with now, which is like, and I actually, I feel this way when I work on personal projects, I'm like, I could do anything now. And I'm like, so, so I have to be, I'm like, even more thoughtful about like, what is that next personal project I should take on? Because like, I could basically build anything I want. And I think I feel that way for our team. And I think the lens that I look at this through is like, what is the thing that we can uniquely do? And I feel like there's, there's obviously many things that Google could do, but like, what should, what should we do? What do we actually have an obligation to build for our customers? Where do we have a bunch of infrastructure or a competitive advantage that sort of makes it easier for us to build these things? And I feel like a studio is like a beautiful manifestation of the story of like the things that are unique about building this type of product. Like it really feels like a product that should be built inside of Google. And we're, and we're trying to build it in our, like, obviously there's other people building products like AI Studio in some sense in the ecosystem. But I think our, our, the way that we see it is like building the version of this product that Google should be building. And that means in a lot of cases, and maybe you saw this with, with workspace and Android today, but like, how do we give you a step into these other parts of the Google ecosystem that like, we can uniquely create this connection point with that like a team outside of Google, you know, they don't sit next to the Android team and they don't sit next to the cloud team. And so we can do all these unique things. And we're going to keep going in that direction of like help Google has such a vast ecosystem. How can a studio be sort of a, a window into a bunch of the opportunity happening across the Google ecosystem?
Sabrina Ortiz: The cool thing too is when you put together or launch a really high quality or to your point, like a product that is solving an actual gap, I feel like the natural, I guess, progress of it is that there's going to be offshoot applications that you probably couldn't have predicted. Like, what I'm thinking of right now, and I'm correcting the details, but Antigravity is being now used in search, right? Just to create like those really neat like visuals where like they're really like mini apps almost like mini games or to better solve your question. I'm sure you, when you started building or Antigravity, you didn't think of like, oh, this is like the use case that we're trying to solve. But then it was like, oh, this could actually work and do that really well too.
Logan Kilpatrick: Yeah. I think it's a great, it's a great point. And I think Antigravity is sort of powering just being this agentic layer for everything. I think actually the best case, and I don't know, I don't know how ambitious folks were thinking about originally, but like the best case actually is like search becomes that customer. Cause like search has such a high bar of, you know, they have to serve billions of customers and they have to do it very fast and in an efficient way, et cetera, et cetera. So it's almost like the ultimate success case is that search goes and adopts whatever the piece of technology is cause they have such a high bar. And I feel like that, you know, is happening with Antigravity, which is, which is really cool to see.
Sabrina Ortiz: I think it must have been a podcast clip or something. I heard you on or something you said at one point.
Logan Kilpatrick: A whisper in the wind. Yeah. Somewhere, somewhere, a billboard.
Sabrina Ortiz: Something, yeah. It must have, maybe through osmosis. I got like a, you know, that you said that.
Logan Kilpatrick: It was probably osmosis. Spent a lot of my time just sending osmosis out into the ecosystem.
Sabrina Ortiz: Yeah. And, but you said something that was really neat. It was kind of, again, I'm going to butcher it, blame the osmosis transmission waves, but that you're kind of building at Google, you're building tech or AI solutions or products that you could then use to build more for Google itself. It's like a fulfilling, and I thought that was super neat because I was like, I wouldn't think about it that way. But in a way, you're almost like catering to your own needs, right? And your own team's needs to produce more. Yeah. Yeah.
Logan Kilpatrick: I think there's, there's great, um, actually the Antigravity harness story is a, is a great example of this. I think, you know, we've were announcing Antigravity 2.0 for the, to the world today, but actually like a lot of the flywheel that's been happening is like Googlers across all of the, you know, vast number of different products and business areas inside of Google using Antigravity to actually like accelerate their own product stuff. We have the same thing for AI Studio. Like, you know, it's like 50% of the, the UX, UX folks inside of Google use, use Google AI Studio to like build UX prototypes and mock-ups of the different things that they want. Um, and, and actually like Gemini is the other example of this. Like we, we as a company, you know, are so invested in Gemini because it's a, not only is it foundational in all of our products, but like it helps us build a better version of all the products and of Gemini itself. Um, and so those, those flywheels are like the most, you can find that, that's like the, you know, chef's kiss. The most natural thing that you could possibly have is, um, the product helps you in building the product, you end up building a better product, um, both for yourself and for customers.
Sabrina Ortiz: Are you ever tempted to be like, well, actually there's this like third party product that could do this thing really well too. Like, why don't we just like, you know, use that and like kind of go elsewhere. Or do you always kind of have the mindset of like, we build in house, like we could do it better. Yeah.
Logan Kilpatrick: I mean, that is, it's, there's definitely tension with that. I think there's a lot of infrastructure things inside of Google that actually like just by nature make it hard to go and use other third party products. Like the, you know, security requirements and legal requirements are very high. And, and, you know, sometimes you feel the tension of that, but in a lot of cases that actually like it does really force Google to make a better product because like, you know, workspace is an example. Like I spend all my day inside of docs and Gmail and sheets and, you know, looking at slides or whatever it is, um, what I'm not coding. And so if, if those products don't have AI infused, um, you know, it will make my job really difficult. And so like not only knowing the workspace team and sending them the feedback when I have comments and suggestions, um, but the fact that like everyone at Google is doing that, I think is, uh, is such an important part of the flywheel ultimately for our customers to make sure we build, you know, the right AI features and things for them.
Sabrina Ortiz: Well, before you wrap up, I'm going to put you on a hot seat really quick. I have like, we talked earlier.
Logan Kilpatrick: It's so cool in here right now. I don't know. The hot, hot seat doesn't exist.
Sabrina Ortiz: You're like, actually compared to how hot it is, how hot it is outside. I'll take it inside. Um, so there's, we talked before about like, there's like the AI studio, like persona who would be using it and it might just be somebody who has never, ever touched any type of, you know, uh, IDE or anything before. And it's just starting for the first time. And then on the other side, we have developers who are using Antigravity and have years of experience and have done this before. You had one, I guess, piece of advice for both completely separately. I'm assuming it'd be separate with that piece of advice. And if there is one for both, that'd be even the kudos to you. If you could get that with that, you get bonus points. But, um, yeah, if you had one piece of advice on like what they should be doing now to make the most of the moment we're in with the tools that they have right now, what would it be?
Logan Kilpatrick: Yeah, I've got one for both. I have to go out for my bonus points. Um, the, the thing, and I feel this for myself and this is a, I put it out in the universe so that through osmosis, it will come back to me as well. Um, is I think the model progress and the sort of agent progress and the product progress has just continued at such an unrelenting pace that like you need to keep resetting your level of ambition as you're like existing in this AI era and sort of what you were trying to do six months ago that like didn't really work and you tried twice and then you're like, well, maybe I'll just ask like very simple questions to this AI model. Um, those things are not possible. And I feel like we instinctively as humans sort of like overtrain ourselves to like, oh, this thing wasn't possible. So I'll just keep doing it this way. And it is like, it's, it's so difficult, but it's also so important to just like keep resetting your understanding of what's actually possible. And it applies so broadly across every person using AI these days. And I fall into this trap myself all the time. We're like, I literally have a reminder in my calendar that tells me like reset your ambition of like how much you should go and like try to ask the models to do, um, because it just changes like literally every three months, every two weeks, maybe it, you know, you can, you can ask for more that you couldn't before.
Sabrina Ortiz: I love that. It's kind of like, you don't know what you don't know, right? Like you kind of always have to reset the expectations, keep on learning, keep on growing. I love it. I love it. Okay. Well, with that, uh, thank you so much for chatting with me. I, I always been crazy and you're like the man of the hour. So I really do appreciate you taking the time and, uh, this has been a lovely, lovely conversation. Congrats on a pretty awesome day, pretty awesome launch.
Logan Kilpatrick: Thank you for coming. Uh, and happy year two at IO and hopefully we'll have this conversation again next year. Yeah. Incredible. Sounds like a plan.
Sabrina Ortiz: Awesome.