← Generative Now | AI Builders on Creating the Future
Loading summary
Transcript93 lines
- [00:00]
A
Foreign.
- [00:06]
B
And welcome to Generative Now. I am Michael Magnano. I am a partner at Lightspeed. And today we're back with another conversation, this time from the Generative NYC stage where I sat down with Logan Kilpatrick at Google's New York City offices. Logan leads product for the Google AI Studio where he and his team built the Google Gemini API into one of the best platforms in the world for developers building with AI. Before Google, Logan worked at OpenAI and he has had a front row seat to the meteoric rise of AI over.
- [00:38]
A
The past few years.
- [00:40]
B
So if you weren't able to attend Generative nyc, you're in luck because we.
- [00:43]
C
Have the full interview here.
- [00:45]
A
Enjoy.
- [00:48]
C
Without further ado, please help me in welcoming Logan Kilpatrick, Senior Product Manager and lead product for Google AI Studio. Come on up, Logan. What's up, Logan?
- [01:03]
A
Hi. That was honestly the best intro of just like an event that I think I've been to in an entire year.
- [01:09]
C
So thank you. I'm very humbled.
- [01:10]
A
Such a great job.
- [01:11]
C
Thanks. Well, thank you for doing this with me. I've been really, really looking forward to this. So like I said, you have had a front row seat to kind of everything AI over the past few years. So we have to ask you about your journey. How did you get here? Maybe talk to us a little bit. What came before this? I think you were at NASA at one point. You're at OpenAI. Give us the Logan story.
- [01:33]
A
Yeah, well, one, thank you everyone for being here. Well, I'll tell my story very, very quickly. We can talk about the exciting stuff, which is what y' all are doing, which is building with AI. I joined Google back in April to work with a bunch of amazing folks who are, who are here in the room to build AI Studio, build the Gemini API before that. Led developer relations at OpenAI joined at the end of 2022 and it was a small startup, much less conviction at that point that it was going to turn into what it did. I'll give the very quick story, which is I had a job offer at IBM at the time, so this will hopefully humble the story a little bit. And I genuinely at the time did not know if I should take the IBM offer or the OpenAI offer. The IBM offer was cool to do something that I was excited about, but it was just much less clear at that point. I think unless you were in some of the circles that everything was about to explode. Before that was at a startup doing machine learning and deep learning for digital pathology called pathai. And before that was a machine learning engineer at Apple, so started my career as a technical IC and have become a product manager over time. Awesome.
- [02:39]
C
What's it been like being at Google after OpenAI? I'm guessing two very different companies.
- [02:45]
A
Yeah. The interesting experience for me personally was I joined OpenAI as a startup and it truly felt like a startup and there was 200 people and I think it became not a startup over the course of just a year and I think coming back to Google has actually felt like a startup. The Gemini stuff happening inside of Google, despite it being a massive company with huge scale, really does feel like a startup. If you have agency and you are excited to do something, you can actually go and build that thing or ship that feature for developers or for customers. So it's felt wonderful from that perspective and I think we have a ton of work that we have to do still, but generally trending in the positive direction.
- [03:28]
C
Maybe before we get into some of that work and what you're trying to accomplish, help us understand what AI looks like inside of Google. Obviously there's gemini, there's the AI studio, there's DeepMind. Help us better decipher what we know from the outside.
- [03:44]
A
Yeah. And there's a bunch of folks here from Google DeepMind. So hopefully you'll find the people at Google DeepMind, I don't want to call them out, but so Google Deamine does all the hard work of building the generative AI models that power all the internal features at Google as well as a bunch of the developer APIs. And then there's a whole bunch of different teams at Google who are sort of commercializing building internal products. And that's everyone from ads to YouTube to search, building AI into the actual products. And then the Google Cloud team takes a lot of the models and puts it in the hands of builders. So if you're sort of an enterprise customer, you might be using something called Vertex AI to get access to Gemini models. If you're a sort of long tail developer just getting started, building startup founder who just wants the fastest possible solution to build the Gemini, you might use Google AI Studio, the Gemini developer API. I think that sort of symbiosis between both building first party products with the models and developing models and putting them into the hands of external developers is a great competitive advantage for us because we feel the problems that builders feel. It's not like Kat, who's here somewhere and I work on AI Studio and AI Studio is actually just like a first party consumer product sitting on top of the models and we Live and breathe, the rate limit problems, the quota problems, the model hallucination problems, the same as anyone does. And I think it's incredibly helpful for us from that perspective.
- [05:15]
C
Yeah, it's got to be really interesting doing this at Google where like you said, you have this advantage where you can distribute this stuff to millions of developers with the flip of a switch. At the same time there's probably also disadvantages. I mean, being such a big company like Google, you can't just throw some random model out into the ether without, like you said, really making sure that hallucination is solved and other maybe trust and safety issues are solved. How do you balance some of those challenges with some of the benefits?
- [05:44]
A
Yeah, I think the DeepMind team does a good job of sort of having the infrastructure to, to know what is actually the bar to get this model out the door. So some of that is abstracted away from us. By the time the model gets to us, it's like, okay, we actually have a reasonable amount of conviction that we want to put this thing out into the world. There is a long tail of other stuff. Like cloud has its own set of more customer specific evals because the model might be safe at the core level and might be useful at the core level, but on the very specific things that we know our customers care about, it might not be the best model. So there is that extra level of check as well at the sort of Google Cloud level.
- [06:20]
C
Yeah. And give us a sense for AI Studio, what does that encompass in terms of products which you're leading? What are the different products that roll up into that?
- [06:30]
A
Yeah, so Google AI Studio is the conduit for developers to get into the sort of Gemini API. So success for AI Studio looks like you come in, you try the model, you realize, hey, the Gemini models are actually pretty good. They have, you know, long context, native, multimodal, whatever else you're excited about. And ultimately I want to build something with them. So it's not focused on like being a true consumer product, it's really focused on get you to that wow moment, building with AI and then ultimately click, get code, get a Gemini API key and like go and go and build the next company, the next startup that's like actually going to provide value for end users.
- [07:03]
C
Got it, got it. And what does a team building for that look like? I mean product managers, designers, engineers, like how do you even like think about and execute on some of those problems?
- [07:15]
A
Yeah, that's another good question. I think this is one of the things that is a benefit of being Inside of Google is there is someone inside of Google and our amazing venture folks who help put this event together and Jason and Alex and a bunch of other people is a good example of leveraging the scale of Google to go and do a lot of these things. Our team doesn't have a model quality specific function. We're fortunate enough that there's other teams in Google that do a lot of that stuff. So it really is. I don't know if anyone, I'm sure someone inside of Google has said externally that Gemini really is this massive cross functional Google process to make happen. But it's true because of how much work happens to get models out the door. It's not just our team and we get to stand in the spotlight in some cases because we have the externalization surface. But there's a ton of teams doing work to make all this happen.
- [08:07]
C
I've heard a little bit about there are these three different main components of the Gemini ecosystem. Gemini Vertex AI. And is it Gemma or Gemma?
- [08:17]
A
Gemma.
- [08:19]
C
What are those three things and how do we differentiate them?
- [08:22]
A
Yeah, the two main model classes. The Gemini models is our set of proprietary frontier models and then there's also a set of open models called Gemma. And the advantage for Gemma is you can actually own the weights of the model. You can go and put them onto a server somewhere and then the apocalypse can happen and you'll still be hanging out with your Gemma model weights doing AI stuff and you don't need to worry about anyone else hosting the models for you. There's generally this lag that happens between the frontier capabilities and what ends up in the Gemma models. It is based on the same research, a lot of the same core technology, but it is not like it doesn't have long context. For example, the Gemma models aren't natively multimodal. They're really good in their class of text in, text out, open source LLMs. But it's not super competitive yet with the main frontier capabilities.
- [09:14]
C
Got it. One thing I was thinking about is it seems like there's a lot of talk about where the value in AI is going to accrue. Obviously there's a lot of talk about the foundation model layer and there are several big companies like Google, like OpenAI, which are obviously going to be big winners at that layer. Then there's also a lot of talk right now about how the apps layer is wide open. I have to imagine you, your team, where you're working in the developer ecosystem. You must see amazing app layer stuff being built. I'm curious where you all feel like the opportunity is at the app layer.
- [09:51]
A
Yeah, that's a great question. I think just to echo what you said, I do think there's an immense amount of value to be created at the app layer. If you just look at one simple proxy for this. If you look at the cost of LLMs over time, basically going down to zero, the consumer willingness to pay, if you graph it on the same exact chart, is not going down to zero. People are wowed by AI. Obviously there's millions and millions of people who are willing to pay $20 a month for insert whatever AI subscription you're interested in. People are really excited and it's creating all of this value. And the cool thing is founders and people building stuff actually are the ones who get to accrue that value. Their companies are the ones who get to accrue that value. And the cost continuing to go down to zero is great for people who are building stuff. That said, I think there's a lot of things that people are building that are not long term differentiated, are not actually different.
- [10:49]
C
You're talking about just rappers around models.
- [10:51]
A
Yeah. And I think in a lot of cases starting as a rapper makes perfect sense. But if you don't really aggressively figure out a way to get out of that position quickly, if there's not a clear taking off point as part of the baked in strategy, it's going to be tough because the model crank is going to keep turning. There's a bunch of obvious things that all the main AI consumer applications don't do today that they will do over the course of the next two years. Those teams are fighting really hard to build great products as well, just like you all are. I think the thing that has gotten me most excited recently at the application layer is all these differentiated actual ways of interacting with AI. And I think NotebookLM is the most recent reminder to me, and this came from Google. But there's a lot of. I don't think there's actually that much interesting differentiated stuff that the Notebook LM team did, other than having conviction that there's a different way that you could be interacting with AI content. And I think there's a whole lot of other things like that that just take time for people to experiment with. But I would push in every conversation I have with people building out the application layer, push on. It's not chat that is going to create all the value. It's maybe not even like voice. I think there's a lot of people who are like voice is chat.
- [12:08]
C
It's Voice. But maybe it's not even voice.
- [12:09]
A
Yeah, maybe it's not even voice. I think keep pushing on what that interaction paradigm might be. And there's a lot of value to be accrued in having a differentiated perspective.
- [12:19]
C
Do you have either through what you're seeing or just a personal intuition around the areas or the categories or sectors within the app layer that you're most excited by, whether it's consumer or enterprise or health care, whatever, Whatever it might be. Like where in the app layer, Are you really excited about the future?
- [12:37]
A
Yeah. It still feels like consumer is early. Like, I think if you grab a random person off the street and like not San Francisco, they don't know or care about AI. I think the New York crowd cares about AI. I'm glad you all are here and building stuff, but most people don't care because it is hard. Like the use cases today where the most value are being created are the enterprise use cases. Like, there's just so much. If I'm 50% better at coding, many hundreds of thousands of potential dollars for some large company to be accrued by me being 50% better at coding. I think the changing point for this is creating a bunch of value where it's not putting the technology first. This has been my big qualm with people building agent stuff, which is everyone thinks we have to call our product. If it's consumer agent stuff, we have to call it an agent platform because that's what's getting the people at lightspeed excited about it or whatever it is. But I think really the value is abstract away all the agent stuff. The idea of agents is great and AI should do stuff for people. I think that's pretty obvious and people can agree to that. But don't force the long tail of consumers to have to care about the technology, because they don't. That's the reality. They're not interested in agents, they're interested in their life being better, easier.
- [13:54]
C
Yeah. They didn't care about the gps. They cared about Uber.
- [13:57]
A
Right, Exactly.
- [13:58]
C
They didn't care about the camera, they cared about Instagram. So it's about the product.
- [14:03]
A
It is.
- [14:03]
C
Yeah. That's super interesting. Speaking of where value will accrue and we talked about some of these sort of big company incumbent advantages. One of the things I've certainly read about in the press a bunch over the past week or so is this notion of sort of reaching the upper limits of skills scaling. You're around a lot of this stuff. I can tell by your tweets you're a Big thinker when it comes to AI. Where's your head at on the scaling limit question? Are we starting to reach the upper limits of what these transformers can do and then maybe we can get into if so, why are we reaching these upper limits?
- [14:41]
A
Yeah, I think it's an interesting question. I think there's definitely people probably in this room who are better, who are closer to the metal. On being able to answer this, my perspective is you have to sort of earn the scaling laws. The scaling laws is not like a law of nature. It doesn't just happen. You have to literally earn it. And that means innovate and do a bunch of stuff that enables the technology. Same thing with Moore's Law. Moore's law doesn't just happen because someone wrote it down on a piece of paper. It happens because there's thousands of engineers that insert whatever hardware company sort of making that the reality. And I think people are very in the moment of capturing like you know, is scaling continuing to work like today? Maybe. And I don't this not saying this is the case but today maybe scaling is not working. Tomorrow there's innovation that sort of enables it to happen and then all of a sudden the scaling law continues back up again.
- [15:39]
C
Right. So like we're going to hit some upper limit or we are hitting maybe some upper limit of data or the capability of the current compute clusters. Like something is capping us. Like what?
- [15:50]
A
I don't think so. I think we're still sort of, I think we're still sort of pushing up is my perspective. But I think even if, again, even if whatever the technique of today isn't working like tomorrow it will, there's going.
- [16:06]
C
To be algorithmic scaling like we're going to find new ways to or the.
- [16:10]
A
Next 100k H100 cluster comes online, whatever it is, you know, meta throws another 100K H1 hundreds at it and then things start to work again.
- [16:19]
C
Yeah, I'm glad you brought that up. That is a very, very interesting moat for big companies like Google Alphabet, Meta. How important do you think size of compute cluster is and how long does that advantage last?
- [16:33]
A
I think it's tough. Training LLMs is expensive and it's tough. I think there are a lot of companies that have proved to, to be solving very domain specific problems like training their own models. And I was talking to the founder of a company called Cartwheel and the founder was actually at OpenAI before and they're training domain specific vision or motion models to be able to do animation and stuff like that. And he was showing some of the relative to the foundation model, doing the capability and the domain specific model they trained because they're only trying to solve that problem. It is hundreds of thousands of times more compute efficient to actually do inference.
- [17:16]
C
For those tasks by being specific.
- [17:18]
A
And it doesn't need to have the weights of writing poetry and doing all this other random long tail use cases. It literally just needs to be good at this motion task. I think there's a future in that. And you don't need to have crazy amounts of money or crazy amounts of compute. You can tackle those problems as long as you're not trying to build the generalist AI agent. I think that problem will sort of go to whoever's willing to throw the most compute data, money, algorithmic improvements at the problem.
- [17:47]
C
So do you think is Cartwheel, by the way, super cool product and model they actually demoed at one of these events?
- [17:54]
A
Oh, no way. Yeah.
- [17:56]
C
Do you think that that is an approach we will see more and more teams go after like highly specialized small models? Is that an opportunity to not sort of get run over by these bigger models?
- [18:09]
A
That's my instinct and I think the proof will be in the pudding because I'm not sure how many of those companies have actually gotten to legitimate scale to know that this is the case. But I think the same is true for just generally people solving any problem, which is if you really focus on whatever the vertical is, whatever the niche is, you're not competing against Google and the long tail of these big companies. There's very little competition in those very domain specifications ecosystems. And I think like Cartwheel is a good example of that.
- [18:37]
C
Yeah, what about talent? I mean one of the things we, you know, we hear and we talk a lot about when we talk about AI is talent. Right? These amazing people, researchers, builders, you know, finding this next algorithmic breakthrough. And you know, we're starting to see some signs that like talent really may be that important and that much of a moat. Obviously there was like a pretty public and notable deal between Google and character AI recently with Noam Shazir and some of the founding team coming back over here. How do you think about sort of people and talent as a moat? And again, like how long is that going to last?
- [19:13]
A
Yeah, that's another. I mean I think it matters a ton, like at the end of the day, like if the talent is the difference between you winning and not winning, I think the same. It's true at Google, it's true for every startup that's building. I think more so than I think anything else. My personal belief is you can win with having talent even if you have a worse idea, a worse, et cetera, et cetera. And I'm actually curious for you, backing companies at the early stages, is it a similar.
- [19:45]
C
Yeah, I mean, we think talent is absolutely critical. And as a newish VC and former product person CEO, when I first came into the the job, I was very, very focused on ideas and products, and still am, of course, but I think as I'm getting more experience and learning from some amazing people, it's really about the talent. It's about the people for sure.
- [20:08]
A
And it's interesting to see the talent flow, I think, which I think the challenge for a lot of things that aren't AI right now, as I'm sure there's like, I'm assuming there's at least one founder in here who's building something that's not in AI. It's hard because the center of gravity continues to shift between all these different things. And if you don't have the ability to pull in the best, it's tough.
- [20:30]
C
Yeah, let's get into everyone's favorite topic. What is AGI?
- [20:38]
A
Yeah, I think I still align with the version of AGI where it's the models are able to do most of the things that are economically productive that humans are able to do. I don't know if anyone listened to the full five hours of the Dario Lex Friedman podcast. You did. You have too much time on your hands. You should be. You should.
- [21:01]
C
It's really good though.
- [21:01]
A
It is good. It's good entertainment. But yeah, I think the 2026, 2027 timeline to make that happen seems. Seems interesting.
- [21:11]
C
Do you buy that? Do you think we're headed there?
- [21:13]
A
I think it's tough. I think the models will continue to go up and do so much more. I think the real question is it's like a. To do most of the things that humans are able to do that are economically productive. It's not like a moving bytes around problem. It's a moving atoms around problem, which is actually just a lot harder to scale. I'm fully bought into the idea that there could be digital versions of maybe it's AGI in the sense that it can do all the things that are digitally productive that humans are able to do. But the long tail of making robots that actually do most of the things humans are able to do in the physical world I think is much longer off than 2026 or 2027.
- [21:59]
C
Just to clarify what you're saying. So it sounds like what you're saying is, yes, we need some breakthrough on the model side, but then we need to have a physical embodiment of it and get into robots and like you said, humanoid robots or whatever, full self driving, Is that what you mean? Like it has to make its way out into the physical world.
- [22:16]
A
It has to make its way out to the physical world. And you could have AGI today, assuming you had AGI today and the current state of robotics was the same. We're still five years away from any large scale manufacturing run of humanoid robots that are actually going to be able to scale out and have any meaningful impact on the world. So it's like we can definitely scale up the digital intelligence, but I think actually manifesting intelligence out into the world in a physical form is going to be much, much harder to make happen.
- [22:49]
C
How does Google think about this? Obviously there are startups out there that they're very publicly stating that their mission is AGI or super intelligence. Is that something that Google talks about and thinks about internally? It's like, hey, we're doing Gemini because we want to achieve artificial superhuman intelligence. Or do you not even really speak in those terms inside of Google?
- [23:13]
A
I've heard Demis not to speak on Demis behalf. I've heard Demis say a bunch of times he wants to make AGI. I think for Google, Google's mission as a company is organize the world's information and make it universally accessible. So I think it's like less, it is much less at the Google level, AGI focus. I think DeepMind and I'm sure Demis probably want to make AGI as sort of the focal point. He's been pushing on that for 10 years now or something like that.
- [23:39]
C
Yeah. Are you looking forward to AGI? Are you an accelerationist or a doomer or somewhere in between?
- [23:47]
A
I think I'm looking forward to it in the sense that I think there's a lot of tough problems in life and I'm excited for AI to help solve a bunch of those tough problems. I think the real practical downside is. And again, back to this narrative of you can grab random people on the streets of any city besides San Francisco and ask them about AI. And if the tool actually does deliver on this promise of everything that everyone says, it's going to the time horizon to educate the world about whatever this new tool is or technology is just very long. And you can Even look at ChatGPT as the most successful version of this. There are billions of people on Earth who have never heard of ChatGPT have no idea how it works, all that stuff. If the technology is actually that powerful, I think there'll be a lot of downside for the people who aren't in the know that this thing exists or have access to it. And I think that has the chance for accelerate the discrepancy in the delta between people who have access to things and who don't, which I think is a very tangible downside. And I don't think there's been enough push in my personal opinion. I think at a government or world level of actually helping get those types of resources and education into the hands of people who are going to need it.
- [25:07]
C
Maybe tying this back to app layer developer ecosystem where it feels like a lot of value is being created and it could be an area that experiences disruption sooner than in other areas as a result of AGI is coding. Right. It's like these things are really good at coding and they're getting better very, very quickly. How do you think about the long term impact of AI on software? What does software become? Is it all dynamic? Is the God model just like spinning up products for us in real time? How do you intercept that future with maybe what you're doing at the developer ecosystem level?
- [25:47]
A
Yeah, it's a good question. It's something that I think about a lot as a weakly active cursor user who I think are closest to this. I think the cursor example is relevant in that I think the path to get to that point is software engineers increasingly having this really powerful tool in their hands, which I think is a different version of the world than AI is going to replace software engineers. I think it really is. The AI augmented software engineer is going to be able to do an incredible amount of stuff in the future, but it is not going to be able to. The fundamental nature of having to prompt models very specifically to do what you want is not going to change. And I think the example of hey AI model, go and spin up this vertical SaaS company for me to do is not actually going to be possible to happen. I think in the way that we think it's going to happen today. My assumption is it's going to look very different of how the interaction with the model or the way that the model is going to go about solving those problems because it's just going to need more guardrails than the current versions. You can't just let the model run wild and do that thing. It's going to burn a whole bunch of compute and then end up with not the Vertical AI SaaS thing that you really wanted it to do in the beginning, which will be. Yeah, it'll be interesting to see. I think there's a bunch of very specific problems that need to be solved to get to the point where the models are able to just generate entire stacks of software themselves.
- [27:21]
C
I do think it's going to be really interesting for enabling people to create software that maybe don't, don't know how to code or maybe don't know how to code very well. It feels like you could see this almost cabrian explosion of software and all these amazing new applications that you really couldn't have before. I think that's something that could be very, very interesting and especially for maybe a developer platform.
- [27:42]
A
Yeah, I agree. I think people have been pushing on low code for a very long time and it feels like this, for what it's worth, the low code people had their heart and head in the right place. It was just like needed a couple more cycles of innovation to actually make it happen. It'll be interesting to see how much that actually comes to fruition.
- [28:05]
C
Where should we expect to see Gemini over the next few years? Throughout the Google ecosystem, obviously we're starting to see it more and more in search. I anticipate home could be next. Waymo, what can you tell us about the future of where we see this thing?
- [28:22]
A
Yeah, that's a good question. I think the wonderful thing for Google is all of these products benefit from AI. So my expectation is it's going to be everywhere. I think on very specific, different things. I think it would be really interesting to see how does it make its way into Waymo. I think there's some interesting research papers they put out about Gemini and how it relates to Wayne Mo stuff. So I don't know, I don't want to represent that work so you can go look it up. But yeah, the domain specific models is also the other angle of this is how many of those problems actually benefit from those teams having their own version of Gemini. And Google has done a bunch of work with that as well. With Med, Gemini, with LearnLM, a bunch of things based on the core Gemini model, but solving the very vertical problem. My guess is actually for a lot of the products for them to be like, if it's really a successful use case, they'll end up doing a vertical and not just use the base model with prompting.
- [29:23]
C
Super interesting. I can ask you questions all night, but I want to make sure we get a couple questions from the audience. Please say your name, what you're working on and then ask Logan your question.
- [29:33]
D
Hey, my name is, I'm a PM at Cash App. Logan. Michael, thank you for this panel. This is fantastic.
- [29:39]
A
Question.
- [29:40]
D
So there's probably a ton of founders here that are early stage thinking about using Gemini, OpenAI llama, et cetera, et cetera. Logan, question for you. How does Gemini differentiate? Why do I use Gemini versus the others?
- [29:53]
A
Yeah, this is a great question. It's something that I spend a bunch of my time thinking about because people ask us this question all the time. I think there's the core model capability standpoint. Gemini is the only model that does long context. It's the only model that's natively multimodal, can take in video, can do audio and all that stuff. And really, if you look at what are some of the most successful product deployments of AI, it's people who are sort of leveraging the frontier capabilities. Like, there's not a lot of applications that are leveraging long context. It's actually like one of the fundamental enablers of really, really cool user experiences. And I think it's leaning into that. And I think this actually carries across not just Gemini specifically, but as you're looking at the models, finding the thing that that model is uniquely differentiated at, leaning into that from a product experience. And there's also, as you explore that state space, there's a lot of things that are less obvious, that are less talked about. But a good example in the Gemini world is the model is ranked one of the best at creative writing, which is not something that's super obvious. And you kind of have to figure that out yourself. But yeah, do that exploration. I don't know exactly what you're building at Cash App, but yeah, we'd love to chat more.
- [31:10]
C
Next question.
- [31:11]
A
Hi.
- [31:12]
E
Hi, Logan. Thank you, Mike, for moderating. My name is Jacqueline. I'm the founder of StarCycle. We help founders shut down their companies faster. Everyone, listen up. I'm just kidding. Quick, quick question is, would love to get your take on how AI agents, what that landscape is looking like, especially from your perspective, because a little bit of context is that we are building using the Google Agent builder actually to have AI agents do effectively many different parts of the shutdown process when a founder goes through dissolution. I would love to see specifically or would love to hear about what you're excited about, what agents are capable of doing. What do you think is a little overblown at the moment and where do you think people are underestimating?
- [32:04]
A
Yeah, I think people are overestimating the consumer willingness to Delegate very specific tasks to models. I think shopping is a good example. There's a whole tar pit of AI agent ideas that are going and have the model buy tickets for a flight. For me, I think all of those use cases are not actually going to be where most of the value is created. You should be doing things that the. Or I guess put a different way. I think the thing I am excited about is the models going and solving this long tail of problems that I'm not interested in solving. I feel like humans. This is a very human behavior, consumer behavior problem in that people actually like shopping is the very simple answer to that question. And you have to go and solve the problems of things that people really don't like doing or create such a better experience that they're willing to have AI augment it's hi, I'm Steve Wiss.
- [32:59]
F
Co founder at OpenAds AI. We use AI to generate custom ads in real time for every impression. So with each generation of foundation models we've seen new use cases unlocked just from the quality of the models. Gemini is interesting since it's running on TPUs. It's running at 200 tokens per second. What use cases do you see getting unlocked by just the speed of these models increasing?
- [33:24]
A
That's a good question. I think there's a lot of real time monitoring use case where speed is very obvious. A lot of things that have to do with taking actions based on what's happening in a video screen. So there's a whole long tail of video image understanding real time. You can imagine sports as a principal example of this where having an intelligent model is actually incredibly useful. But it has to be. It's incredibly latency sensitive. If I've already kicked the soccer ball and ran 20ft, it's no longer very useful if it takes five seconds for the model to say that that happened. I think lots of interesting use cases around vision and image understanding specifically, which also historically have been spaces where people have had to build domain specific models. I think there's this huge long tail of enterprise value that's gone to companies that built those domain specific vision models or internal teams. This is what I started doing in my career at Apple was training domain specific computer vision models to solve very, very niche problems. I think the LLMs of today with vision capabilities could probably do most of those use cases today, especially with how much faster they are.
- [34:40]
C
I also think we're going to see use cases that are already getting adoption, getting much, much more once they get faster. Even just answers through whether it's the Google AI answers or perplexity. These things are awesome. They're not that fast yet, especially when you compare it to just Google Search. Google Search is instant or maybe some of that software on demand stuff we talked about earlier. You need low, low, low, low latency to pull any of that off.
- [35:07]
A
I think a lot of the demand for tokens is actually rate limited by how quickly you can get tokens. Also the cost, but how quickly you can get tokens. I think there's a whole long tail of cases to be built, assuming that tokens were coming out at 1000, 10,000 TPS. Whatever it is. Hey, how's it going?
- [35:29]
D
I'm Andy at nyu. I was kind of wondering, like, what's.
- [35:33]
A
A question that's been on your mind recently? What are you thinking about?
- [35:39]
C
Is AGI? God is AGI. I'm just kidding. Or am I?
- [35:46]
A
I've been having a bunch of interesting conversations with people about how much having frontier capabilities that don't actually provide value matters in the context of getting people excited about what you're doing. I think there's a lot of flashy AI stuff that it's very clear is not creating the value. A lot of the value is much of the boring stuff. And this is probably generally tracks across AI and is not unique. But trying to find that balance of like, you know, do we do the thing that we know is going to be useful for developers and is going to create all this value, or do we sort of allocate our resources to do this, like, frontier thing that like, we know isn't actually going to be that, like, create that much value today, but is sort of an important signal to, to tell people where we're going and what we're capable of and, and trying to find a balance between those things is tough. And yeah, spend time thinking about it.
- [36:45]
C
Thank you all for the questions. Logan, thank you so much for speaking with all of us today. Everyone give it up for Logan Kilpatrick.
- [36:57]
B
Thank you so much for listening to Generative now. If you liked what you heard, please rate and review the podcast. That really does help. And of course, subscribe to the podcast so you get notified every time we publish a new episode. If you want to learn more, follow LightSpeed at LightSpeedVP on YouTube X or LinkedIn. You follow me at McNano M I G N A N O on all the same places. And Generative now is produced by Lightspeed in partnership with POD People. I am Michael Magnano and we will be back next week. See you then.