What Google Cloud Can Teach Enterprises Developing & Rolling Out AI Tools, With Kawal Gandhi

0:05This week, Alad and I are joined by Kowal Gandhi. He works in the office of the CTO at Google Cloud, where he's the lead for generative AI. Gandhi comes from a long history of working on search and ads at Google before cloud. Welcome, Gandhi. Thank you. How did you end up working on cloud and then AI in particular from other projects at Google? Sure. I really worked deeply with a lot of our advertisers around search and ads for shopping and travel, especially commercial-based queries. And as working with them, they required a lot of storage, compute, infrastructure constantly around making their ads perform better, which led us to cloud and then cloud solutions around using some of that to create smart analytics, machine learning pipelines, more around documentation AI, conversational AI.

0:55And here we are. We've been doing it for a while, but now it's generative AI, where how can you make that customer experience much better with the information that they have? And just in terms of beginning to broadly Google incorporate AI into GCP, like what was the origin story of that? Is it TPUs, APIs, some customer need that you specifically saw? As we were getting into the Google Cloud, this goes back into how can we provide them with latency, high response, better experience with data is how our customers started kind of leaning towards Google, in my mind. From the beginning, it was more around machine learning and AI as a differentiator to work with Google.

1:43And how could they use that data better on our platform was a constant ask. So as we were on the journey of Google Cloud, it was all about data, AI storage, privacy, security, and kind of having that same deep technology that we used inside Google. How could we leverage that and offer that in market? So lots of learnings because what we built internally, we had tools, frameworks, et cetera, that took us time to kind of make our platform rich for our customers from regulated to non-regulated environment. and how to leverage some of their current investments on our platform. What are some of the internal use cases that have really driven that behavior in terms of the stuff that you ended up building for your customers?

2:28I know that a lot of what Google does is sort of dogfood its own APIs or products, and then it starts launching them externally as sort of a service that other people can use. What were some of those first applications of generative AI that occurred internally that then caused you to decide to do these things externally? Yeah, the early ones, I think it's all public now, was around workspace. Just using our documentation, our email, you'll sure use it. So it was like, can you summarize this better? Can you personalize this better? Can you offer me a suggestion? And all this kind of gradually as we, as we tested it internally in dog food, we gradually launch it externally.

3:08Because we see a lot of progress that we make internally in terms of efficiency, productivity gains. Folks can use it in their spreadsheet creation, etc. was gradually launched and now it's Duet AI, part of Workspace. So these are constantly being dog food and tested. We call them experiments. And as the research team leans in and looks at some of these, we add it to the platform and bring it forward in our products. Is there a single feature or product you launched within the internal Google version of Duet, the Workspace AI products that have gotten the most uptake or that you're most proud of?

3:45Yeah, I think we're seeing it across the board. aha moments. As we launch, we're seeing it in documents in terms of generation, summarization. We're seeing it now in slides with suggestions on images and new image creation, which is to take time for someone to go ask a studio or an agency, and you have a prompt that you can give and say, here's something I'm thinking about. So Sarah, we've seen kind of intake on all those features. Also, email generation has been super helpful from productivity perspective, not only for the consumers, but for enterprises as well. Security, we don't talk about it a lot.

4:25I think it's super secure, how it's sent, how the links are used. Those are something we really take it seriously. How far do you think we can take it with email generation? Because I maybe spend four hours of my workday just trying to keep up with my inbox. So this is of great personal importance to me. I'm not going to predict because you predict something and something always comes back and surprises you from a technology perspective, from a modern perspective. So I'm looking forward to see what surprises you. And I'm sure we'll speak again in six months and you'll be like, Gandhi, I saw this feature inside Gmail.

5:01It's really helpful. I like the translation feature for my mom because she speaks Hindi. She's really fluent in it. And when I write my emails now I can just say, translate this when she reads it. Now it does it automatically because she's using that app, which is fantastic. So I think it just bridges the gaps for a lot of our users internally and externally. I know that there's a lot of different services that Google Cloud provides, particularly around Genitive AI. There's a series of models. There's lots of domain-specific models that Google has really been forward-thinking on, things like MedPalm 2 or SecPalm or other things like that.

5:35Which of those are currently available and how do you think about that roadmap the things to expose over time. Sure. And I think it's helpful a lot to know that we have invested on the AI infrastructure. So you've seen people training, not only using our first party models, but also training their models and bringing it onto our platform. The next level up is the Vertex, what we call is capabilities of models that you can use from our model garden. That's where we have domain-specific models. So it's MedPalm, SecPalm, these are early days domain-specific. You can chat with it. You can kind of get validation of things, the notes that are written out from a nurse to a patient health care from a doctor.

6:15And it kind of puts it in the right format that can be saved up. So it's early days. We are seeing it from just foundational model capabilities to models that are built by our users and deployed on our platform to also open source models. So I'm really excited to see like Lama, stable diffusion, other variations coming onto the platform. But what's key, Elad, in all of this is your data is secure. You want to be like, you know, fine tuning it on the platform. And then all the model operation should be easy because we've spent a lot of time really specializing around the model drift, operations, tooling, safety around it super early.

6:57But I think those are the elements that will differentiate on the platform from the next few quarters. That's super interesting. Yeah, there's this guy, Ankur Goyal, who runs a company called Braintrust. It's focused on eval and a couple other things. And one of the points he makes, which I think is pretty interesting, is that there tends to be a very sequential adoption of LLMs by larger enterprises or people training models for the first time. So often they jump straight into fine tuning and then they suddenly realize, wait, wait, wait, let me just prove it out in GPT-4 or BARD or some other APIs.

7:27and then let me iterate towards something that actually works. And then maybe I go and train my own model. Do you see something similar in terms of the pattern of behavior that a lot of your partners adopt? Or do they just jump to your APIs? Or what's the most common sort of sequencing of people really adopting this technology? Yeah, I think it's always the case if you look in history. It's like, hey, I got an API, I can build a prototype, and it creates a lot of excitement. It's when you really start thinking about deploying it, managing it, monitoring it, using different models to chain them, have responses that really answer up a capability, you start thinking deeply about it inside an organization.

8:07So I see this as early excitement. It's not hype, so I want to separate a couple of that. It's real excitement because engineers, we love it. Like I've been one, you want to grab something, you don't want to have restrictions around it, and you want to show the art of the possible. And I think from those experiments, we're going to see in the next year or so capabilities that are available to the users and to inside the enterprise where they will see differentiated output and capabilities. That's how I see it. So I'm excited in all those phases, by the way, because it's the best creative time for engineers to solve problems that they've been thinking about solving for a while.

8:49and now they have the tools that they can download, whether they can grab an open source one, a closed one, and we can then kind of see how they're using it and evolving the platform along with it. Do you advise your customers and major Google Cloud customers how to think about when they need to train their own models or when they should be using domain-specific models? What advice would you have for them? Yeah, sure, Sarah. So we deeply think about it because you have to be responsible once you're building and fine-tuning a model? You know, what are the guardrails? What are the use cases? What is the cost you want to put behind that?

9:25Because it's a continuous learning process. And it also depends on the maturity of the organization. Do they have a team that understands how models are built, tuned, and then brought forward? And this, it doesn't mean that you cannot invest in it. It's just that where you are in the cycle is very important. So we lean in, talk about that from a board level perspective, from a strategy perspective, and then thinking through cultural transformations as well. So it's not limited to just one dimension in my mind. And it really helps them to think through how they see this transformation also happening inside their organization.

10:03Do you track the common use cases that your customers end up focusing on? Because it seems like there's almost three different things that people tend to do. Let's go experiment and just see if this is interesting. There's, I'm going to use it for specific internal tools. I want to make customer success better, or I want to make some ops workflow better. And then there's the people who are actually doing it for external products. I'm actually going to launch a feature that includes generative AI. Do you have a sense of sort of how that breaks out across your customer base and the proportion that's doing each at this point in this early cycle?

10:35Yeah, it falls into like, for me, it's like efficiency, productivity gains, and also creativity. And I did it in order because inside the enterprise, folks want to be creative, but every time they think about efficiency games inside their workflows. So how can I make my workflow better so I can invest back into my group? And then how can I make productivity better than I can make them creative? And I think we're seeing the flow inside the organizations. So what you touched on in a lot is, right, how can I make my customers successful? Like, let's start that use case. efficiency, how can I make my support better?

11:12And not drop my CSAT. That's the KPI I'm going to look at. And if that matches, then I go into productivity gains, offering promotions, next recommendations, then the trust increases. So let's put another dimension there of trust. So now you have your efficiency, you're going into productivity, you trust more and more. And the world we look at is like, how are these things going to work in the system of intelligence or agents in the future that the trust goes up and then you really free up your human capital to be creative, right? So I think we're on that trust cycle of like, how do we trust these models?

11:47They do what we think. They don't go off. They don't hallucinate. All those things are important. So that's how we think about it and then gradually make progress towards it and not get super excited and then find that the KPIs are not working, for example. Are there particular verticals outside of technology you see the strongest adoption in so far or the strongest interest? Yeah, I see a lot of interest. It started from sales and marketing. As soon as these came out from a horizontal perspective, I think it has an intersection between horizontal and vertical for every department, whether regulated or unregulated.

12:21There's bottlenecks in content creation, distribution, all that became like you could see the efficiency gains immediately and creative gains and then making that process and workflow shorter. So that was one. Customer care, you all touched on it, is another one that goes vertical across horizontal. Even if you are a B2B, you're touching a supplier and a supplier wants a good experience with their current provider. So how can that experience get better? And then now we're seeing them verticalize all the internal experiences. So how do I make my employees experience better? So you're kind of seeing that same technology applied internally.

13:01It's like, oh, my God, I'm looking for my HR benefits and do this X, Y, Z. Why can't we make that easy, right? So we're seeing the internal process now, whether it's regulated or unregulated. And I think there are these huge areas that we can provide opportunity with AI. That's my belief. I'm assuming that the vast majority of this starts with text, right? Text analysis, generation, et cetera. Where are you seeing, if at all, multimodality, voice input, text-to-speech output, any sort of other, even within sales and marketing, perhaps other modalities in people's use? Yes. So I think the core right now is language and text.

13:42And conversation fits really well into it. Like, you know, what we're doing, can you convert it into a script and you do translations around it and you have a distribution mechanism. Sounds easy now, but it was really hard three, four years ago. So multimodal, I think we're early stages now. So you're going to see the early models, which are audio based. So I think about it like text, then images, then images and text with audio. And then you're kind of combining these medias together. So as I said before, the trust on these, as it increases, you're going to see multimodal coming forward. There are things to think about like identification, creation, deep fakes that get created, voices that are used out of band.

14:25How can we provide that layer of safety around that, especially when it's used internally? And the data and the creation becomes an element of cost. How do we store that? How do we retrieve that? How do we scale that? So if I give you an example, I think gaming is a really good industry, which is kind of scaling out multimodality. And we have a lot of customers who are using that. Now think of that when you're going shopping and then the latency and the progress bar, like, why can't I see the video coming, right? So that's why you see it's centralized with some organizations who are experimenting, but we got to bring it at scale out, which will be like fantastic to see across not only the web internally as well.

15:07Are there any particular patterns that you see either organizationally or the types of projects that people start with that make your customers more and less successful with their AI efforts? Yeah, I think it goes back to kind of the, how much investment, how much belief, how much, you know, they want to be in the vision quadrant, first of a kind versus fast follower, all plays into it. It's less about the technology. It's more about like where they see themselves in the industry. And if they're seeing themselves as like, hey, we want to really build something and then scale out and then keep investing in it.

15:44I've spoken with a lot of customers, even governments. The interest in AI, I have not seen this. It's so high. We have just scratched the surface of this right now. So we can see a lot of these gains, which then can get invested back in the business. So I see a lot of positive signs around that. What's the most expensive part of the investment cycle, you think? I'm sure it varies dramatically from customer to customer, case to case. But when you say, you know, it depends on commitment level and investment, like what's expensive? The expensive part is now becoming cheap. It's the models, the availability, the usage of the platform.

16:26Those were the things that were really expensive. I think if you do a cost curve right now and the investments you all are making into the ecosystem, I think we're seeing that just phenomenally coming down, which will allow people to adopt more. So I think we're really entering a phase of Cerala and Elad, like a growth phase of just people adopting this more and more. We are seeing signs of just like, how fast can you do? How fast can we learn? We've had training classes and we've had people certified. I've never seen this for a year. We have more than 10 ,000 people certified now. Just more on GCP, generative AI, globally, right?

17:08Ready to write code, take advantage of code capabilities, engineers' productivity going up, which used to take time, migration of systems. Can we make that easier? There's a ton of use cases that are helpful. We're just bringing this forward on the platform. Are there any anti-patterns or big mistakes you guys have made internally or that you see customers make when they're trying to get these efforts into production or even choosing use cases? Not mistake, but we think deeply about the data that customer bring to our platform. And they're not mistakes, but we worry about any of this, like, you know, not being used in a way or something happens.

17:51So we take that very seriously. So we run a lot of drills. We make sure that data is kept secure. the models that are trained, even if there are early days, no leaks happen. The adapter model, the certification, everything stays in that tenant. So from the beginning, we just made sure that all of their data, all of their models, all of their weights, that's like their IP. And we want to safeguard that. So none of the mistakes, just deeply have to think about how fast we move and what are the checkpoints we need to make in between to make sure we don't move too fast that we get into mistakes. Those rollbacks are expensive, by the way.

18:29Then you have to again educate the industry, make sure they understand. Rather, not commit those is where I'm going. Yeah, Google's always been very good at data security and ensuring that customer data is well secured. I guess related to data, there's a number of verticals that are very data intensive or alternatively where the data can be quite sensitive. Healthcare is almost a canonical example of that where, you know, bespoke data can really help build a more interesting AI service, sort of like what you've done with MedPalm, but at the same time you want to secure it properly. Are there specific verticals or use cases that you see adopting AI soonest?

19:04For example, do you see healthcare moving ahead of education or do you see fintech or financial services as sort of the early adopter wave? I think they're all kind of early and adopting in all those verticals that I was mentioning or horizontal like sales and marketing, creative approaches. I'm seeing engineering adopted much faster internally now. Cody models with the open source. It's a surprise, by the way, if I had to share my personal. I thought engineering was so productive and they're writing the right. Just design elements of UX and the discussion you have around that. I've been there.

19:43It takes a long time. But then having a bot in the room saying, here's how you can approach it or here's how code could be written capabilities, it's fantastic to see that. So I see a lack of resources which was causing engineering projects to be bottlenecked. Now the transformation is going to be much faster. So bringing that across to vertical is fantastic. That's the IP that's getting created now inside those organizations. Yeah, I'm seeing something very similar in the startup world where a lot of the early startups are basically technologists building stuff for themselves. Or mid-market tech companies is sort of the earliest adopters outside of a Google or a Microsoft or sort of the really cutting edge large tech companies.

20:25And so it seems like, you know, at least in the startup scene, there's a very similar pattern being mirrored relative to what's happening on your cloud in terms of, you know, developers are building for themselves first. Mid-market companies are kind of next simply because they're often driven by a developer or CEO. And then it kind of is starting to eke into the other parts of the world in terms of adoption patterns. So it's interesting to see that parallel. It's very interesting. And think about the investments you all make. Now I have a platform in GCP, which gives me models. I have a coding model that helps me code.

20:59And now I just need capital allocation to go experiment. I think that's great. And then now I need to make it secure and scale it up. That's where we come in with our infra and then make it cost effective. I think we're like early stage of new era of even startups and big companies coming out of this and different solutions getting built out. What you said earlier was really interesting about having a bot in the room, so to speak, helping us with UX discussion or speccing a product. Where does that live for a Google product team? Does it live in a chat interface? Does it live in documents? Does it live in the IDE or in version control?

21:38I think a lot of people are trying to figure out how this should integrate into existing developer workflows. Yeah, I think today in my mind, it lives in docs. So you get a lot of like, you know, benefit in the documentation and that's shared across. I think you've seen it like any project is like, let's start a doc. Any hypothesis or experiment, let's start a doc. I really think there will be an evolution as the platforms mature where the docs can make suggestions. And those suggestions could be here based on what you've specced out. Here is output. Now, you can choose to ignore it. It could be a notebook as well.

22:13You said IDEs and UX. It could be just code written with put the API key here. It's great to kind of do that. But then when you're scaling up, that's where I get scaling. How does it scale up? You don't want these things to become something that doesn't have the guardrails around them. Who is entitled inside the organization to do this? So we have to make sure logging is on. Simple things like logging we will not think about, but has to be who accessed, when did they access, etc. But I think once we get over that, if you're writing a monitoring tool and it knows the schema, you don't need to go back and look at it and have an engineer work on it for like four weeks, right?

22:53Like, why can't it be generated out of, I just need a monitoring tool for model drift? So those benefits are going to just accelerate adoption in my mind. One of the things that Google really pioneered on top of all the cloud services and APIs and everything else is actually the silicon layer. I think it was seven, eight years ago when TPUs were first really rolled forth as something internal to Google. And they said, let's invent silicon that's really good at dealing with AI and ML systems. And so that was my understanding of the origin of the TPU. and it obviously was dramatically more performant than GPU for a long time.

23:26Could you talk a little bit about the TPU versus GPU trade-offs and how Google Cloud approaches that? Yeah, I think how we think about it is like, can the platform be capable enough and have features, as I call them, that takes away the developer or engineer or team thinking about a project from the infra piece? Early days, we used to write Windows app. I've written like Java application, and it's always about abstracting it away so you can manage it, scale it, and then roll it out faster. So I think we'll see a world gradually, other than training and where it's trained and deployed. Is it abstracted out from that layer in a way that it gives the fungibility and the adoption and the scale out that helps you kind of do that?

24:13That will speed up in my mind. And then to your point, innovation on that layer is going to also continue. Are there specific trade-offs you see in terms of when you should use TPU versus GPU, or do you think they're reasonably fungible? They're reasonably, in my mind. It's just more around how many folks know how to do it directly is something top of mind for us. Because there is this that was externally available, and a lot of people are trained up. As more people get trained up, we're seeing costs and benefits across the board, right? So having optionality is what we want to kind of bring forward.

24:50constantly. The people that I know who know how to use TPU very well are very comfortable in that environment. And to your point, that's often people who are at Google and therefore they're trained to know how to actually use that underlying silicon versus the GPU in terms of optimizations or little tricks that make things more efficient. How can we make that more prevalent and available? It's something we should invest in in the future. I'm going to see that. More than that, I'm on the team as such is thinking about inference because you get the trend, you get the application ready and then the scale is where you don't know about character ai is like really thinking about scaling it out we have other examples in quora and po adams thinking about scaling out the application so now you're going to see the shift on like we have v5e etc we are investing uh heavily and you know others are doing is how do you scale that's and then keep the cost curve kind of going in the other direction.

25:48We're going to invest and going to put our effort around that. Yeah, and I guess since Google trains its own models and they have an advantage in terms of thinking through how that scalability could work for others. That's right. And that may be an interesting differentiator in terms of that understanding. Yeah, I think inference will be something that really needed. We can see the numbers, right? Like an application launches and it goes from like 5, 10, 15, 100, right? And you're like, how do I scale this further out? And Sarah touched on it. What happens when we get into multimodal world? And if I'm an educator, thinking about an education site, I'm not having that kind of luxury.