On this episode of The Six Five – On The Road, hosts Daniel Newman and Patrick Moorhead welcome Will Grannis, CTO at Google Cloud for a conversation on the latest from Google and Google Cloud, Generative AI and a preview of Google Cloud Next ‘23, happening later this month.
Their discussion covers:
- What to watch for at Google Cloud Next ‘23, happening August 29-31st in San Francisco
- An introduction to Will and his role, which sits between Customers and Technology
- How this current chapter for Google and Google cloud is unfolding from the customer perspective: Why Google Cloud?
- How Google Cloud is helping businesses take advantage of generative AI, from content generation and chat summarization, to building custom chat interfaces and digital assistants
Watch the video here:
Or Listen to the full audio here:
Disclaimer: The Six Five webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.
Sponsored by Intel
Patrick Moorhead: Hi, this is Pat Moorhead and we are live here on The Six Five. We are back to event season again. Very, very excited to do that. I think I’m on the road for, I don’t know, five, six weeks, but that’s what US analysts do. It’s part of the job. It’s in the job description. I think I may have written the job description. But anyways, Daniel, how are you doing my friend?
Daniel Newman: I’m doing great. I guess you didn’t get the memo, but all the events are remote and we can do them all from home now. I was told that at some point in 2020, 2021. I was also not told that that would become completely untrue and that we would be on the road 47, 48, I think it’s going to be 48 weeks this year, Pat. So you know what though? Who’s complaining? These events are awesome. I love technology. It’s so much fun. We are out there and this next one’s going to be great.
Patrick Moorhead: No, it is. And we are here, we are talking Google Cloud Next 2023, which Dan, you and I will be there live on site. But we are excited to have a conversation talking a little bit about what to expect in Google Cloud Next. But the great part is you’re not just going to have to listen to us talking about that. We have Google Cloud, CTO Will Grannis will welcome to The Six Five and thanks for being a first time guest.
Will Grannis: Pat and Dan, an honor to be here. Thanks for letting me crash the party.
Patrick Moorhead: Yeah. Gosh, I think I went to my first Google Cloud Next, I think it was on a pier somewhere in San Francisco and the company has just come a long way and it’s… really looking forward to the next one. Sorry I interrupted you.
Will Grannis: Well, I was going to say the first… so I’ve been here about nine years now. First one I went to, we had in a ballroom of a hotel and you had the little, the popup stand, pop up stage and you had to watch you didn’t trip over it. It has certainly come a long way.
Daniel Newman: Yeah, it’s been a really great journey to watch. Last couple quarters, the company turned profitable. I think there was a lot of eyeballs on when was that going to happen, Will. And I know you being in the CTO’s office, you probably think about that a little bit less than maybe some other people that are in the business. But at the same time, as part of the leadership team, I’m sure that was a really proud moment for you and the company.
But it’s not just about the dollars, it’s about being able to offer the value to the enterprises driving the type of services that companies are looking for, enabling not just hyperscale cloud and cloud infrastructure, but also software services and connecting to that old hybrid cloud. Pat will do a victory lap about that at some point in the show. He says he called the hybrid cloud early. But…
Patrick Moorhead: 10 years ago. Okay.
Daniel Newman: But the cloud isn’t… it’s almost old enough to drink now. I mean legally in the US. It’s definitely legal in Europe. But how about a quick, just give us the quick background and rundown, Will. We mentioned you as a CTO of the company, but of course, the office of the CTO and its purpose, its direction and what you focus on day in and day out over there at Google Cloud.
Will Grannis: Sure. So my function is actually a new build with our cloud organization and it dates back about 2016, 2015, 2016. And so I come from the enterprise, I’m an enterprise technologist. Most of my background has not been in pure tech, it’s been applying tech in the context of a business, to build things with, application, services, experiences.
And so when I first showed up at Google, we were really just laying the seeds for our cloud business. We had Compute Engine, we had BigQuery, we had App Engine, we had some services. But we’re really mobilizing on the business. And to the credit of leadership at the time, we realized that we were still approaching the technology development very much like, hand things to developers and let them go build. And what we realized very quickly working with enterprises that they needed some guidance, “Watch out for these pitfalls. Who are my peers that are going to have the experience building with these tools that can really help guide me?”
And so we started a little group called the Office of the CTO around that time. And the whole purpose was to go attach ourselves to the top end number of customers and make sure that they had an awesome experience building with our technology and kind of took off, which is why I guess I’m still employed here at Google. And over the years we’ve built a team and we’ve added a little bit a new spin. So not only do we work with our top customers now and help make their path to cloud a little bit less bumpy, but we also take what we learn and we connect the dots from what we hear from these top customers. And you can imagine how valuable their feedback is and input is when you’re building multi-tenant platforms and multi-tenant technologies.
So now we also take some of those dots, connect them and we pilot some new capabilities, some of which you’ll see at Next and some of which are already baked into the platform. So it’s really a privilege to sit as a technologist, at the junction of the world’s preeminent scaled tech company and all of these companies who are trying to, outside of Google, who are trying to apply technology to solve really, really important problems like precision medicine or sustainability or commerce or supply chain. So as you can tell, I think I’ve got the best job around.
Patrick Moorhead: Wait a second, I thought I had the best job around. So maybe we can compare notes at the event in real time. So yeah, over the last 30 years, I’ve met with so many different types of CTOs. There’s CTOs that kind of stay inside of the hive and never go out. There’s some that, like you talked about, talking with customers a lot and having this bi-directional conversation and more than a conversation, actions that go in it. And by the way, I think there’s value in both, but if you want to really move the needle from a customer viewpoint, you have to have the latter, which is what you’re doing.
So as industry analysts, we pay attention to a lot about what Google is doing, your competitors, your complimenters, your entire ecosystem. And there’s been a ton of news lately and I would even say a lot of momentum that I’ve seen in terms of events relating specifically to things like AI, and at your AI event that you had for analysts and others, you rolled out a litany of customers there. I wasn’t personally there, but I got the download. But why are they choosing you over your competitors?
Will Grannis: Yeah, it’s a great question. And I’ll extract it a little bit because quite often the decision to work with Google happens for reasons that are not just about product features and roadmap, but actually are more enterprise institutional relationship trust based. So I think of them in three different categories. First, our customers want to partner with the proven leader in AI, meaning they want to partner with an organization that has operated production grade machine learning systems, platforms, applications responsibly and securely at scale for years in front of billions of users. Every single day, all of our machine learnings production systems are on display and ready for scrutiny by almost everyone in the world. So they want to partner with a true AI first company.
Patrick Moorhead: You mean like PlanetScale?
Will Grannis: Someday. Not to tease too many things out in the future, someday interplanetary scale. But more on that I think in a future episode. So they also are very interested now that Google Cloud is this enterprise grade business within Google, we’re also externalizing these AI innovations so that enterprise customers and developers, any organization really, can build with and on the same technology that serves those billions of users every single day.
One example I like to give is Tensor Processing Units. These were originally designed so that we could efficiently scale our own machine learning production systems and we started externalizing them to customers and they started to find the same benefits and efficiency and scale and now we’re in multiple generations that have been available to customers. So for years our customers have been able to take advantage of these purpose-built chips for AI. So that’s reason number one, right? AI first company, a true leader.
Second is they get a real voice in what we built. And I mentioned a little bit of this in the intro, a little bit of what my team does, but if you look at our product roadmap today and at Next, you’ll get a view of both what we’ve built and we’ve landed and also some things that are launching, it’s a direct result of customer feedback. So I’ll give you an example. We know that databases are going to play a really large role in generative AI. And many of our customers wanted an open way, a site to bring vector search to traditional databases for example. And so this may sound like a small thing, but we recently announced pgvector, which is the leading open source similarity search extension. It’s now available with Cloud SQL and AlloyDB. And I could list thousands of features and products that are now available to all companies based on the feedback of our customers. So now our customers get the best of Google technology, like TPUs, and the best of open source and third party technology all in one place.
And the third reason is that no matter how big or small these companies are, it doesn’t matter what geography they’re in, what industry they serve, our customers know that they’ve got a fully featured set of platforms that are tailored for what they need to build, deploy, and manage their critical applications. So this shows up really in a couple of different ways. One is the unified data and AI platform. So scaled, storage systems, data stores, databases, we just mentioned some features and databases that are going to help them in gen AI. So from data to pipeline to application, it’s you can move from one place to the other in the platform very seamlessly with horizontals that are very consistent across the platform like security, monitoring, logging, things that are really important to operators.
Second, we have a dev cloud. So this helps businesses and governments build able to quickly, securely and cost effectively. Purpose-built infrastructure and AI optimized platform. Third, we’ve got AI powered features in Workspace. I think Workspace is one of the most exciting surfaces we have because you can immediately get the value of generative AI in a SaaS application that feels almost like a little bit of magic sprinkled in to what most people, it’s like collaboration, communication, but now in this new way. And then finally, this engineered insecurity and world-class threat intelligence teams. We continue to invest in key technologies to progress towards what is our North Star, which is invisible security. This is strong security that’s pervasive and simple and mostly invisible to everyone.
Daniel Newman: One of the big waves, Will, is definitely generative AI. And that’s been a bit of ride. I kind laugh because it was like, we always kind of talk pre-pandemic, post pandemic. Well, how about the pre Thanksgiving, post Thanksgiving? We legitimately saw this industry get turned upside down in a very short period of time when all of a sudden we saw what generative AI could do. For Google and Google Cloud, the company has been all over it. It’s been very exciting to watch. Pat and I always talk about innovation drives innovation. It’s been exciting to watch the back and forth with Google and OpenAI. It’s been fun to watch the different cloud providers all rolling out different strategies.
But I was at your event, the customer event in San Jose or in the Valley and you guys launched some updates to Vertex, you launched some updates to your generative AI tools and I was super impressed. I really said at that point, Google Cloud has really done a great job. It’s innovating, it’s bringing solutions to market and it’s helping customers do what they ultimately want to do, which is not all be data scientists and not all be programmers and not all be GPU experts, but really just get their data rights and then utilize it for applications to help them be more productive or more efficient in their business. Talk a little bit about this whole generative AI journey that you’re on and how this is really driving momentum right now for Google Cloud.
Will Grannis: Yeah, absolutely. And it’s no surprise to me that being a Google person, that generative AI, large language models are creating this next wave of innovation in scaled computation. We’ve seen this since 2015, 2014, 2016, the publication of our papers around transformers, with DeepMind’s publications around reinforcement learning with human feedback. We have been not only a part of, but actually in my opinion, leading a vanguard of innovation in this area. And the moment that you mentioned last year, the reason why it’s so exciting to us is because it’s really opened up interest and curiosity from kind of the entire ecosystem and how to build with these technologies. And so that for us, that means as an AI first company, we’re going to focus on making it easy and scalable for everyone to build with AI, not just our research teams and our product teams here at Google. So that takes a couple of different shapes.
First, I believe we have the most comprehensive platform now available and ready to go. We have strong support from leading organizations that you’ve probably heard about at the event. We’re helping companies create content for marketing objectives. We’re helping synthesize and organize information. True knowledge management. One of the projects that every company has always wished would come true, but generally they fail because it’s so hard to pull information together across silos and make it universally accessible and useful, if you will. We’re helping to automate processes and also build engaging customer experiences with this platform. So first, the most comprehensive platforms. You can summarize content, you can generate content, you can build customer experiences.
The second element that we’re really focused on here in order to democratize this, and this is the number one question I get is, “Do I have control of my data, brand, IP risk, ability to meet regulatory requirements?” This is the trust question. And so unequivocally, the short answer is yes, our customers have full control of their data, their brand, IP risk, and ability to meet regulatory requirements. We don’t use our customer’s data to train Google’s models.
Third, to your point, Dan, everyone can be an AI developer. You can even drag and drop now and build applications using some of these assistive and generative tools. And so all users with varying levels of expertise can create innovative enterprise search, chat, even vision apps now. And we’re enabling both business and tech practitioners to be more productive using these AI assistants.
Then finally, and this is a really big deal because the excitement is always in the use case and that first application, and then companies that want to build and scale out, they go, “Oh, maybe we hadn’t thought about the infrastructure and the scale out that’s going to require, it’s required to serve however hundreds of millions of users they may have or maybe their B2B, so they’ve got an ecosystem of partners as well. So we’re also delivering infrastructure that’s optimized for AI workloads. So you can get the latest GPUs and TPUs. At Google, you can get a rich choice of deep learning VMs and the ability to easily build custom AI software. So it doesn’t matter what stage someone’s in, our goal is to make sure that the platform, the data approach, the development tools and the infrastructure are all ready to support that.
These are rapidly emerging use cases. And so we’re also kind of as a wrapping layer now. We’re already putting out a lot of educational and consulting programs, blueprints for specific industry use cases across a growing partner ecosystem. One of the things you’ll see at Next is just an unbelievable explosion in our partner ecosystem, developer ecosystem, third party startups. And I think we’re seeing it because they see the commitment that we have to help them learn, build with and deploy generative AI.
I like talking about customers because my job is like 50% customers, 50% internal R&D. So I did want to give people who are tuned in here, a couple of examples just to make it a little more real. So for example, in financial services, regulated industry, Deutsche Bank is testing Google’s generative AI and large language models at scale to provide new insights to financial analysts. So this helps them drive operational efficiencies and execution velocity. They believe, and so do we, that there’s an opportunity to significantly reduce the time it takes to perform banking operations and financial analyst tasks. And this empowers employees by increasing their productivity. So that’s in financial services.
Who doesn’t love a drive-through? Every time I think about this customer, I get hungry. So I’m sorry if I’m doing this to you.
Patrick Moorhead: You can probably look at me and tell me…
Will Grannis: Yeah, so I’m going to hurt you here. This is all about frosty’s and fries and dipping the fries in the frosty’s. Yes, that’s right. Wendy’s is piloting a groundbreaking AI solution. It’s called Wendy’s FreshAI, and it’s designed to revolutionize the quick service restaurant industry. And the technology has the potential to completely transform Wendy’s drive-through food ordering experience. And this uses Google Cloud’s generative AI and LLMs by initiating a conversation with customers that is as fluid, some would argue, maybe more fluid sometimes than the conversation people are used to having in the drive-through. And it can understand made to order requests and generate responses to frequently asked questions.
And maybe a third example, because I like 3s, telecommunications. Orange France. So moving over to Europe, they recently launched Orange Bot. It’s a French language generative AI enabled chatbot. So not just English, but now we’re also into other languages as well. It’s embedded on their website, use the company’s support knowledge to independently generate precise and immediate responses to customer questions. It’s also a conversational search engine and an entry point to their help and contact website. This is one of the most common things that we see today is these chatbots use, as a natural language experience, it’s like the front door in to the knowledge that a company has and quite often, the support experience. And they believe at Orange that they’re going to be able to transform the customer relationship, optimize management costs, and offer even more helpful and user-friendly experiences in the future.
Patrick Moorhead: Well, everybody does like customer examples, really where the rubber meets the road and the tech connects, and that’s when you really see the value. I like cool tech, just like everybody, but will customers get benefit from it, will they pay for it, that is a completely different story.
So hey, we’re here because of Google Cloud Next, and it’s literally right around the corner. What should we be keeping an eye out for? And if you want to pre-announce everything on this video that’s going out beforehand, you can do that if you’d like, but we’re not press.
Will Grannis: Only the really good stuff.
Daniel Newman: Hold on. I need to find my badge so I can shred it.
Patrick Moorhead: No, I’m just kidding. What should we expect? What are we going to see out there?
Will Grannis: Well, first, for anybody dialed in, I mean, Six Five is going to be there, right? So you want to see more Pat and Dan, who doesn’t? I’d say right there is probably the big headline.
Daniel Newman: Cut. Cut. That’s it.
Will Grannis: Yeah. Is that it? Are we done? Yeah, that’s all we need.
Patrick Moorhead: Can we get your wiring information please?
Will Grannis: Yeah, absolutely.
Daniel Newman: I love it. Will. Thank you.
Will Grannis: Yeah, no worries guys. Listen. So first, Google Cloud Next, August 29th to the 31st at Moscone Center in San Francisco. First and foremost, it’s a user conference, so it’s designed for practitioners. So it’s going to be heavily leaning towards giving knowledge and experiences that help people learn.
So three big themes at Next. First, generative AI, as Dan would say, is everywhere. So for a couple of examples, we help the world’s leading customers build AI with our optimized AI infrastructure and Vertex. So you’ll hear about that. We help every person use AI with Duet AI. So this is this always on AI collaborator that provides help to users across all the surfaces here at Google. So they’re going to hear about that in Google Workspace and Google Cloud platform. And with our ecosystem of partners, we help organizations successfully incorporate AI. So that’s the first thing that they’re going to see is generative AI is everywhere.
Two, that they can get the most out of generative AI when it works with their cloud. So this could be people, processes, security. We’re going to really dive deep, not just about some cool chat app or the latest LLM, but we’re actually going to get into operations. LLM ops. How do you secure a chain that starts with a prompt and ends up with a piece of generated content and everywhere in between. So really making sure that they have practical hands-on, operator level knowledge that they can take back with them to their companies.
And third, regardless of your role, taking advantage of gen AI is so much easier than most people believe. So for today, today what’s happening in companies like in finance, accounts receivable and accounts payable, summarization with SAP. In HR, HR Help Desk, job post creation benefits, summarization with Workday. In procurement, vendor contracts creation summarization with Ariba. Sales, generating letters for prospects, marketing, advertising, copy creation, social media publishing. Just so many things happening already and we want to expose that really to embolden people to take the next steps with gen AI. But we’re also going to hint at some of the things that might be coming in the future and how AI will transform industries.
So a couple that are near and dear to my heart, coming from manufacturing industrials, government first, factory floor monitoring, quality management, using vision and image AI on the factory floor. On field service, generating information from a manual into an application, speaking with the application, using voice-based search, taking a picture with a camera and auto generating summaries for each field service person. Super interesting and really popping up in Europe quite a bit with quite a few of our customers.
So we’ll share these and a whole lot more. There’ll be keynotes, there’ll be experiences. Attendees will be able to find just about anything I think they’d want in terms of code labs to executive level presentations. We’re talking about business outcomes. So it’s not just about the tech. And we’ll clarify across tech people, process and open ecosystem, how we can help customers learn, implement, and be successful. And most importantly, our customers are going to be there, and they’re going to share with each other and they’re going to share with all the attendees, the ups and downs, the sweet and the sour of gen AI at scale in production today. And no more authoritative source than our customers. So a lot to consume.
Daniel Newman: I am very excited, Will, to get there. And Pat, I think I can speak for you. I like hearing it from you, I like hearing it better from your customers and then I like it the most when it comes from me. But I would not have been an analyst had I not said that. But in all serious, you’ve really laid the groundwork for a very exciting event. Of course, customers are thinking about everything from how to just scale traditional workloads all the way to how do we AI everything. And we know that AI, it’s not a technology that sits vertically, it sits horizontacally… horizontacally? I made up a word. Horizontally. And it’s going to impact every vertical within industries and every part of the business. And so I want to hear more. I think I can speak on behalf of The Six Five. We want to hear more. And Will, we do look forward to hearing from you and TK and the leadership at Google this coming week. Thanks so much for joining us here on The Six Five and for previewing Google Cloud Next.
Will Grannis: Absolutely. It’s a pleasure and honor, gentlemen, I’ll see you there.
Daniel Newman: Sounds good. Looking forward to it.
All right, everyone hit that subscribe button, join us. We’re going to be talking more with other experts throughout the Google Cloud Next period, weeks. We’re going to be doing it before, we’re going to be doing it during, we’re going to be doing it after. So stay with us here on The Six Five, hit that subscribe button, join us for all of our shows. If you love what you hear, send me a direct message. And if you weren’t so happy, you know Patrick’s Twitter handle. All right everybody, we’ll see you all later. Thanks for tuning in. Bye-bye.