Day 1 Keynote Analysis | Google Cloud Next ’24
[Savannah Peterson]
Introduction
Thank you for joining us. We’re going to start in about a minute. Good morning, cloud community, and welcome to fabulous Las Vegas, Nevada. We are here at Google Cloud Next for three days of coverage on theCUBE. It is my first time at Google Cloud Next, and I can tell you, it is pumping in here. So excited to be joined by the cloud dream team with me. We’ve got John, we’ve got Rebecca, and we’ve got Rob. Thank you all for being here, for bringing your insights, your smiles. You’re reflecting the energy in the room. We all had a chance to see the keynote. So many announcements today. SiliconANGLE’s already got 10 stories up. Very impressive. Rob, you were doing a lot of homework before we got down here. Give us a little bit of a breakdown on what to expect this week.
[Rob]
Keynote Overview
Yeah, I think the entire keynote was about how to do cloud differently, and I think that’s kind of the theme going through this. There’s 30,000 people here, and you can feel it in the hallways. I think what they’re looking at is how they’re differentiating themselves from the AWSs, from others like Azure, and really staking their ground beyond people doing partnerships with NVIDIA, which they had some announcements around that.
[00:01:08]
It’s just been crazy the amount of announcements in the last eight months since Google Cloud Next in San Francisco.
[Savannah Peterson]
Keynote Takeaways
Yeah, it’s been a lot. Rebecca, what’s standing out to you, first impression, day one?
[Rebecca]
I think, to piggyback on what Rob has said, I think that that is exactly it. The fact that companies have gone from sort of nibbling around the edges to just experimenting with AI, to really trying to fully integrate AI agents, and he had some really heavy hitters up there. Derek Karushowy of Uber, David Solomon from Goldman, of course, Mercedes Benz, Ola Kalinas. And so they were really trying to show how here are some really brand name companies, how they’re using Google’s technology and integrating into their workflows.
[Savannah Peterson]
You know, it’s been a theme across our shows all year, that 2024 is really the year of making AI real. Last year, lots of hype, lots of interesting potential applications, but we’re starting to see it at scale.
[00:02:02]
John, would you say that you’ve seen something similar? I’m sure you had conversations about AI at the analyst event last night.
[John]
Google’s Position in the Cloud Race
Well, I mean, AI is everywhere, and I think Google’s really trying to find their footing in the cloud race. They’re still in third place. Some are saying Microsoft’s kind of getting ahead of them with open AI. Others are saying that Azure technically just isn’t as strong as people think it is. But Google’s really trying to find their footing, and you can see their posture. Hello, next big thing. Doing cloud a new way, the new way to do cloud. They’re trying to find that cool factor, and they’re really racing fast to fill in the blanks, if you will, Rob. So we’ve been seeing this keynote, a lot of AI, a lot of Gemini 1.5 Pro going public preview. Vertex AI is the key story, but if you look at what’s going on in the announcements, the trend is very clear. They’re going in and checking all the boxes. First thing, workspaces. That’s their big numbers. Gmail and all the AI going to be integrated into the productivity software. Two, the security cloud. Big time announcements there.
[00:03:01]
Data cloud, developer cloud, modern infrastructure cloud, networking, and then cross-cloud optimization around getting the hardware, the GPU, CPUs. So you can really see them laying out essentially the cloud stack for AI and working really hard to fill in the blanks, and they got to get more horsepower. The number one thing we’re hearing is it’s an IaaS game, infrastructure as a service, the new GPUs, the new TPUs, and they’re just grinding. I mean, again, finding their footing, be cool, but they got to do a lot more work.
[Rob]
Well, even their Axiom CPUs, where they’re going ARM-based CPUs that they’re announcing, and we’ll dig into that later today as well.
[Savannah Peterson]
Yeah, yeah, they’re talking about TPUs. I mean, there’s announcements across the board, which is so exciting. We were chatting a bit before the show got started. John, you’ve been coming to these events since the very beginning when it was just a tiny event. This is huge, and every division of their cloud offering is represented. I’m super excited.
[John]
Google is now officially, Google Cloud is on their own, even though they’re part of Google. They’ve really been fixing a lot of the things under the covers, like they go to market a lot of their operations, and they got a great team.
[00:04:04]
They’re attracting some good talent to actually stand this cloud up and be successful, but they have to kind of get faster with this wave. AI is going to give them an opportunity, and the theme that we’re seeing on the end-user side is they’re in line with a lot of the things that people want. They want the new user experience. They’re integrating into Workspace. They’re grounding Google Search with Vertex AI and Gemini. That’s a huge thing. That’s going to give better data quality, and you’re starting to see them taking the leverage of their user experience, so you’re going to see a lot more Google-esque front-end work, and then the Google big iron back-end, so you’re going to see two things emerge, more horsepower, the big Google brains that are building stuff at scale, and then the front-end is going to be all about how to make AI better, and that’s why agents are going to be the most important thing that’s going to happen in this show. You’re going to hear more about it in the coming months this year. It’s the year of the agents, Rob, intelligent agents.
[Savannah Peterson]
Key Theme of the Show
All right, all right, so I like that. You think that’s going to be a key theme of the show. Rob, what do you think is going to be a key theme of the show?
[00:05:01]
[Rob]
Security, Governance, and Sovereignty
I think security, governance, and sovereignty is going to be one of the big ones because Google always gets knocked on that, and I think you heard it earlier today on stage around, hey, we can bring that to the edge. We can bring it to where you need to be. You can bring it to your sovereign cloud, and I think so that distributed cloud, they’re leaning in hard. They’ve kind of rebranded what was some of their GKE everywhere type of stuff as well.
[John]
Google gets a lot of crap in the industry for inventing stuff and not taking advantage of it, and I think they’re focused on not having that happen again. They invented machine learning pretty much, and AI, as we know it, all came from Google DNA and some meta Facebook folks, but you’re starting to see them get focused. We’ve got to get in the game with the Google brains, and the Google brains, and they’re making big, bold claims on the keynote. They said they’re the only cloud that can do first-party, third-party, open-source models.
[00:06:00]
Amazon would say they have Titan, but that’s not as good as, say, Gemini, and then they laid out, Rob, the three things I thought was interesting around last year we heard about the model garden. Here they have model garden, model building, and then agent builder, so you’re starting to see where the trajectory is going. Get the models going, start building stuff, and then start deploying values. Rebecca, to your point, this is where the year is going to be. The rubber hits the road. If there’s no value, and by the way, it’s easy to get some value quick with data. If you have good data, you can get these agents up, and that’s why I think agents will be popular, and the rag is a good example. The retrieve augmentation generation is a tell sign that people are going to where the value is, and that’s going to be where I can make my data do something compelling to get a win. It’s a single, Rob, to use a baseball metaphor. Just get on base, you know, and this is the theme this year is show some value.
[Rebecca]
The Cool Factor
Well, one of the things I keep hearing from all of you is that hello, next big thing. This is the cool factor, but there’s also a lot of little things that they were talking about in terms of people’s workflows, in terms of Google Vids, in terms of Gemini.
[00:07:05]
These are the things that most of us non-big time technology nerds, kind of like all you guys, sorry, but this is how we do our jobs. We get on Google Meets. We have meetings. We need to find something in a document.
[Savannah Peterson]
We’re in sheets. We’ve got slides.
[Rebecca]
Exactly. These are sort of the little niggly things that we do during the course of our day, and to have technology powering them, making it easier, making it, refining these things. Here’s one thing that I thought that, I’m interested in your perspective, so I don’t know if you guys have ever heard of the term office housework, but this is something that women invariably do.
Office Housework and AI
We do more of it. These are things like, oh, we need someone to take notes. Who will do it? Oh, let’s get Susie to do it. Oh, we need someone to plan a luncheon. Who should do it? Well, Beth. Beth’s great at that sort of thing. These are non-promotable tasks. Invariably, women do more of them, 29% of them more. How about with the introduction of AI, with Gemini being the note taker, being the someone who can, the someone, the technology.
[00:08:02]
[(Unknown)]
AI Agents and Productivity
I’m going to give you a sentient being. Sure. We’re moving there.
[Rebecca]
Let’s give her a really a dude name. Let’s give this Gemini. Chad. Chad. Let’s get Chad to do it. But I think that this could be really revolutionizing for women at work, too.
[John]
The productivity angle’s off the charts.
[Rebecca]
Yeah, right.
[John]
And also, the level’s the playing field. So, I’ve always loved that term, democratization. Savannah, we joke about it. And democratize media, democratize data. This is truly level setting. We start to get into AI where creativity and intellect will be a big driver of the AI prompting and or human in the loop. I think that’s going to be a big productivity gain. It’s going to have not only the productivity, but who’s going to do what job? And what does it mean to be an IT person? Pressing buttons, use voice command like Star Trek? Hey, computer, fix that.
[Savannah Peterson]
Democratization of AI
Yeah. One of the things that excites me about the greater conversation we’re going to have today is Google is one of the companies that meets users where they are. Every single human in this room has used Google Search. Most of us probably have a Gmail or a G Suite.
[00:09:01]
And so, when we talk about democratization, think of all the people who may end up interacting or building with these tools because it’s already in their existing Google workspace. When it’s there and you’re not, oh, I’ve got to go over into this other sandbox and learn, you actually open the door to your point with office housework as well as general innovation. You open a playground in a space they’re already comfortable in with a UI that people already understand. So there can be a bit more of a gateway. I’m personally, I mean, we’re on camera right now. I’m really excited about Google Vids. I’m pumped that we’re going to be talking about it with some of our guests this morning to start off the day. I’m really curious to play with it. I mean, we’ve got tools that the team’s developed and that we use all the time. And I’m keen to see what’s up their sleeve with that.
[Rob]
Customer Presence at the Event
Yeah, it was funny because I think a lot of the announcements and they’re bringing a lot of customers to bear. They said they’re going to have over 300 customers talking about their journeys with AI and with a lot of different services. And I think that to me is telling, is that people gave up their time and may have gotten a discount, who knows, but they’re still showing up and doing that.
[00:10:06]
McDonald’s Use Case
One of the big ones that I know will be interweaved throughout the entire week is McDonald’s, which is interesting because they had Wendy’s here last year and McDonald’s is using it out at the edge and in, I guess you could say, edge deployments of distributed Google Cloud so that they can be within five milliseconds of each of the little hubs. So they’re using the actual Google Cloud WAN, which we’ll get to talk about a little bit later today. But I think it’s interesting when these big brands come out and say, hey, here’s how we’re changing, how people are changing our work style and how we’re using it to change the interaction with our customers as well. And I think that’s super exciting.
[Savannah Peterson]
I love that you brought up McDonald’s. I’m smiling to myself. Beginning of last year, I was creating some content in far north Queensland, up in the Daintree Rainforest, way up in the tip of Australia. There is no cell phone service for days up there, and I desperately was trying to get one last email out to a client before we lost service.
[00:11:06]
Eight hours into the abyss, the tour guide gets this ping on his phone and I’m thinking, maybe I can send an email. No, there’s no service, but what is it? A notification from McDonald’s telling him that he gets a dollar off his next coffee. And I thought, my gosh, what an example of edge. Here we are, could not be more isolated, quite literally, and yet McD’s is able to reach us out here in the middle of nowhere.
[John]
AI and the Future of Work
They’ve automated all their inbound, all their customer service. You just order at the kiosk, go to the counter, get your food. I mean, this is the, again, this is why I always tell my kids, I pay with cash because you have to hand over money. It’s a social construct. Hey, look, thank you. You have the touch of a human being. But this is the whole story about the AI. Workspace is going to be the future of work. You’re going to see the news they announced there, and that was, you’re going to see a ton of Gmail. You get the Gmail piece on the workflow there. You’re going to have Teams, not Teams, Meet, Meet is their thing, AI transcripts.
[00:12:04]
And the thing that got me was the context window for Vertex and Gemini being a million tokens. That’s a huge deal.
[Rob]
It’s an hour of video or 11 hours of audio. I learned that this morning.
[Rebecca]
And 700,000 words. I mean, whoa, mind-blowing.
[John]
And cross-modality analysis. I mean, it’s just going to get so much easier, not only to do tasks, but do real old-school work, grinding through and analyzing papers. So it’s going to be interesting to see how that plays out.
Grounding AI with Google Search
Now, I want to get your reaction, Rob, to the statement that he mentioned in the keynote. He announced that grounding with Google Search, which I thought was a pretty cool thing. I think it’s going to be a big deal because Google’s a lot of data over the years. So that’s going to help with the hallucinations. Then he said, ground it with enterprise data. Then he said, quote, tune it and connect with enterprise truth. And then prompt tools, they talked about prompt tools. So that is interesting. Tune it and then connect your enterprise truth to it.
[00:13:00]
What was your reaction to that? Because I’m like, okay, that’s the small language model. That’s the proprietary data or the company data.
[Rob]
Fine-Tuning AI Models
What was your takeaway on that? I think they’re getting at the whole fine-tuning of the models and how people, it goes back to the security and sovereignty, where they’re trying to say, hey, we’re not going to make you suck in all that data to train the models and train our models. We’ll have you fine-tune it at the edge with your information, your intellectual property. And I think that’s going to be just a massive theme this entire week for them is how we protect your data, and it’s not about us sucking it in and trying to make Gemini better. They’ll get enough of that out of Samsung with my phone and everything like that and workspaces.
[John]
Gemini 1.5 Pro and Tools
Well, I’m psyched to get my hands on Gemini 1.5 Pro. It’s in public preview. Other announcements, they showed some customer agents. I thought that was cool. I think them releasing an automatic side-by-side rapid evaluation, I think Google’s going to probably dump a lot of tools out there because I think they’re going to pick a playbook out of Amazon, which is win them with tools and abstraction layers and monetize the compute, monetize the cloud scale.
[00:14:11]
So they’ve still got to get that cloud scale horsepower up. And I’ve heard from startups saying, you know, there’s not a lot of compute available. Only the big customers are getting it.
[Rob]
Axiom CPUs and the Hardware Gap
Well, that seems like why they announced Axiom and where they’re going with that. I mean, that was one of the things that I was getting text messages about that from people on the East Coast this morning going, are they going to announce this today? And I’m like, well, hold on.
[John]
You’ll probably hear something at 9 a.m. And when he announced it, he called it the ARM-based CPU designed for the data center. Data centers are not going away, Rob. They’re hanging around.
[Savannah Peterson]
They’re having a little bit of a comeback, honestly.
[John]
They never left.
[Rob]
I think, again, part of that whole discussion was the distributed cloud aspect of it as well.
[00:15:00]
And I think those are going to be the themes, is privacy, security, do the compute where you need it, but do it on our stuff.
[John]
The Data Center’s Comeback
The data center is dead. Long live the data center, as they say. And what came out of supercomputing two years ago, remember, Savannah, we started tracking this early, and then last year at supercomputing, we saw it right out in public view. These purpose-built clouds that are coming online are changing the nature of these new kind of tier two or now specialty clouds for GPU, for instance, because it’s a managed service. The other thing that’s happening is we learned from GDC, which, by the way, was a total nerd conference that turned mainstream. Talk about flipping the script. The data center’s being re-architected. It’s not your old-school rack-and-stack, top-of-the-rack switch. It’s being redesigned for AI systems. And I think what’s coming out of our research, Rob, on your team and the group, is it’s the same game, different components. So Dell’s got a huge growth there with their AI factory PCs. HP’s probably going to do well.
[00:16:00]
Intel’s just trying to catch up by announcing processors, the pre-announcing stuff that’s kind of comparing benchmarks to products that are old. And so Intel’s struggling. So you’re going to see that game really change. And whoever can build those AI servers and put them in clusters together, so it’s a data center on-premise or edge, it’s the same game, but new systems. And I think that’s going to be the real undercurrent here. Is the TPU going to be in place for that system? I think yes. Google knows this. And we’ll see if they can produce.
[Rob]
Hardware and Skills Gaps
We don’t know. I was going to say, I think there’s two gaps I want to hear that are being filled this week. One is the hardware gap. And I think everybody talks about the shortages there. The other one, they briefly hit on this morning, and kind of right up your alley, was they had partners do a million hours of training on their AI. And there’s a skills gap. And I think a lot of the people we have on today are really going to talk to that in their global service, their GSIs and their system integrators are going to really talk to that.
[00:17:06]
But that struck me, a million hours of training already, which was just crazy.
[John]
Google’s Public Sector Push
Yeah, I mean, we have 10 stories up on SiliconANGLE. One story that’s a little bit out there, Rebecca, you’ll relate to this, because you’ve done some public sector events with us, with Amazon, Google announced hosted private cloud platform that can run public sector workloads. I bring this up, Savannah, because the news this past week, Microsoft got handed a huge blow with that government paper from a breach that they had from the servers from the government, the customer. So that’s going to put a black eye into Microsoft’s public sector business. Amazon’s, since Teresa Carlson’s left, is kind of like, I won’t say gone sideways, but they’ve kind of dispersed a little bit, not as strong, but they had the CIA cloud. So you have Google in the fray now for the public sector. And from what I’m hearing, they’re winning a lot of lift and shift business, Rob, in the VMware side of things and in public sector.
[00:18:01]
So you’re starting to see kind of like new land grab going on with public sector, and Google might mop up public sector.
[Savannah Peterson]
The Need for Complete Solutions
People are going to need a complete solution that they can trust. And that’s what you see everyone kind of scrambling to do is finding the best components, the best partners, the best way to plug in, the right APIs, the right hardware, the right GPU, the LPU, TPU. It doesn’t matter what we’re talking about. It’s this, how can we solve for these huge companies to be their long-term solution and keep it all safe? And I think that’s actually one of the bigger challenges. It’s only going to be more and more data. It’s only going to be more and more people. I mean, we were talking at the Women in Data Science event recently. Cybersecurity attacks have gone up so much with AI. So this is such a two-sided coin, and I think it’s a really important thing when we’re talking about government data, national secrets. There’s a lot of things there that need to be, it’s why we always see the CIA when we’re at supercomputing, which is fun.
Google’s Public Sector Expansion
It’s not what I expect to see there.
[Rob]
But we even have Karen DeHunt, who’s the CEO of Public Sector, coming on tomorrow as well.
[00:19:03]
[Savannah Peterson]
That’s going to be great.
[Rob]
That’ll be awesome, because I think, John, to your point, is that they weren’t only just talking about U.S. government, they were talking about other governments. And I think that’s a big thing, because if you look at where Google hasn’t been strong, it’s Europe, because of a lot of the privacy laws. So this will be an interesting potential coming-out party for them with governments, not just in the U.S., but abroad as well.
[John]
Google and CloudNative
Yeah, Google’s looking good. They can get these new workloads. I want to get your thoughts on Google and CloudNative. KubeCon just happened in Paris. North America’s coming up. Hen Goldberg’s going to be on, CUBE alumni. She’s now VP and GM of Engineering for Kubernetes and Serverless. Fantastic individual, human being, great CUBE alumni. She’s coming on, Rob. What’s your analysis from Google’s perspective on where they stand vis-a-vis CloudNative? What did you guys hear in Paris, and what do you think about their current position and their prospects?
[Rob]
I think given it was only like three weeks ago, they really were very quiet on what was going on from that.
[00:20:01]
And I think I’m really excited to hear Hen talk about it tomorrow because I think part of it is they have a lot of history, I mean, being the creators of Kubernetes. Yeah. And I think, you know, we had this discussion just over there in Paris that they were kind of absent from like the main stage, but they were in the different working groups, I think.
[Savannah Peterson]
Kubernetes and its Linux Moment
Yeah, yeah.
[Rob]
What did you think, I mean, coming out of that?
[Savannah Peterson]
Yeah, no, I agree with you. There were certain things that weren’t quite out in front as I was expecting. And so I am curious, I know Kubernetes is a big theme here. We were talking about if it’s having its Linux moment when we were over in Paris. It’s going to be a really fun week.
[John]
Guest Lineup
Yep. Well, we got a lot of guests coming on board. We got Palo Alto Networks, McKinsey, we have Arm coming on, talking about the processor Elastic, we got tons of Google executives.
[Rebecca]
And we have a cloud therapist, the cloud therapist. Yes, we’ve got Bobby, he’s wonderful. I mean, I’m so excited for that one.
[John]
Cloud therapist.
[Savannah Peterson]
Bobby’s great here, too. He’s wonderful.
[John]
Analyst Angles
I’ve never met him. And Thursday, we’re going to have a special analyst, I can’t see the person’s name, who’s going to come on, analyst, our analyst angles now, the plural, because we have so many analysts coming on Thursday.
[00:21:04]
We’re calling it analyst angles where we’re going to have all the industry analysts come on and talk about it. This analyst is going to come on, Rob, and talk about how VMware is losing customers that they thought they could lock in. That’s going to be very interesting to see. And Google is picking up a lot of that business. And we’re going to have Bobby Allen, the therapist on, and Gabe Monroy, VP of Developer Experience at Google. We’re going to have Data Bricks, and we’re just going to have all the top analysts come on. So that’s Thursday. It’s kind of open mic night here on theCUBE for analysts. Don’t tell me it’s an open mic night.
[Savannah Peterson]
Conclusion
Danger zone. Danger zone. But that said, we have an absolutely outstanding lineup. We have over 30 segments over the next three days. Really cannot wait to do all of these interviews with each of you. And thank you for your insights. And thank all of you for tuning in from home. We’re here in Las Vegas, Nevada at Google Cloud Next. My name’s Savannah Peterson. You’re watching theCUBE, the leading source for enterprise tech news.