 Hello, and welcome to theCUBE pod episode 54. I'm John Furrier with Dave Vellante, the CUBE hosts. We extract the city from the noise every week. This is our podcast. We break down what we're working on, what we're seeing, what we're doing. Obviously, we're kicking off event season. We just came back from an amazing week at Google Next 24. Dave has had some real proprietary NDA meetings with some of the big players in the industry around the country. And of course, we're digging into all the research on SiliconANGLE, Rob Ho from the team there, doing a great job. CUBE research continues to put out great content. Dave, great to see you. We're kicking everything off. Again, the season starts, but what a week. You can hear in my voice, I'm a little bit tired from just really kind of cranking heart at Google Next. We did like 40 interviews. We have over, almost 50 interviews and over 500 videos we produced. I think SiliconANGLE has over 30 articles posted and more coming. Just a barrage of content. I did a segment on the New York Stock Exchange. We did a live remote. Just Google Cloud kicked ass. We're going to dig into that. Andy Jassy was on CNBC. You saw the interview. I could not get it because it's on the pro blogging. I did see a clip from Brad Gerster from what Altimeter Capital had a post around his really strong posture against the regulation and he used iRobot as an example. I know you want to talk about iDo too, but you watched it so I know we have thoughts on that. So make sure we hit that hard. Intel had an event and basically all the analysts blew off Intel and they came to Google Next. So you saw a lot of people either not go to Intel's event or just snuck out the back door. Okay, did an Irish exit, as we say. And so Intel's got challenges there and they continue to try to polish up their position, which is not good. TSMC, I think outplayed them again this week with their funding from the government. Intel got their subsidy and TSMC is investing as more. So we'll see what Gowdy does. Gowdy 3 is a seven nanometer, Davis. It's starting out behind. And I want to get into this because I did some research on this, okay? Talked to a lot of experts on this. Gowdy 3 might not make it, okay? It might be a little too late and a dollar short or in this case, nanometer short, too big. We're going to get into the energy aspect of it. And also people are smoking, you know what, when they get the benchmarks. Had a great conversation at Google Next around how people are sandbagging the performance numbers with the new tokens. So in AI, it's all about how many tokens per second can you do? Well, guess what? There's now controversy brewing around tokens per second because you can juice the tokens number with more energy. So we're going to get into that cybersecurity consolidation. Big blowback from our segment last week where you introduced the paper with Microsoft. That's interesting. And just a lot of news from Google Next that surprised me, we'll get into that. But it's just an unbelievable week and the whole team is cranking. I mean, the Google stories is off the charts. Maybe it's more than 30, but you see, you're seeing that hands with. And a surprise note, and I knew, I knew it was around that time, but it was actually an event in Vegas with Google Next. This is the 10 year anniversary on June 6th where Kubernetes was launched. First commit was June 6th. So unbelievable time flies day. Remember when we were sitting there saying, hey, we're going to make a bet on Kubernetes. This is going to be big. This is going to be the clue that unifies the industry. We don't want another open staff. We want to see something. And we made a bet. We went to every single KubeCon. We invested our own money into the community and got to meet those people that were starting it. And there was some stop starts and risk almost didn't make it. Those people are now running senior positions in the industry. And Kubernetes has been such a big success all around. I will get to that. And you and Stu were there, right? I think you were there with Stu. Weren't you there when Kubernetes was announced, like DockerCon? Stu Miniman and I, he had a red hat, former KubePoser, where we know Stu. Stu and I were having a beer with Lou Tucker and JJ, who runs OSS venture capital firm and the other guy from Kismatic. I forgot his name. And it was talking about the Google paper and spitting that out. And even before that, our first ever Google event, when the App Engine was around with Craig Mclucky, Brendan Burns and Tim Hawken and all the Google people were like the nerds working on this. I ended up going to their after-hours party and we're all sitting there drinking. They didn't know I was with SiliconANGLE or theCUBE at the time. And I just sitting there, they thought I was a Googler. I fit in. And they were riffing about how Eric Schmidt was killing the whole initiative and that they did not want Kubernetes to be another MapReduce. If you remember the MapReduce paper, Dave, they took that out and made Cloudera. They're like, we do not want to see the same thing happen to Kubernetes. We don't want some company to hijack it and then make a business out of it, which Cloudera did, which that was smart move, by the way, by Cloudera, because Google just gave it to them. So it almost got killed. Kubernetes was almost killed before it even started. So great little legacies in story, but 10 year anniversary, they had a big party, big cake. Really, really cool to see all of them. I would say, I would go so far as they say that if Google didn't open source Kubernetes and give that gift to the world, it would have been completely irrelevant in the cloud. We all know it's a distant third, but we'll talk about this, but the AI is changing the game. And so they're new game, new rules. So without Kubernetes, Google would have been just completely irrelevant. Well, the entire cloud-native industry might not have been irrelevant, because it basically creates the enablement for things like more container usage, serverless. Kubernetes is a really important technology that has made cloud-native really work at cloud scale. So you saw that really, really prominent at Google Next. I was really interested to see that, but all good. I don't want to get too fast into Google, but I wanted to make sure I had all my, all our topics for this podcast. Yeah, the Intel, TSMC, investment was big. The, Andy, we're going to start with Andy Jassy interview. What do you want to do? Let me just make a quick comment on TSMC. So yeah, they got their subsidy, but to me, the big thing, watching some of the vision stuff, you talk about Gaudi, they were comparing Gaudi performance to last generation NVIDIA chips. Even NVIDIA does some bench marketing. If we read the semi-analysis they just did, which was outstanding on Blackwell, even they're playing games, but here's the thing, IDC just put out a report, PC volume is going to grow 1.5% this year, okay? Intel, Intel's monopoly is from PCs. PCs peaked and volumes peaked in 2011. They're now back to pre-pandemic highs, only going to grow 1.5%. Why is that important? That means for Foundry to succeed, they're not going to get volume from their own x86 PC business. The only way they're going to get volume is if they steal major share from TSM and Samsung. So I think the probability that they're going to steal major share, enough share to drive volume, to drive their cost down, to get their yields up, is a very, very high bar for Intel. And I think it's a relatively low probability that they're going to surpass Samsung and get to a confident number two by the end of the decade, which is their goal. Well, what do you think about the seven nanometer announcement? I mean, Gaudi is starting out like a leg behind. It's almost like playing with one leg, right? So, compared to five nanometer and three nanometer and their benchmarks were comparing, as you pointed out, old NVIDIA benchmarks. So the not yet even released Intel chip is already claiming that it's better than a multi-year old NVIDIA chip, which when the new chip comes out, will be completely different from both nanometer perspective, energy management perspective. Just overall, everything's different. So is it okay to have a seven nanometer chip? Who cares? I think what I've announced is not, I mean, I can't, I'm not an expert on this, but I have been told by people who are that it's hard to compare Intel's process at seven with others at five or three that there are significant differences. I think Intel is trying to leapfrog, playing what Floyer calls as parlor tricks to try to catch up five nodes in four years. So I think I'll give a pass on that, but again, it just comes down to volume. How does Intel get volume? The monopoly is basically gone. The monopoly is gone. They no longer have a monopoly. You know, as a monopoly is NVIDIA, we can talk about that. I just think, I've said it before, I'll say it again. I think if your objective, if Gina Ramondi's objective is US access to onshore advanced chips, the higher probability is to invest more in those that actually are delivering those advanced chips, i.e. Samsung and TSM, TSM in particular, but both of those companies are investing in the United States. And so I feel like they're watering down the investments, our tax dollars, a little bit too thinly. I want to see Intel succeed. Don't get me wrong, it pains me to say this, but the question is, is it more important to have a US-based supply or is it more important to have a US-based manufacturer? And that's not necessarily an easy question to answer, but I would say the former is probably more important than the latter. Well, Intel has always been secret, recently Intel is very secretive. They're like everything's scripted. That's a red flag as far as I'm concerned. So I'm very skeptical on Intel right now. I don't have a good vibe on them. I see them, you know, groping for relevance, obviously getting lapped around the field by the competition and the TSMC, Arizona win, not only on the grants from the US government and but their own investment, they're winning. And whoever gets on their plate ends up probably having the best chips. Everyone's designing custom silicon. So the ARM Google announcement. So I think TSMC is continuing to kick ass and that's not good news for Intel. But since you brought up Intel and the policy side of it, the government subsidies or the government handouts, what should we call it? Government funding? What is that? What are they calling it on paper in the press release? The CHIPS Act is obviously a strategic move to bolster US manufacturing capability and supply of advanced chips, but it's very clearly a subsidy for Intel. But in fairness, they're spreading money around. Samsung's getting money and... Would you call it a handout? ESM? Well, they call it a handout. I mean, yeah, it is a handout. It's taxpayer handout. But I mean, I don't think it's the wrong thing to do, by the way. You know, but I will say this. It's not an entitlement. It's an investment. Handout investment, what do you want to call it? And they have to produce. And your question's a good one. Can they even deliver? But I'll say this, John. 52 billion, to me, 50 billion's a drop in the bucket. You hear numbers like trillions of dollars thrown around. The likes of Google, Amazon, Apple, who are all big, big meta, big, big players in this game, big, big consumers of chips. We're talking, they're trying to build, we heard this at Broadcom, million GPU clusters. Okay, these are big, big consumers of chips. They also make their own chips. All three of the big hyperscalers do. And so they have so much money on their balance sheet. They could easily write a hundred billion dollar check to bolster US based manufacturing. They could, and I think the smart thing for the US government to do would be look, we're going to dial down all the rhetoric around not letting you buy, you know, high robot. Well, let's get into high robot. And let me just finish that thought. And you know what? But you guys got to step up and start writing some checks for US competitiveness in the semiconductor business. So let's start figuring out a way to do that that is additive, you know, long-term to their balance sheets. And at the same time, additive to US competitive and doesn't necessarily take taxpayer dollars. They've got so much money, they don't know what to do with it. So they're buying GPUs. Anyway, we'll get into the GPU architecture later with Google next conversation because they did a lot that they announced. Let's get into Andy Jassy and Amazon and Amazon web services. First, Andy Jassy had a couple of things happen this week. One, he had the shareholder letter went out and then he was on CNBC for an interview and he was very vocal. And by the way, I give him amazing props for what he did. That takes leadership and that takes guts. Okay, and it was ironic at Google, the new head of public sector. First, within the first couple of minutes of the interview, she brought up China directly, volunteered, we are not doing business in China. This is from the CEO of Google public sector. And by the way, as Kevin Mandian on her board. So I'm learning a lot of new things. Look at that in a second, but this China rhetoric is huge. Andy Jassy basically said, and this is my takeaway. He basically said to the world, look it, we try to buy a company. We thought there was synergies and it got rejected in the EU and the US told us they would reject it too. And as, because two Chinese companies were the competitors and they were worried about us having unfair business practices, meanwhile, in order to do the product that they make, you have to have data on everyone's home, right? And so you don't trust an American company, but yet you'll trust these Chinese companies and he was blatant. That's, let me say abuse of the law. Yeah, so. Okay, that was massive. And I think that's leadership. People are throwing haymakers at him. That's hypocritical Amazon, cause always the haters of Amazon, but he's stepped up. Oh, that was 100% John. I gotta add to that. So basically, let me double click on it. So the, right, they blocked the iRobot acquisition. And so Andy's point was, when you have these Chinese manufacturers, these home vacuum cleaners, they have to map out your home, right? They're collecting all the data on your home and you're trusting them. And yet at the same time, you don't trust us and you're blocking this acquisition. It's really absurd. And then he did make the statement that they're overstepping their bounds and they're probably not following the law. I hope they challenge that in court. The problem is this, the current sentiment in this administration is to just pretty much block everything, which is, to me, it's just swung too far. So that, the other thing I would add, I put out a tweet on this, Andy Jassy did exactly what we said he was going to do with Amazon. Everybody was, people were calling for his head last, in 2022, remember? Oh, how long is Jassy going to last here? We were like, wait a minute, the guy just took over. You know, he took over and the market was booming. So they had to hire people like crazy during COVID. You couldn't even get toilet paper. So they had to like figure out the supply chain, hire people like, of course they overhired and then interest, Ukraine hits, interest rates start going up, the Fed starts tightening like overnight. So, you know, you just can't just turn around the steamship, but what they did was remarkable. Stocks at an all-time high today, they've got liquidity in the balance sheet. AWS is kicking ass, even though you read Charles Fitzgerald, you think they're about to go under. Fizzi is like out of control with his vitriol against anything that's not Microsoft, but he's funny as hell. But anyway, Jesse did exactly what we said he was going to do. He was going to cut costs naturally. He was going to make sure that the retail business started cranking again. He's doing media deals. Thursday night football is a huge win for those guys. He's got advertising revenue kicking in. He's taken over AI basically, helping Adam, you know, figure that out because it's so important. And Amazon is back with enormous runway. So kudos to him, great leadership, everything we thought he would do and more, John. Yeah. Well, I liked the fact that he stood tall and he said what we saw on everyone's mind. And he's also listening to the podcast because he basically said what we've been saying in so many words, but he did it in a way that was very respectful. He was very opinionated, but he brought up the entrepreneurial equation. And to me, what resonates in that, his little diatribe with the New York Times reporter or CNBC reporter is he used an example. He says, look it, that's not our model to go to try to screw people over. I saw some more stuff and yeah, we offered third party. But his real point was iRobot suffered. They didn't make it. They weren't going to make it. You had Chinese competition. We know in China there's no entrepreneurial action. It's all controlled by the government. It's all the whole system works as one. The state runs everything. So you have two companies that are scaling up and using all their anti-capitalism to have a competitive advantage over the American company. And so he basically saying, this is an entrepreneurial success story. They started a company, created so much value and that had competition and Amazon synergies would have put them into more competitive synergies on leverage on scale and supplier components, et cetera. And then they rejected the company lays off a third of its staff, stock price tanks. It essentially goes into a tailspin. And Dave, you know, we always say in the queue when the boat starts taking on water, it's hard to really make it work. And sometimes it's irrecoverable. You can't recover from that when the boat starts sinking or the plane starts going to a tailspin. So that was really the problem. And that bothers me because I think that's the government that is saying we would rather see entrepreneurship fail over some hypothetical scenario of a company intentions. Okay, that I have a problem with right there. So I think, you know, Jassy standing up for all the entrepreneurs out there and people who start businesses, it's hard to start a business, hard to run a business. And, you know, yeah, did they do anything wrong? Nothing wrong, it was a good deal. So, and then it's the China angle. And that's the whole point. The US government is so clueless on this. And the regulators. To your point, it's like the tail of the parable of the Chinese farmer. You know that where his horse runs away and they say, oh, sorry. He goes, oh, maybe then the horse brings, brings back, comes back with seven other horses. Yeah, that's great. He goes, oh, maybe. And then the sun breaks his arm. Oh, that's terrible. And then this, you know, you've, you've heard that parable, right? Yeah, yeah. And I'm talking about, yeah. But the point is there are these unintended consequences of these decisions that it really hard to predict. And so the, okay, great. So you hurt an American company. You're essentially helping these Chinese companies with your public policy. Is that the outcome that you're intending? Really? And so I just, I feel like it's really hard for the government to predict the impact of their actions. And, you know, who knows, John, maybe AI will be smarter than the government soon and they'll be able to figure it out. I'm really disappointed in the public-private partnership and the discontinuities between, on the one hand, we want to be more competitive and you hear, you know, hear things like, you know, tariffs against China. And on the other hand, we're just constricting business, trying to do regulatory capture in certain industries, which doesn't really work out that well. But yeah, to back to Jassy, he was stellar. And I thought, I gotta, I have to say, I thought Andrew Ross-Sorkin was excellent, but that he did a really, really good job. He does, he does good interviews. He consistently asked the right questions. He's obviously got his homework, which is, which is awesome. But, you know, the Lena Kahn at the FTC, the problem is, is that, you know, in the AI conversations, not to bring back into AI, because like I said on theCUBE, AI drug these days, everything's about gender AI. We have an era where there's master societal change going on and shift. And people like Lena Kahn, who's at the FTC, really is not plugged into what's going on in the world. And because technology is not this evil section of people, or section of industry where there's evil people and there's screw people over, it's just, it's becoming pervasive in everything we do. So the number one conversation with gender AI is societal change. Better government, be a better government, be a better, provide to your citizens, better services to streamlining antiquated systems in the public sector, to offering entrepreneurship opportunities, and also democratization, Dave. You know, AI is a technology that will, if handled properly and not regulated and killed by narrow-minded people like Lena Kahn, is to say, hey, you know, people can level up. You can have someone with a degree in Harvard and someone who never went to college can level up within months and be peers. Okay, we're talking about the biggest democratization opportunity in our history. So the whole idea of underrepresented minorities getting a fair shot, getting access, these are things that we work on, the opportunity is now for that. And I think, you know, the stuff's changing so fast that you're having some pedigree or having some status, elite status or privilege is could be at risk tomorrow because the opportunities are everywhere and the ability to capture those opportunities is available and the skills to do it are the same for any people. So to me, that's a big problem. And so my rant is that the regulation of AI, the other side of the coin of safety is what opportunity costs are being blown by trying to regulate something that's not well known. You don't know the outcome. You can't make optimal decisions. It's like a startup. It's always hard and you got to keep iterating with the decisions. You can't figure out unknowns that are evolving. So. I think the problem I have, the biggest problem I have with Lena Kahn is she's trying to rewrite the law through her public policy, at least how I interpret the law. The history of the DOJ and the FTC meddling in tech is frankly very poor. That markets generally moderate the technology monopolies and market forces have done a much better job in my opinion than the government. Now, having said that, if a monopoly is colluding or they're bundling or they have practices that hurt consumers, that's breaking the law. But essentially what she's doing is saying, if there's a potential for those things, we are going to declare a monopoly and then we're going to take action. So there's a big difference between, these things are actually happening and they have the potential to happen. And I think that she's conflating them. And I think that's where Jassy is right that she's overstepping her legal boundaries. And it's, to me, it's very disappointing. All right, well, let's get into while we have Andy Jassy on the docket here. And by the way, we should ping him and get him on the podcast. I'm going to reach out to him and see if he comes on and kind of have a Elon Musk moment for us, how he made Joe, remember Joe Rogan and had Elon Musk on and everyone went crazy. So we should try to see if we can lure Andy on to do us a favor and come on and share his thoughts. Let's get into his dear shareholder letter. They had the annual letter. It's a tradition that Jeff Bezos started. Now that Andy's the CEO took it over. You read it, would you think? Things jumped out, obviously the numbers are incredible. But they overspent in the pandemic. He told me that when I last time I met with him in November with hockey game, he said, yeah, we spent a lot. We built up. We didn't know it was going to end. Well, plus they were responding to customer demand. I mean, what are they going to do? Not respond? I mean, he was looking at a shareholder letter. It was really upbeat versus the last year, you know, hey, we got some challenges and we're going to take them on. He basically did everything he said he's going to do and more. To me, it's just, it was, people are saying they're cutting back though. That's my point. My point is that they got to cut back. They have to. Yeah, yeah. I mean, people are always going to complain at those things. If they've cut back and cutting back on CapEx, which of course, you know, Charles Pence is like, oh, see, they're in trouble. I think, you know, to me. If you overspend, though, if you overspend, Dave, then, you know, it's like, right, you got a lot of back. So, so to me, just the, just briefly, the combination of what's happening at AWS or Amazon, you know, retail and AWS, you know, coming into play. Now, AWS has its own challenges that we should talk about that, but their businesses are humming and that's why the stock's at an all-time high again, which we were saying when the stock was getting beat up and everyone was like, oh, wow, Amazon's done. Hmm, could be a buying opportunity. Again, we don't advise stocks, be careful, do your own research. We're not, you know, stock advisors, but, you know, we do watch companies and we watch fundamentals and we know good management when we see it. The one thing I would say to AWS, I think it would be really interesting to see if they can accelerate growth as a result of AI. We should talk about their AI strategy. He did say, he said now a number of times, and he said in his letter that cost optimization is attenuating, has attenuated. That doesn't mean it's gone away. And I tweeted out some data, cost optimization is like muscle memory in golf. You know, you hit that shot, you know, I don't really know that well, but you know, when you hit that shot, you're like, okay, I got this down now. And it becomes like your best shot, your go-to shot with your seven iron, you can always hit it. Like I've seen you hit it, boom. I'm like, you can't miss, it's unbelievable. So you got that muscle memory. Well, I think that's what cost optimization is like now. They, the IT buyers and Amazon customers, they have muscle memory on cost optimization. They say, hey, we know how to do this. Let's do that. Now, a lot of people think, oh, that's bad for the cloud companies. It's not bad. Here's why. What they do is then they save money. This market's always been elastic. I've just never had the data on cloud, but now I do, and I'm seeing it. When prices go down, when cost goes down on something that's valuable like technology, they spend more. And that's what happens. They'll invest more and they'll invest more for innovation. So when I've seen this for years with storage, cost per bit would go down. People would buy more, why? Because we can store more. We can do more data and analytics on that. It's the same things happening with cloud resources. You're going to see the same thing happen with, with GPUs, costs go down, people will consume more. Now, is it a straight line up? No, it goes in step functions and it goes up and down. It's not always a pretty straight line. That's how markets work. And so he did call that out, but I think that there's a gain sharing going on with that muscle memory, meaning people are taking money from their savings and they're putting it into AI and they're also taking it from other places and putting it in AI. That's clearly what the data shows. Yeah, I agree. And for me, what jumped out at this letter was, it was a very long letter, it's on their website. They've been cutting back, but their delivery times are up. That's a big deal for Amazon is delivery times. The emphasis on prime, again, with the original content is core. You mentioned Thursday Night Football, they got programs, they had that successful Boys in the Boat movie. That was phenomenal. Just hit, that should win awards. I thought that was a phenomenal movie, but their advertising business day was growing. So you don't think about Amazon as an advertising company. So what you're going to see is, I think their whole TV offering is going to be disruptive. So let's keep an eye on that. I remember I had a one-on-one with Andy Jassy for a pre-reinvent briefing. And he's like, why do you love Twitch so much? And I'm like, well, first of all, Twitch is huge, it's fun, it's gamers, but the numbers are massive. And then that was the beginning, they started getting into the, now they got Free V, they got over the top. So you watch their free media, remember they bought MGM, so MGM originals, all that stuff. AWS, we cover that like a blanket, so no need to go in there. The key's going to be there to see how their custom chips, will they have a lead on performance and will they drive that home? But I'm really fascinated by the video, prime video service. That really got my attention. So advertising and prime, they're in the entertainment business, right? And they went out to hold the vision, talking about primitives and primitive services, and then the math. That was interesting, the primitives conversation. So he applied that to Amazon.com. I found that really, because we've been talking about primitives in the cloud forever, for a decade. And we used to say, Andy, why don't you guys do more solutions? And he remember on the cube and he said to us, look guys, here's why it gives us optionality and flexibility to move fast. And he was talking about primitives in the Amazon.com business. I loved that conversation. Well, you know why they're doing it. First of all, this is my speculation. No one told me this. I'm just connecting the dots. It's an Amazon thing. This is an Andy fingerprints are all over this. What he's doing here, he's cleverly tying together Amazon services so he can tee up that we have technology on Amazon. Because remember, Andy's been very vocal about, we've been doing machine learning for years. Now, everyone says that now, David. You know what I'm saying? Everyone's like, oh, we've been doing machine AI for years. Before it was all hyped up. I mean, that is- We even say that, John. We use that. It's true. I thought about AI once. I programmed in Lisbon in the 80s. That was AI. So I think what he's doing here is saying, look at Amazon as a company, including AWS is using AI everywhere. And then they have been using data and machine learning. So this primitives is basically saying, our cloud has been serving our existing business for since it started. I mean, remember Amazon started, Amazon web services started with one customer. Amazon.com, okay. And they grinded their way through and iterated through and as a customer of themselves, Amazon web service became the behemoth that it is. So this primitives is basically saying, this is how we think about our business. And I guarantee the next step is going to be, tie that to machine learning and AI so they can go back to tell the world, hey, we are better than Microsoft and anyone else at AI. Now, Google is going to be a good company to watch when it comes to, is Amazon's AI better than Google? Because- Before you go to Google, can I just say one last thing about Lina Khan? I just want to say something about Amazon. So the reason I get so pissed off about this is, I think about Amazon. I think I'm like a loyal Amazon customer. Somebody told me one time, years ago, you're a Prime member? I go, yeah, of course I'm a Prime member. They go, do you know that on average you spend like 20% more on products, blah, blah, blah? I'm like, okay. They said, would you ever give up your Prime? Like, no, never. My point is, I have choices. I can buy online from Walmart. I can buy online from a thousand different places. But I personally trust Amazon. I know they're going to deliver. They're going to deliver fast. They freaking deliver on Sundays. If I have a problem, they fix the problem. You know, if I got the wrong thing or it's damaged, I send it back. They're great to do business with. So they're not hurting me as a consumer. I love the way they've helped my life. They've driven so much convenience to my life. So when I see all this vitriol against Amazon, I'm like, look, you don't have to buy from them. Buy from somebody else. Drive to the store, whatever you want. But, you know, anyway, let's talk about Google. No, no, I was segwaying from where Google about the AI. Because Microsoft, I mean, Amazon, and Amazon Web Services, Amazon Proper, they have done a lot of technology. So like to call them like falling behind is ridiculous. Now, Google has the same problem that Amazon has. They don't get credit in mainstream for all the work they've done. I mean, if it wasn't for Google, mostly I'd say almost 90% of the AI success is probably due to the route back to the lineage. It's Google people. Google and Meta's trying to take credit for another. They hired all the Google people. So Meta basically is a little mini Google in terms of the engineering prowess. It certainly didn't come from Zuckerberg. He wasn't like known for a great engineer. He's a great product guy and he can code PHP. He's not the killer deep tech engineer. They've been poaching Google. They live in the valley known public secret. Google DNA AI route goes back to all the papers that were written. They've been doing it for many, many years. And it's in the company, Google search and all the scale and everything they've done like Amazon. So at Google Next, you saw them bring all that out. You saw Google Flex and I said this many years ago, Dave on theCUBE, remember we had a big session on this. We said, if they can get their act together and go back into the back room and get all the stuff and have someone organize it and just this department pull it from there and just get it into one coherent thing, they'd smoke everybody. Remember that conversation? It might have been like three years ago. And when everyone was like, why isn't Google winning? They have large scale experience. They know how to deal with data and massive scale. They basically invented the SRE concept and Borg, which is now Kubernetes is they have all that. It's all coming from Google. So Google doesn't get enough credit in the industry. Insiders know it. In fact, there's a lot of poaching going on from Google. Either people or the ideas. I have to open AI people used to work at Google or came from the DeepMind team. And so that's the creative process of Silicon Valley. The fertilizer is the people and the technology. So Google's a big, been a big part of that. So I think Google and Amazon are going to really flex their AI. I think Microsoft right now is looking good. If you look at the, from a leaderboard standpoint from perception, but I'll tell you right now the conversation I've been having with people in the trenches, startups, no one likes Azure. Okay, they like Google. No one complains about, have you ever complained about Google price? No. No, not really ever. I have data on this too. I want to share it with you. Yeah, please share. Cause I'm telling you right now, they can make a really power move right now in terms of moving faster up the leaderboard or from the third place challenge position to a leadership position. And I'll tell you, they have the chops and they're bringing their stack right now. So I just want to say something about, so when Frank Slutman took over for Snowflake, and I think he might've written about it as a book, but he told me this company, Snowflake, they had product market fit like for years, all they had to do was figure out the go-to market. So I came in, I scaled go-to market. And you know, you don't want to scale go-to market until you have product market, but when you get product market fit, you want to scale your go-to market. Well, Google's got product market fit, but they can't figure out the go-to market and the service model. And they've struggled with that, right? Rob Enslin was kind of running sales at Google, and now he's at UiPath. And so they'll eventually figure it out, but the data that I have suggests that AI is really allowing them to close the gap. Not so much broadly in cloud, but let's talk, let's talk narrowly in AI. Okay, so when you look at the ETR data, and they have a very consistent time series methodology, which measures spending momentum on a platform and spending momentum, it's only measured by percent of customers that are doing something. So it's just a percent customer measure. It's not a actual dollar measure, but it's called net score. Net score represents the net percent of customers that are spending more on a platform. So they ask a bunch of questions, are you adding new, are you spending more, are you spending flat, are you spending less, or are you churning? And they take the lesses away from the mores and that's net score, okay? And anything over 40% is exceedingly good, okay? So I just ran before this Q pod, I ran numbers and I ran it back to January of 2022 for Google, AWS and Microsoft. And the net score in just the machine learning and AI sector, the net score for Google was 61%, AWS 60%, Microsoft 71%. That's back in January 22. Now the other thing they measure is how many people actually respond because the more people that respond, it's an indicator, a proxy for install base. So Google was at the time 93, AWS is 151, Microsoft 175, okay? So all three had very high net scores, a lot of spending momentum, but relatively small ends, 93, 151 and 175 for Google, AWS and Microsoft respectively. Now fast forward to April 24, the survey that just hit guy, it's not even, it's kind of quasi-public. Net scores, Google 56%, AWS 58%, so pretty comparable, Microsoft 74%. But check out the ends, the responses. Google 340, AWS 370, Microsoft 611. So what's happening is, let me explain this, what's happening is Microsoft has stayed elevated in terms of customer percentage, spending more, and even actually accelerated a little bit. Their numbers of N went through the roof because of Azure API, getting open API access. But Google and AWS have come together in terms of those Ns. It was 93 versus 151. Now it's 340 versus 370, so you can do the math. But when you look at the data over time, and they have these graphs that, you know, they use Power BI, which kind of sucks, but it doesn't, you can see these coming together. And Google, in ML AI, Google and AWS really on top of each other. So that says that Google is in the AI sector, really catching up in terms of customer momentum and customer install base. So to break, to simplify that in my mind, it's a lot of data there, I was trying to grok it. So is it, is the gap closing? And if you can scope it, scope the gap between two and three here, between Amazon. How would you kind of scope that? And if you can scope between one and two, that'd be great as well. Are we talking about a horse length? Multiple horse lengths, you know the horse racing. Yeah, so lengths. So if you go back to January of 22, 151 over 93, so there was 1.6. So Google, AWS had 1.6 more installations in this survey than did Google. And now when you do the math, it's 370 and 340. So it's 1.08. So basically, you know, you're talking about 60% lead down to an 8% lead. Okay, so, you know, if it's a horse race, the plotter at the back of the pack, coming down, you know, the far turn, it's here comes Google on the outside. And Google's going to the whip, you know, the right-handed whip, as they say. So they're in position. They have a shot of coming up faster than you think. So the data says no question about it. And of course, Microsoft just, because they're everywhere. Microsoft, let me say this about the Microsoft interpreting the Microsoft data in these surveys. They're so ubiquitous. They're so large and they have so many customers. It's, they just, they overwhelm the survey data. Whereas, you know, Google and AWS, it's a much more sort of apples to apples comparison. So, given that, I would say, being at Google's event in Vegas this week, I would say that it's very clear that they got all the package together. BigQuery actually now is a data platform with the vector embeds and the capability and all the multimodal reasoning is looking really like a viable platform, scalable. I found that was my, that was my favorite little nuance point is that BigQuery with the data there is really set up well. And they have tons of stuff around the analytics and databases. The Gemini product is phenomenal. I actually did a little test. I said, is SiliconANGLE an influential publication? Because everyone's trying to figure out who's influential. Of course, the analyst relations departments at Google's struggling with that as well and others. And it really, it said, yes, they are extremely, and the answer wasn't generic stuff. It was specific to our business. It was really, really good. I was blown away. Of course, I did the vanity search, you know, who's John Furrier and then gives you who it is. And they said I worked at Oracle and VMware. So, stellar career at Oracle and VMware. No, IBM and HP, they got it wrong there. Close, big corporate company. And then I started putting in stuff around things that I knew about in the industry. And it gives you a good answer. And then it gives you reasons why. Evidence. So that's perplexity that does the same thing which I like, and we do it with our RAG system. So I think this multimodal thing is huge, Dave. This cross modality reasoning is going to be the real AI. And I think like our last conversation on theCUBE which people loved, I thought, I'd tie it here, is that during the early days of the web, you had web pages and you had search engines. Here in AI, you're going to have big power, highly, highly capable smart things, mechanisms, not just a search engine, but like technology that's built by people who are really deep and deep tech engineering. And then you have users of AI like the web page. So it's very clear to me at Google is that that's happening. And the consulting businesses are booming as well. And what's happening there is that unlike the old IT transformation projects, the new consulting system integrators are doing coding. They're actually engineering stuff on behalf of the customers. So that is a major differentiator for the ones that have engineering talent because the companies need to move fast with AI by board pressure. And so they're putting pressure on the management of these companies saying, where's our AI straight? We need more growth. And they got to hire people to build it. So we saw this at the web too. Remember the old interactive shops? I'll build your web page for you. Very similar concepts going on here. So, you know, Google is going to be a supplier. They have all the shops. The one area that I was impressed on was still I feel is going to be an opportunity to keep an eye on for Google is their ecosystem. They would do a lot of people with it with booth, big names. But I think they still need another year to make sure to validate that. So I think Google should make sure this year their focus is 100% on making sure the ecosystem doesn't fail. They have to make the ecosystem work by driving business through those partners, helping those partners and then getting business for themselves as well. So if the ecosystem fails with Google Cloud, it's they'll never catch up to number two. They'll be, in fact, they'll fall away in the distant third. And so I think that's my takeaway. And I think the partnerships are good, but there's a little bit of cognitive dissonance in the air with the partnerships. It depends on the layer of the stack, right? Again, this is something else Slutman told me. It's like, you know, Google actually wasn't Slutman. I think it was Scarpelli said that Google, that really doesn't, that friendly doing business. Now that was with Snowflake because they get BigQuery and BigQuery is awesome. And it's really head-to-head with Snowflake. But I wanted to go back. I just got to tell you a quick, make you laugh for a second. Before Gemini, when it was barred, I asked, I was vanity, you know, barting and searching like, who's Dave Vellante? It said that I started Silicon Angle with Jason Calcanus, which is hilarious. I mean, you had a background with Calcanus, right? You guys used to work, you know, kind of. He was a producer at PodTek, my podcasting company. He did podcast. Maybe that's how it connected the dots. Who knows? He's a prolific media person. I mean, Jason has roots, but his newsletter, he started, that's how he got in the business. And then he started WebLog Zinc, which was the original. And Gadget was the first real tech blog that brought it. So maybe it was confusing you with Calcanus. But then the second thing I wanted to say is. Easy to confuse. So I was in Texas this week. I was in Houston, saw HPE and saw Fidel Maruso. Pat Osborn was down there, although you can connect with them. Jim Jackson, I went down to see him, but he was busy. So I just kind of waved Jason Newton and guys. And then, but then I went to drove over to Austin and I met with the CFO of Yvonne and Rob Williams. And so, and I saw Michael, which is kind of interesting, but I also saw Matt Baker. Michael Gill, you mean? Michael Dell. Yeah. Yeah. And I saw Matt Baker and we were shooting the shit like we always do. Matt. Our favorite Matt Baker. Are you Matt Baker? Yeah. Yeah. AI Matt Baker. He's all over the whole rag stuff. He loves the AI game right now. He's really been digging in hard. I love it. Oh my God. I love his LinkedIn posts are phenomenal. He's unbelievable. I mean, so Matt Baker, for those who don't know him, he used to run strategic planning at Dell. And then when the whole AI thing took off, he said, okay, Matt, they hired, or they brought back Vivek Mahindra to take Matt's job and strategic planning. And I said, Matt, you go figure out AI. So now he's like charging ahead with AI. And every time you meet him, he's like, like, I think the last time I saw him was November and we were talking. And then since then, you know, he goes out and does all this research. He's doing a bunch of internal stuff. He's learned so much. Basically, I'm talking to the CFO of Dell, Yvonne and Rob and talking to Matt Baker and certainly Michael. But I'm convinced that Dell, because they're like, you know, nearly a hundred billion dollar company, they're doing some really interesting stuff in AI that could put them ahead of a lot of their customers, which is rare for Dell, right? Usually Dell's, you know, they're selling boxes and they're going into customers and the financial services guys are telling them what they need. I think the reverse is happening where, you know, Dell's learning a lot because they're investing a lot. They have such a big company. They realize, wow, if we drive AI, we're going to drop money to the bottom line. So doing a lot of cool stuff with RAG. And Matt's right, we were talking about a lot of customers haven't figured out how easy RAG is to do. And, you know, he was pointing to us. He goes, well, you guys did RAG, of course you were like way ahead of the curve. He was pointing to us. But he asked me a really interesting question and I wanted to ask you because we were talking about Google and I'm saying, hey, how about AWS? They're going to have Olympus and Microsoft's got open AI and the data suggests they're running away with it like we just talked about. I said, yeah, but, you know, there's a lot of really interesting stuff going on with Lama. And you said about this when we talked about the gen AI power law. He said, don't you think that open source is going to win that all this proprietary stuff is at some point, the customer's going to say, well, wait, why am I going to pay, you know, monopoly prices when I can get just as good a functionality for, and I could do mass customization with open source. What's the question? Don't you think open source is going to ultimately win all these closed systems and what's going to happen to these closed systems? Anthropic, Olympus. Okay, so this is an, let me just, one more thing. It's a really interesting question I've been having with leaders. Like, will these LLMs become commoditized? One school of thought says, absolutely, we talked about this last week. The other school of thought says, no, they won't. So then there's that dimension and then there's the open versus closed dimension. So piece that all together for me. Well, we talked about last week on the pod about commoditization. I thought you agreed with the thesis. It's sometimes when people say commoditization, they're talking about something different. So when something's commoditized, what does that mean? That means it's a common thing. It's not as expensive. It's cheaper. It gets lower in value. It's everywhere. It's commodity. And then commodities change based on scarcity, right? About volume, what abundance there is, demand is blind. So I think with LLMs, these foundation models, that world is growing and changing so fast that commoditization is not really the question because there's so much growth and value in these models that they're changing. So the question is to ask is who's obsolete? Who's not? Who's better? If someone's not good at their LLM, they're not going to do business. Although they'll have a bad experience. I think our power law research that we put out where you have that power law and at the top of the power law and volume size of model is definitely where the big players are. And they're called, we call them proprietary. That's the original name for them. Some call them pioneering models, but they don't cause they don't like the word proprietary because open has AI is open in their name. So like they don't want to be labeled as proprietary because that's pejorative. It means- Sam Altman said he wished he didn't take that name. Did you hear him say that? No, I never, of course. He said in retrospect, I wish we didn't call ourselves open AI. Basically saying we're not open. Of course everyone knows that, but that's the whole point. That's just optics. That's just, that's like- Yeah. That's just like in the, our computer industry, being proprietary means you're owning it. You don't want anyone else to get it. You don't, you just, you're selfish and open source software is open. So opens, then the question that Matt's asking is will open source win or will open win? Open always will win and not. We are, we are like, it's like, it's like, I don't know, I shouldn't say this. It's like, I have to say it's like the, it's like the right to choose. Okay. You know, that debate that goes on. So, you know, women's rights to choose. You know, we are in, we can't put that back the way it was. It's has to, it's never going to change. I think so. I think that discussion is a pretty political one. It's happened in our business. It's become political. Proprietary is bad. Okay. Open is better. Open will always win. You can be proprietary and the norm, the normality of privacy, proprietary is I own it land. It's proprietary. You have a deed on your house, right? You have land, you own it, right? That's proprietary to you. That's not a bad thing. That's just called ownership. So ownership of something is bad. So the word proprietary means I built it. I own it. Means no one else can co-own it. You could use it and charge your fee for it. That's not open source software. So that's where the proprietary name got associated because it was the defaults conversation. Open AI built their own model. They own it. They spend millions and millions of dollars to do that. And that's that where Matt Baker is getting at is that open source models are getting just as much traction and functionality and adoption as those proprietary funded for business companies, which means it's not as expensive and it's open. So that's a good thing. So that is going to put pressure on the big guys that have spent all that money to do one of two things, make it commodity in the sense of, okay, it's commodity, use it. It's not as good as anything else. It's got hallucinations and I'll make as much money as I can till it dies. It's a race to the bottom. Or they can say, let's pour more money into it as a proprietary business and make it better. And that's what they have to do. So when we were talking about, I said, if anyone's going to be commoditized, it's going to be the big LLMs. That was my answer. Matt would agree on LinkedIn, he pinged me. So the big guys are going to be proprietary because as you go down the power law, the smaller models actually are around people's data and that's proprietary and that's not bad. That's a good thing because I own my thing. Now open source software is going to adopt open models that can be programmed with. So that's kind of the way I see it. And I think what Matt Baker is smart about is is that he knows that staying open will give them the best price performance because open source tends to be free. Now what will happen with these open source models, just like an open source, a company will form around it to support it, to provide what Red Hat did. So we used to ask, is there a Red Hat for the cloud? What was that we would always say? Is there a Red Hat for- For big data. For big data. Well guess what? There could be one coming soon. I think that was a no a decade ago, but there could be a Red Hat for AI in the sense that if open source continues to get more popular, if you look at hugging face, look at the leaderboards, Mestral, Lama, all the open source models are doing great. Why? Because there's a huge appetite demand for them, Dave. So if anything, that's kind of, I guess that's commodity, because it's free. Well, I think, I think- Open source is- But I think to your point, as long as there's innovation, there's probably not going to be commoditization. I think Matt's other point was, look, it's all going to be, it's about mass customization. And open source is going to support that mass customization, you know, at massive scale. So I don't know, it's an interesting- Let me ask you a question, because you, you know, because I saw some of these fake analysts out there, talking about stuff that they're paying customers. You know, one of them said NVIDIA's a monopoly, or is it a monopoly? Is not a monopoly? Oh, I was going to say, yeah. Let me find that tweet. NVIDIA's a monopoly. NVIDIA is, of course, NVIDIA's a monopoly. I mean, the question with NVIDIA's, not whether or not it's a monopoly, it is. I mean, it's got, where else are you going to get, a GPU, the class of what they make? You know, you saw Intel trying to do, you know, Gaudi and doing some benchmarks against NVIDIA's previous generation, but Blackwell's not out yet. Do you think NVIDIA's a monopoly? Of course. Yes, absolutely. I mean, they've got, you know, over probably 80% share of the high-end GPU market. Maybe it was closer to 100%. They've got complete pricing power. They've got 77% gross margins. They've got a one-year backlog. If you don't write a check to them for 10 million bucks, they're not even going to return your phone calls. You're never going to see a GPU. So that's a monopoly. So the question is, are they going to keep their monopoly and how are they going to keep the monopoly? How long is it? I saw this consortium that's getting together that said they'd have sort of Cuda equivalent functionality by the end of the year. I'm like, wait a minute. Cuda was first shipped like more than 15 years ago. You're not going to compress 15 years into one year. Now, maybe they can more narrowly apply it and compete. I mean, the whole industry wants competition for Nvidia. I want to see competition for Nvidia. But, you know, they're packaging the Melanox move that they made years ago. It's going to take a long time for people to unseat that. We know, we sat through Charlie Cowess' presentation at Broadcom. We know what their strategy is with Ultra Ethernet and where they're winning with connectivity. But it's going to take a long time for the XPU folks to catch up, in my opinion, to Nvidia. I think they will have a monopoly for at least, in this market for at least five years, it could be longer. It could be a 10-year monopoly. I just find it's embarrassing that analysts who get paid by Intel and the other companies saying it's not a monopoly because they want to give hope, I guess, to the- Is that why they say it? Because they're trying to help Intel? Well, they're either incompetent or they're being paid or both. So, look at, but that's- That's why Intel doesn't invite me to their, I said, Pat Gelsinger, I love Pat Gelsinger. We know each other. We always had great interviews. We've had debates and arguments. You think that's why they don't invite me to their stuff because we've been so right and critical of- I don't think they know you exist. You've got to go knock on their door. I mean, they're closed for business. Intel's closed for business. They're hiding the ball. When you don't have anything, you close the doors and you put window dressing around it. So, if they're not inviting you to their analyst meeting, they don't want you in the kingdom because you're going to probably discover stuff that's bad. So, that's- Independent analysts do that. So, they think you're already briefed. They don't want to waste a seat. If someone else, that's not up to speed. They want to train someone else. Or they're just not competent. I don't know, there's variety of reasons for that. But we're clearly covering it. You're covering it. So, the reason why is that most people don't do things because they're either have dogma or some other religion around how they do things or they don't want smart people in their camp. Okay? I remember having this conversation with the top Amazon people years ago that were like, hey, we want to work with you because you know the most. And we want you to understand our stories so that all these people throw in haymakers out on the internet, you can at least be objective and tell the truth when it's time to like step up and say, no, no, it's not true. Because when companies have conjecture and make statements and hire people to be a mouthpiece for that conjecture, okay, and they're in the media, then what happens is customers will think that's real or not real, and then there's credibility gap there. But if a company says they do X and customers think that's not true, then you don't want the analysts that talk to customers in the media saying, wait, you're lying. It's like, what you're saying doesn't match what customers know or think about what you do. Or worse, if the third party data doesn't back it up. So I've found that when you have strict command and control PRAR teams, the ones that aren't talking means something's wrong. And that's classic marketing deal. You don't market when you're not good. And that's what's going on, I think at Intel. They're not doing well. They're licking their wounds. They had a big loss. You know, as I said this last week, and companies that don't want to amplify bad performance. Yeah, well, there's plenty of data out there on Intel. It's easy. I mean, I watched the keynotes and the presentations and I worked with Floor. Floor and I are doing a breaking analysis soon on just doing one of our semiconductor updates. And we'll address this. I mean, personally, I'm not going to chase Intel, but I'll chase Nvidia. I want to spend more time with- My philosophy is that they probably don't want to, shouldn't invite you because you're already briefed. You're already on top of it, you know? So maybe they feel they don't need to educate you. I don't know. I don't want to get into that now because there's so much that this whole analysts ranking stuff has been a real black eye for the industry. And- What do you mean? What do you mean? You mean those sites that ran? I had a conversation with the CEO of AR Insights yesterday morning with a bunch of other analysts firms. And I was saying, hey, I see a lot of bad behavior, people using rankings and gaming the rankings so they can put that into their contracts. And so that's not how the industry works. So it's not being monitored. It's an honor system in the industry analyst world. You know that more than anyone, you're even an analyst all your career. I'm just new to it and like, you know, this is weird, that guy is not number one in his analyst way. He's not number two. I didn't even know what this thing was until somebody mentioned it to me. And I was like, am I on there? And I, because I write, I write, like I drop research every week. And so I emailed those people. I said, hey, do you count my stuff as a blog or as a research? And they said, a blog. And I'm like, well, what's the difference? And I'm like, look at what I write. It's like data, it's survey work, it's analytics. And they're like, well, this is research. I was like, okay, well, you should count as research. All of a sudden I hit the list. I don't know, wait tweet more, I'm on it. No, the list started as a member, member of the cloud ranking. Yeah, yeah, cloud ranking was a joke. It was good at first. Then it'd be like, oh, number of followers make your reach. I'll just buy more followers. Well, that's what I'm saying. You could just game it. You could easily game it. But a lot of these lists start as promotional elements. So the AR Insights company has a business model to service analyst relations departments. So what they do is they have software that they sell and service their core business to analyst relations departments. Meaning helping them with things like organizing news and having like a SaaS service. It's actually a good business model. I actually like it. I think it's important for the industry. And the leaderboard came in as more of a way of a list. Here's some top analysts you could work with. So it ended up being a list of the people who are the best for getting paid, who will pay to get work. So if I'm looking for someone to build a deck on my house, I got the yellow page of who's the best cheapest deck builder for me. That's highest quality for me. And that's what this list is. It became a list of people who I can hire to market. I can pay them for services. Not who's the best analyst. It's who's the best to get taken cash and doing whatever I want. That's what it is. So I told them that lists is hurting their reputation that they should kill the list or recreate the list and recategorize it. And look at real research. And then they said something funny. You'll get a kick out of this. They said, and it was legit comment because it was an old school kind of mentality. Well, the media citations are critical. I go, okay, so if I put my name on every Silicon angle blog post, would I, we are media? Well, also how many times they're on TV? Well, we are TV. So clearly the metrics are not there to do these lists, Dave. So there's no data. And so what turned out as a marketing opportunity for this good company, AR Insights, actually made the list more like PR Insights. Who will get paid to do public relations for you? So I think they're going to fix it. The CEO is very seemed, very cool. Their community person is really, really nice. I really liked her. She's really friendly and very smart. And they're listening. And I got to give them a lot of props. They listened and they're going to make changes. So AR Insights is going to make changes to their algorithm or maybe even discontinue the list. I told them they should do like what Net Promoter Score does, do surveys. Yeah, I mean, I don't know. I don't really understand this though. I was just looking at the list and I was like, some good analysts on there, but I don't know. I don't know how it works. I don't really care. I just, I mean, I want to keep doing good work and serving our audience. I won't name names, but being exposed in our area of the industry, I've been 13, 14 years going to events and being involved in all the analyst relations programs out there and PR with theCUBE. We see all the analysts every day. And I can tell you, I personally can spot that there's been a good analyst and a bad analyst out of the box. And then over time you say, consistently you see the winners. Like, you know who the players are. It's like people who put it on a good game on the field. It's like sports, you know, you can see the athletes, you can see what they do. And you're like, wow, that person looks good coming out of the tunnel, but doesn't do well on the field. But that's in obvious. And so again, I think the industry's going with a lot of change. Media in general, Dave's going through a lot of change. And I'll tell you right now, what's going on with theCUBE and SiliconANGLE is really exciting because as we get more TV and the AI clips coming out and as we get these events, more blanket coverage with team coverage, the performance of the engagement is off the charts. And the content coverage is phenomenal. Google Next, just as an example, I know you were traveling some of these NDA meetings in the industry, but I'll tell you, we had great production, we had great storytelling, we had great news analysis, we had great news writing, Rob and his team. We had 10 stories on our site before I even did one interview after the keynote, 10. And then I think we had now over 30 to 40 different stories on there and more is coming. We're going to have over 500 clips. When you get the right team and focus on these events, it's really amazing how in real time, the ability to acquire, the ability to program and distribute the content with digital is phenomenal. And the combination of our multimodal capabilities with SiliconANGLE, theCUBE and CUBE research and the platform is really coming together. I got to tell you, it's so exciting to see because it's like watching performances go up. Multimodal is the hottest thing in AI right now. And we just got lucky that we do multimodal media programming. So I got to tell you, it's super exciting for me and looking at this, because we think about this for years, right? We're like, okay, we should do more videos. We should do more research, pull it all together, build a platform. Well, guess what? We did. Google, we did. And Google now has tools for everyone else to do it too. So, and it's going to make us and everyone stronger. So I think we're going to, you're going to see a whole new media landscape. I think this whole digital subscriptions will continue. Like the information does great work. I see those reports, they're phenomenal. But that's not the only game in town. I think you're going to see a mix of new monetization models kick in on scale, high quality, truth telling, storytelling, news analysis, getting people featured, personal stories, customer case studies. I mean, customer case studies was always viewed as a, in media as, eh, it's not really news, but on digital, when you feature a customer success story and you tie it with news, it flies around. It really distributes because the people are involved in it. You got the person who's the customer, they're seeing story about their environment. You got the people who deliver the solution and make the products involved. So social networks allow you to bring this neural vibe to the table, this network effect that has people involved who amplify their own stories to the good content spreads around. So I wonder if we should, I wonder if we should talk about that, this, that our business and how it's changing on one of the cube pods. You know, Rafi Myerson, Rafael Myerson, he, you know him, you'd know him if you saw him. He pinged me and said, you know, I'd like to learn, I'd listen to the cube pod. I like it. I'd like to learn more about how you and John got started, how you see the business evolving, what your business is like, changes in the business. And you have a lot of good perspectives. I'd like to, I think people might dig some of that, you know, inside baseball, you know, but anyway. I think, well, first of all, yeah, definitely. Because one, it is, I think the, it's good to know what the origination story is, but I'll tell you right now, it's not just telling the story. People want to know what we're doing because they want to work with us. So we haven't been really good about marketing it. So when my big walk away, again, in these past, this year and these events and especially Google next is that we actually have more capabilities for to get paid than we thought. So customers value certain elements of our business model. So what's exciting is that you're going to start to see, we'll see new stuff come out and that's going to be fun to talk about. But again, the most important thing is how to get the content out and scale content that's quality that hits the mark where the demand is. Google next, you search Google next, Silicon angle the cube, you'll see all the stories. It's awesome. And you guess what? That's all important because it's not just, it's not Google saying right about us because Google's having an event. It's like covering the game. They're making moves on stage. You got to cover the touchdowns, Dave, you know? You know, if they make a good place, the commentary is easy when it's happening right in front of you. So to me, again, I think media is changing and more importantly, the ability to make money, to fund better talent is coming into clear view. And it's going to be fun to see how this all plays out. So again, I'm just super psyched. And yeah, and again, what we did with the cube this week, we did another first. I did a live remote hit to New York Stock Exchange TV at lunch on day three of Google next and did a event roundup. So if you go to NYSE on Google, you'll see April 12th, the Friday edition. My segment was aired. Actually it wasn't a live hit, it was a prerecord that went live in their program. And essentially at the end of the segment where they go, let's go to John Furrier for what's going on at Google next, a really big player in the cloud business. And they asked great questions. Trinity, their host was like, are they catching up? And I was just really easy to do. And then did the interview because it was short. And then I went back to doing more cube interviews. It was awesome. Yeah, it's all good. That's great. Well, and we got, we'll see. Next week we're at SAS, right? In Vegas, you and I, we'd be together. That's ironic. We didn't do Google next together. We'll do SAS next week. Be awesome. Well, we'll have to get the numbers. I'll put the published numbers from Google next. Well, Dave, great, great pod. Again, a lot of energy lost this week. I'm tired. I'm burnt out right now. I had- You guys did a great job, John. Congratulations. I mean, what an awesome team. You, Savannah, Rob, Rebecca, Dustin, you guys were terrific. I was, I felt like I was there. I was able to catch up on all the news. I'd watch at night after I got back. And it was a great job. Everyone just did an amazing job. And, you know, Google was leaning in. So, and what's great about this event is when, when Google brings talent to the set, brings us access to the execs, it's awesome. And these were execs that had something to say. It wasn't just scripted talking points, Dave. So that was really good. And two, we have a lot of friends at Google now. And kind of remember, Kubernetes is 10 years old. And we've been covering Google as an industry. We know people from Silicon Valley. But more important, that Kubernetes team and the, the Google management team, this is a big part of their cloud strategy, right? And so it's fun to see people that we've been kind of, you know, on a journey with over the past decade have positioned power to make, build, build Google cloud. But also we know them. We knew before they were famous. So it's great to see them successful. And they, they're happy for us too. So again, I was pretty pumped. It's very similar vibe to AWS from, around 2015, 2016 vibe with Amazon web services. Remember that feeling where they became, we knew them before they were super popular, all the management team and the people building it out. And then it became like really popular and everyone wanted a piece of them. And we were just one of the team, we're like a team member. It felt good. The similar type of thing happening at Google. But all right, more of that later. And next time we'll get some, let's really get Andy Jesse on. I'm going to hit him up to see what if he comes on. It'd be great. Love to have Andy. Dave, have a great weekend. Thanks, John. Are you listening? Hey, if you like it, give us, drop us a note on LinkedIn or Twitter. Go to SiliconANGLE.com. That's where the, all the traffic is. Cube.net is where all the videos are stored. What's where the events going to be. And some goodies there. Cube alumni database and a lot more stuff. And of course, the cubeai.com. Check it out. That's our language model, proprietary language model. You can say it's- Our rag. Our retrieval augmentation generation. Say that 10 times. All right, Dave, see you later. All right, John. Thanks, see you.