 Hello everyone, welcome to the Cube Pod episode 41. I'm John Furrier with Dave Vellante, weekly podcast. We extract the signal from the noise. This week, it's getting into the holidays, Dave. Christmas is right around the corner, Friday. Kind of a busy week kind of clearing the decks. A lot of work to do, prepare for the end of the year and getting ready for 2024. Great to see you. Good to see you, John, Merry Christmas. Yeah, same to you. And happy holidays, everyone listening. Hope everyone has a good holiday. It was supposed to snow and tub. We've got a little rainstorm here, as they say. It came from the south, so it didn't get a little warm rain. So it didn't get the snow in the tops of the Bay Area, Silicon Angle, Silicon Valley people are going to be bummed, it's not much snow. No skiing. Well, there'll be some snow up there, but it's only two foot base, so it's still not bad. We can play. No ice. It's not like Mad River Glen when you've got all ice in Vermont. You can't ski in Mad River unless it dumps. So obviously a lot of stuff to talk about in terms of end of the year. Not a lot of big news here, but I mean, I think the big story I'm seeing is a lot of deals across the spectrum. Obviously the market is tight. We're going to get into analysis here around the S&P 500 stock performance all time where that is hitting compared to the tech stocks. I'm sure you're going to have a good analysis on that. A lot of deals, M&A deals, companies going private, funding deals, some good funding deals around AI, obviously, and then the walking dead. A lot of companies and a lot of VC is publicly coming out, Dave, and pretty being doom and gloom around what's happening around the startup scene. As we said in the pod many episodes ago, the startups will be falling out of the sky. It's the walking dead right now. That's kind of happening. Also big news this week, Apple has stopped selling Apple Watch Series 9 and the Ultra 2 days before because of the ban going into effect of, again, litigation and the travel import ban, I should say, not the travel ban. The import ban goes into effect. Huge issue. That's been the dominating the news. And of course, the year in review, Dave, is the prediction time. All the people that's pontificating on their predictions, even from the analysts, and then pretend analysts as well. VC's are weighing in. Obviously, always fun to see what they do. The fake analysts. The real analysts and the fake analysts are all put in their predictions. And people always have hot takes. I mean, Twitter acts as turning into quite the platform. Threads is where all the cool people seem to be. All the hot takes are on Twitter. We have a lot of stuff being discussed. It's really awesome, actually, some of the predictions. Obviously the ones we made years ago and continue to happen in the day around date and cloud. Anthropic just got a $15 billion valuation in the reported kind of funding process they're going through. So they're in progress right now to raise a bunch of money to a $15 billion valuation date. Remember, Anthropic, that's the company that headlined in AWS. And of course, Mobile World Congress is coming up. We've been preparing for that. Looks like theCUBE is going to have a massive set at Mobile World Congress now called MWC. We're not going to be at CES this year. We'll be there remote covering from the studio as well as some folks on the ground there. So MWC is our next big show. And then finally, a section I want to put out there because it came back up again, Dave. Loud repatriation done right. Patrick Thornhill, a LinkedIn friend commented on DH's post. He's the Basecamp CTO co-founder around looking back at the money they're saving. So I'll see what, 6,000 Dell servers. And of course, Patrick works for Dell. So interesting comment. I want to get your thoughts on that. So good lineup tonight. And of course, the rant sections up as a mixed bag and maybe we could bring in some memes, meme me up some memes and there's some good commentary going around the internet this time of year. So I'm looking at my 2023 predictions. Every year I go back and I publish on how I did. I did okay. I don't think I did great though. I said, tech spending increases four to 5%. That was wrong. Cost optimization, big theme, that was right. Security, I said, Cisco's going to buy. They did obviously big $28 billion acquisition. Zero trust gets real. That's true. Gen AI hits where metaverse missed. That was right. Cloud expands to super cloud as edge computing accelerates. Cloudflare wins. That's not bad. I mean, they were winning. They're continuing to win and hurting. Here's a miss. Blockchain struggles to find enterprise home. That's true, but devs will adopt in 2023 and solidity wins. I don't think so. AWS Databricks Google. That was a prayer. It was a prayer. It was a prayer actually. It was a Hail Mary. Hey, look, I don't just. That was your optimism. You were just hoping for that to happen. I said, automation makes a resurgence. UI path and power automate separate from the pack. That's true, but they really didn't win because the gen AI kind of messed them up a little bit. Number of physical events doubles. Big events get small. Digital becomes a first class citizen. I think that's true. That's not bad. I'll have to do my grading. I don't know. I mean, I think we successfully predicted super cloud next generation impact of multiple clouds. I think we also predicted the generative AI wave. Again, we underpredicted that, that over, that came in heavier than we thought. And I think a big miss from me, this last prediction of pretty much we got it all right. My big miss was not so much blockchain as, I didn't see the AI agents, agent side of it coming. I thought the chatbot market was just a crappy market. I didn't think that there was more headroom on the quote, chatbots. Remember the enterprise chatbots? Yeah, that's hot. I just saw ETR data. That was like the number one use of AI and it's exploding. So that's a miss, but I think that's more of a miss on the fact that I just didn't see the app yet. And given the chat GPT focus, it was just more of like more, I went more into the data platform, but what we did get right, I will say, this came up on Sanji Mohan's predictions. He posted that on his blog. He's a prominent analyst in the data space. We predicted the intelligent data platform. That was something that we picked big. We also predicted that the data would be upside down, flipping the script on how data is organized horizontally and vertically. That was a huge prediction. I mean, I think that was probably the most successful original prediction we had last year. And I'll see that reinvent, if you remember, we were on the whole next gen cloud prediction from last year. That actually was happening. So if you look at next gen cloud, Dave, think about it, all this GPU cloud stuff, that's what we're talking about. So I think that was something that emerged. Not as obvious as we had thought, but it did get clarity on that. That next gen cloud became like the MongoDB. Look at Mongo's success. Look at their stock, what they're doing. Okay, Mongo Snowflake Databricks. We predicted they would rise. They did. You see the new magic quadrant that came out? I did. We didn't share it with you this morning, actually. Yeah, that's right. That's where I saw it. Thank you. And Mongo looks really good. Databricks, I got to go compare with last year. It looks like Databricks ticked up a little bit. Snowflake maybe, they're a leader, but they're fighting neck and neck. Oracle's in there. We had a deep dive the other day with the head of Oracle engineering for Golden Gate, which is kind of like their open integration platform. I said open, but their integration platform. I was really impressed. So they saw the work that we did with the six data platform and they pinged us and they're like, we got to show you what we're doing. I had to tell you, I was impressed. I'm going to have one of the guys on breaking analysis. It was that good, super technical and it was great. So that was interesting. I just got to do one more thing. So I get a big stack of predictions every year because I've been doing this for like a decade. And so all the PR people send me predictions and say basically please use our predictions. And I can't use them all, but so I pull a couple out. The one here, I just want to share with you, this guy Scott Stevenson from DeepGram, this prediction was AIR's recession proof in 2023. He got that one right. Nailed that one, Scotty. Well, I mean, look at, I mean, it's going to be fun year because 2024, it's going to be very interesting because what I want to talk to you about in this podcast is the, I want to weave in some financial analysis. I know we tend to talk about stock prices and stuff, but it isn't the indicator. And we can jump into it now and get into some of the predictions if you want. But last night I was checking out the S&P 500 all time. And we were speculating on this market because the prediction was that 2024 is going to be a comeback. And if you look at the tech stocks and the S&P 500 all time, it's the graph looks just like it's a steep curve. When you go back to 2009 and even go back to 2019, it's a massive step up and all the tech stocks map the similar trajectory. It's interesting, Apple was the best bet of the tech stocks. Amazon is still at an all high, but they had that big drop. Apple didn't really drop as much. Meta had a killer drop, but they're back up to an all time high again, pushing that number up. So you got Amazon, Apple, Meta and Alphabet all at the same kind of graph. I mean, Berkshire Hathaway and JP Morgan, again, similar graphs. I bring them in there because they're not just tech. They're a good balance alternative to bounce around with. So you say, okay, tech stocks are doing great. They're matching the S&P 500, but also Berkshire Hathaway and JP Morgan's stocks are at this all time high. So if you're an investor in the S&P 500, you're doing great. I think that to me is a bellwether. And the question is if you go back 10 years on these graphs, okay? Go to Google, type in the S&P 500 and check the tech stocks. I just mentioned Amazon, Apple, Meta and Alphabet, which is Google and then compare Berkshire Hathaway and JP Morgan, which is a bank stock and then it can diversify the fund. You see kind of a big picture. The question Dave for 2024 is, we could have a financial meltdown, right? I mean, the growth of the value in the past 10 years and then even go back five years, it's really incredible. So the question that's all coming in these doom and gloom predictions is, there's a lot of signaling from VCs coming in saying, it's gonna be a horrible year. I'll tell you one right now. Jonathan Helinger, general partner at Vertex Ventures says, a lot of companies are going to die. Direct, that's a statement, okay? So big time macro environment tightening, that's gonna be there. All the hurdles for business problems are done. So the people who are spending and not having a sustainable revenue model right now are gonna be in a lot of trouble according to him. Insik Ray, Wade didn't say- What's he say? He says, Insik Ray is the general manager, great partner there and a friend of the cube. Bigger and bad are cyber attacks are coming, okay? And he's really specifically talking about AI now, driving better attacks, specifically social engineering. And then there are other partners, Sam B. Badger says, startups going back to the office, that's gonna be interesting, that's gonna be collaboration. And he also says standardization of open formats. Were you in predictions now or are we going off the stock market? I can't even comment on your stock. I'm just giving the gloom and doom. So just kind of hinting on some of the, we'll get into the prediction later, but that indicates, Dave, that 2024 could be a reset year financially on the stock market specifically. So I'm not gonna- I don't know, there's a flip side there. So first of all, we should say, we should do the disclaimer, like don't invest in anything we say, because we don't know what we're talking about anymore. I say the opposite. I say, do you invest everything? I say, you'll be rich. Okay. But the disclaimer, we're not giving financial advice. But there's a flip side of that, John, which says it's an election year, interest rates are coming down, don't bet against the Fed. So I mean, I don't know. I just, I like to be in the markets. And I mean, I feel like, I don't think this, it's very difficult to beat the S&P 500. I mean, why would you bother trying? I mean, Warren Buffett always said, just put your money in the S&P 500 index stock. But I will tell you, I mean, Metta, take Metta, you're talking about two billion users. It's like every time Metta drops, like you should probably buy the stock because how do you get two billion users? I mean, that's like incredible. I mean, Apple, it's like, you look at that chart on Apple. Apple's on Apple's incredible. But then it's probably gonna hit $3 trillion valuation this year. I really like Amazon. The reason I like Amazon is because I think Amazon web services is gonna accelerate growth again this year in part because of Gen AI, but I think IT spending is gonna pick up a little bit, especially in the second half of the year. So I really like Amazon because that's gonna drop right to their operating profits. And I like Google. I mean, it's just company prints money. Despite the fact that I think GCP is kind of underperforming. I mean, Google still, it's like the same comments on Metta. So I mean, and then the bank stocks, I think as interest rates drop, JP Morgan, I think is a winner. I think you got some good ones here, John. Berkshire, I don't own any Berkshire. Well, I wanted to give a market basket in there because it's not just the tech stocks. I mean, Apple just went straight up since 2019. Just incredible. But look at NVIDIA. NVIDIA is the craziest one. Yeah, NVIDIA. Well, NVIDIA is just one. I heard Kramer the other day saying, and I like Jim Kramer. Jim Kramer was my first Wall Street client when I started the IDC investment service with Alexa McClown and Profit actually was heavily involved. And I heard him the other day saying, when stocks go hyperbolic, it's time to sell. You got to take some profit and then later on in the same segment, basically the same breadth. He's like, don't sell NVIDIA. I'm like, wait a minute. That thing went hyperbolic, so. Well, look at Broadcom stock too. So NVIDIA was a good stock this year. So that was a good one. Honestly, if you got NVIDIA anywhere after 2014, it's incredible. So even after, anytime after October 22 and before March, I mean, the thing is just exploded. Look at Broadcom stock too. That's another good one. Yeah, Broadcom had like kind of an iffy quarter, not just past earnings, but the earnings before. And so people were selling and I was like, it was a buying opportunity. I mean, Broadcom just got a great business. They doubled their stock price this past year, over doubled it. So, I mean, last year they came in, they were 500 and changed by 44 in December. Yeah, to me, what's impressive about Broadcom is how they run their business. And it's gonna be really interesting to see, you know, how VMware fits into that, right? I mean, you know, they're gonna run it as a, they, you know, Hoctan doesn't rely on business A to help business B. They have to be like standalone businesses, right? So, and I think VMware is an awesome standalone business. I do think, you know, customers are concerned, you know, about the acquisition, they're concerned about pricing, but they don't have anywhere to go. I think, I think I'd put Broadcom in my little market back at the basket here. And I'm gonna do that right now for the graphs, for the show notes. I think you should. I think it's representative of the chip because I'm gonna put Nvidia and Broadcom in there as well. So, we'll put them in there. I mean, do you think, I mean, I mean, I don't mean to be loom, I'm a very positive person, right? So, but, you know, I think Adam Salipsky, I said this in my breaking analysis, he summed it up very well, CEO of Amazon Web Services at Reinvent. It was very, he had words of wisdom, I thought. He goes, I've seen worse times, I've seen better times, but I've never seen such uncertain times. And I think that sums up, it's very hard to predict right now. Is it supply? Is it demand? What's the consumer gonna do? You know, inflation, interest rates coming down, but what about employment? It's really, really hard to predict what's gonna happen. And so I think the key is businesses have to remain agile. I think they gotta be resilient and flexible. They're gonna be investing in those types of things. That's why AI is such a huge investment vector in 2024 because it's going to help you be more business resilient, be more productive, be more automated. But it's very hard to predict what the impacts are gonna be on jobs and the economy. It could have a massive, wonderful impact on the economy. It could have a near-term bump, bump downward in the economy as it could affect jobs. To me, it's very, very hard to predict. So I think you just gotta lean in and try to get ahead of the game. This is one of those waves that you're gonna look back and go, shit, why didn't we do XYZ? Just like you did with PCs, just like you did with internet. I mean, we've been riding all the waves, I know, John, because we're media and analysts, so you can ride every wave. But tech business, you gotta go hard on AI. No doubt about it. I'm not a skeptic there. No, it's definitely the big wave, but I would doubt. Again, we'll see how the external market, how the external market... You think Broadcom could be a trillion dollar stock? I think it could. I'm looking at the chart. But it depends on VMware, obviously, but they're halfway there. VMware, I think I've said it in my prediction before. I think VMware pulls their entire portfolio together on the software side. So as I said, when I was, when I was conspiracy theory on this thing called that, I said from chips to software, and that was what the whole theme was at re-invent and at supercomputing. That from the chip to the application and that whole stack's developing and Broadcom's gonna have an answer for each layer of the stack, from chips to the apps and the data. So again, the data piece is key. Now let's get the predictions in because I think that was part of my point of showing the stock market was to saying, can that continued trajectory of S&P and these key stocks hitting all-time highs continue? Is this just a base to build on? So my thinking is there is a huge macro common. Jonathan Hellinger and Insik Ray wrote it two points. A lot of companies are dying and are going to die. That's his quote. And then Insik said, bigger and better cyber attacks are coming. This is huge. What did he say? Say it again? Bigger and better cyber attacks are coming. The weak link in the modern economy isn't code or encryption, but rather the people who get tricked by bad actors by giving access to confidential systems. The new generation of AI tools like Chatchapiti give those bad actors a powerful new set of capabilities to manipulate the unaware and vulnerable into enabling new schemes that could well be on the ransomware epidemic of today. Expect to see at least one major new AI-enabled cyber security-related incident in 2024. I think it's... I think that's like a lock. That's like a lock, one, one cyber. How about a hundred? But I mean, that's not anything new. I mean, AI in gen AI is new, but it's not really new that we're going to see. I mean, it's always an arm tree. I think his point is, is that outside of ransomware, which is right now, everyone knows, it's the killer app for the number one app that's profitable, the social engineering that we've been seeing that we saw at MGM and Caesar, that's going to scale. So I think, and we heard that on our last super cloud event we had, Dave, that we heard that on the Dell Data Protection. So again, when last week, we had our road to cyber surveillance with Dell Data Protection, Palo Alto Network, she said, the ransomware is not the number one, it's going to be social engineering. And this new cyber attack. So I think this is an accurate prediction. Yeah, Wendy Whitmore said that. Yep, she's right on, she knows. I mean, they, like Kevin Mandia, Wendy Whitmore, they see all the trends before, you know, they hit the mainstream media because they have the threat intelligence units, like the smartest people in the world on the grid, like watching all the trends, using AI. I mean, they're like really, really advanced. They work closely with governments. So yeah, I mean, if she says it, I believe it. Well, and that's what InstaxBase is saying. And by the way, we've reported that too. At the Mandian Conference, we saw, and again, the MGM hack that we reported, the social engineering is the new thing. So that automation is coming. It's going to be... But that's all the more reason to love, like a CrowdStrike, a Z Scaler, Palo Alto Networks, even CyberArk. I mean, these are pretty good companies, you know? By the way, InstaGray, remember he was shitting on me for my Oracle comments at SuperCloud One? He's like, oh, you lost me at Oracle. Hit an all-time high this year. Let's pull back a little bit, but my point is it relates to security. Mission critical, highly secure. I think that's the 10th time you brought that up on the pod. Yeah, it was a little sensitized by that. Do you want to talk about it? You want to talk about it? I do, I kind of do. Go ahead, Dave, talk about it. I think he's one of the best, I think he's one of the best VCs out there. So you kind of, he was just tongue-in-cheek because it's Oracle. By the way, Oracle buys a lot of companies, so he likes Oracle. That's what I said, too. You just want to have your companies bought. I mean, we were going back and forth. He's great, I love InstaGray. But I have any VC, I always say, okay, what's their agenda? They've got portfolio companies. Is he holding a lot of security assets that he's trying to pump up? Is he trying to jam valuations down? Like in these bad times, a lot of VCs will to try to get better deals, and how much of that is like gamesmanship. And you know them better than I do, so. But everybody's got an agenda. You and I include it. I have no agenda, Dave. No agenda. Just to get the content out to the users, that's my goal, to get the truth, and I do the best analysis. That's an agenda. That's how much area possible. That counts. That's a mission. That's our mission statement. Okay. All right, so getting back up to the... Sorry, I digress here. Are we making predictions here? Do I have to like... No, we're reviewing predictions. Oh, great, we have to make our own. How about a Bitcoin prediction? What do you think Bitcoin is going to do? I mean, we can make predictions all you want, but I was going to go through other people's predictions and tell if they're good or not. I wish I knew we were going to do that. I'm not kidding. I have a stack of inbound predictions this big. I'd like to go grab them. I'm going to have somebody bring them to me so I can go through them here. They're in real time. So back to the office, there's a big discussion about back to the office coming back. I think people have to be back at the office, and we debated this actually before, and there's two schools of thought remote companies are doing some, some of them are doing really well. Some aren't. And there's a blowback on going back to the office. So if you're a startup in this market, again, given the climate, the Silicon Valley Scuttlebutt right now, and I'm sure it's around other areas and other entrepreneurial circles is, it's a tough time for entrepreneurs right now. It's funny, you know, people look at our business and say, wow, you guys are in the cash flow business have been bootstrapped. We're doing this fashionable now when we started the company 13 years ago. So, you know, the right now, the fashion in Silicon Valley is longer runways, cash flow positive, be very strict with your cash, right? Make sure you have spending your cash in the right places and the whole SaaS metrics table upside down. So it's a very tough economy. There's no public market window and startups are in a non-luxury position relative to cash spending, which means their jobs 10 times harder because they don't have the cushion. Again, I've always said that's always been the alpha version of entrepreneurship. Everyone always asked me, oh yeah, I just raised a huge round of funding. That's actually kind of easy, no offense, but that's kind of easy. The hardest entrepreneurial pinnacle, the Mount Everest of entrepreneurship is building a durable company without outside funding, then taking outside funding when you've already cleared the runway for either liquidity purposes or super hyper growth. GitHub did it, there's a bunch of companies that have gone down that road. Alasian was one, that to me is the alpha entrepreneur. Opportunity recognition, self-finance, customer-driven and scale, they still recruit the team and build a durable company from scratch, from zero stage to profitability. The ones that are, the second tier down are ones who can see an opportunity, raise the capital and go for it. And then the third is the ones who do it and fail. And then below that, you work for a big company. But that's the startup market's hard. So a lot of people are fleeing this entrepreneurial scene unless they're in AI. And even some of the AI startups, David's saying, are saying they might not have a future because of the big companies like Amazon and the Cloud guys creating their own stuff. So huge conversation around the entrepreneurial market. Again, I'm with you on the AI opportunity. I think it's a wave that if you don't have a company, just start serving the waves, you'll get something going because it's too good. The market's very bifurcated right now. I mean, there's definitely the AI haves and the AI have nots. Everybody who's a have not is AI washing. And that's gonna catch up to them for sure. And the guys who have real legit AI are gonna ride that wave and do better. I mean, I think it's very clear. And in budgets, it's not like all of a sudden, budgets are gonna open up. We've been saying that AI has been stealing from everything else. And I think that's still the case. And I think that's gonna continue to be the case. There's no evidence that that's changing. There is evidence in the survey data that folks are expecting a bump up, like four or 5% in spending next year. But that's what they expected last year. I think that a lot of that's wishful thinking. They don't really know. And by the way, when you look at it, they don't expect that in Q4 and they don't expect that in Q1. So what that says to me is it's wishful thinking. It's like, okay, plan for X, but be agile because we might have to cut your budget if we gotta hit earnings. And so that's the big concern people have right now is what are earnings gonna do? Are the valuations gonna be supported by earnings? And if they're not, I mean this thing, markets to your earlier point into the negative scenario, the market's priced pretty much to perfection, at least a lot of the market is, certainly the magnificent seven are. And if there's some event, did you see the deep fake of the Pentagon on fire the other day? Did you see that? No. Deep fake of Pentagon on fire and the market reacted. The market dropped. So you can game the market now with deep fakes. Was there a trade behind that? Was somebody shorting the market, waiting for it to drop with a deep fake and then closing out of its short position? Absolutely, absolutely they were. That's gonna happen. And this is gonna be the guardrails are gonna go up. I mean, this is the whole prediction piece. I think the cybersecurity attacks are gonna be much different than we've ever seen before. I do agree with Instac on that. The other prediction that these guys had over it for a decision I like was now some deep, Bodgera, another partner mentioned to go back to the office. That's not as cool as the next one. He likes the standardization of open formats, I bring that up because we reported this at data bricks event, remember data plus AI and you read Snowflake Summit. Data is the fuel for generate AI boom, but ingesting it to LLM's others can be seen the headache for users. We expect to see industry coalescing around open standards like Apache Hoody and Apache Iceberg with many other major plays realigning their strength only. This is directly in line with the work that we're doing at theCUBE research. And the work you're doing on specifically this thing that you're calling the sixth data platform. I don't know why you call it six, but is there a reason behind six? Yeah, because the big five, and this is what got Oracle a little bit triggered, I think, but the big five are the three big clouds, AWS, Microsoft and Google, and then data bricks and Snowflake. And the whole concept was they each have their own way of separating compute from storage. There's other guys that do it, but these are the five that are really dominating the market conversation. Of course, Oracle is the king of database, so you can't exclude them. But so the sixth data platform is not just separating compute from storage, but it's separating some compute from data. Because if you separate compute from storage, you can scale them independent of each other, but you're still sort of locked into that same environment. It's from the same vendor. If it's Snowflake on AWS, it's AWS compute and storage. If it's on Microsoft, it's Microsoft and data bricks. So the future is, when you see companies like compute AI, they're trying to, vast as a vision to do this and others are trying to, what we say, create like an Uber for all, where you can have access to any data, any storage type, many, many queries. When you say Uber for all, you mean Uber as the company. Yeah, so Uber, think about the Uber as a company, not Uber as an adjective. No, Uber is the company. Think about the app. You're standing in a street corner. In real time, you call a car. It knows you. It knows where you're going. It knows where you are. It knows where the drivers are. It matches you. It knows what the price of the route is and tells you how long it's going to take. It does all this in real time. Those separate disparate data elements, they could be sitting in, like you said, an iceberg table. They could be sitting in a data, in a transaction database. And by the way, they are, they're some kind of combination of Spanner and I think some Postgres and some document database, all these different formats that Uber, back in 2015, figured out, how do we create semantic coherence across all these different data types? And they had, you know, 3,000 engineers that built this amazing application to do that. Again, mid last decade, wouldn't it be very amazing if for everybody, Uber for all, if anybody could get that capability and do people, places and things in real time. And that's what the six data platform is really all about. Right, that's a good description. That's a good segue into Sanjeev Mohan's trends. That was a good post. That thing was amazing. He did a deep dive. So Sanjeev Mohan, he's a cube collective member of our team. He's on his own. He's got a great his own firm. Sanjo. Sanjmo. Sanjmo. Sanjmo, yeah. Sanjmo. That was all wrong. Former Gartner analyst, Sarbjeet. Those guys are great. But what's interesting about XGartner, he's like one of the best analysts in data. So he's solid. What's interesting, he has what's, he had rising trends, stable and declining. What's declining in his mind is the modern data stack, the promise of a decentralized approach to moving data from IT to business. It didn't ever got off the ground. Is that a typo? Cause he's got the same words under data mesh, right? I think he just copy pasted. I wonder if that, I think he means data mesh because they're both kind of the same thing. But both run into the same umbrella. Once infrastructure, one was more control play. But he's basically saying, data mesh is passe date. We're kind of declining. I want to know your thoughts on that. I mean, I think I said, I posted on his LinkedIn, I said, you know, it makes me sad that you say it's declining. But I mean, I get, his point is that the concepts that Jamak Degani, who was on the cube at SuperCloud was amazing. The concept that she put forth, they never really talked about specific implementation and technologies. They avoided doing that. And the, but she spun out and started her own company to do just that. It's going to take some time. I always saw data mesh as an organizational construct, decentralizing, you know, data management and giving access to business people. I think that, I think in concept that trend will continue, but it's not a technology that you can go out and buy. And so, yeah, I don't know. I mean, you go into perplexity and say, is data mesh rising or falling? And it'll tell you it's rising. So I don't know. He's right from, I think, a technology standpoint. I think I trust Sanjeev over perplexity app. Yeah, but sure. But though, so for instance, data fabric and data mesh are oftentimes sort of confused. And I would say that data fabric is gaining in popularity. And I think that's an infrastructure enabler for data mesh. And the modern data stack is again, separating compute from storage. It's cloud native. It's software as a service. It's a consumption model. I wouldn't say that's declining per se. I would say it's hitting its peak. Let me put it that way. I think it was on the steep part of the S-curve. And I think that part of the S-curve is flattening. So the interesting thing is gonna be how do Databricks and Snowflake jump S-curves? We know that Snowflake is doing it through things like data sharing, through the container services, basically it's App Store, if you will, App Store for data. That's their big move, Databricks, sort of breaking away from Microsoft, which helped them with their steep ascendancy. And then you got the big three cloud guys. For instance, how does AWS bring together all its multiple data stores and its data platforms and all the different metadata locations and stores that it has? How does it unify its metadata? That's, I think, really what he's talking about is the momentum that we saw in the last five years for that so-called modern data stack is starting to peak. And I would agree with him. And I think it's moving toward the intelligent data platform. Now all these modern data stack guys are gonna move to intelligent data platforms. They're already headed in that direction. But I love this chart. It's called Top Data and AI Trends for 2024, Rising, Stable, and Declining. And the rising is intelligent data platforms, AI agents, personalized AI stack, and he's right on AI governance. So he's nailing it and then stable. He's got data products, metadata plane, cross-cloud, unified data plane, and then declining modern data stack and data mesh. And he defines each one- I guess a cross-cloud is super cloud, basically. Yeah, absolutely. So he defines each one or at least goes into it in a really great detail. And this thing is a tone. How many words is this? It's like bigger than a breaking analysis. He's got nice block diagrams and architectures. He's really well thought out. And so I guess I would agree with you. I trust him more than perplexity. I just, I love Jomak Tagani. I think she's brilliant. And I hope she does really, really well with her startup. Well, it's a lot to digest. We'll try to get a link to the graph on the show notes. But the point of this is to take away if you're watching and listening, the rising is the intelligent data platform. It's funny how the modern data stacks declining. The word modern is the relative word. I think that's not declining. I think it's the specific old way they did it. But what's rising is the data platforms. Data, this intelligence data, that's key. That's what we've been doing a lot of work around. So you look at our content, you'll see us covering all the companies that are essentially making that. Again, the AI agents, I was skeptical on this, but after seeing the demos from the SAS conference we went to, to reinvent, the applications are going to be agent and personalized. So to me, and he mentioned, and he gets into this personalized AI stack. There's another debate going on around the notion of the word stack, because AI is not a stack, it's a graph. So the question is, is that, can you even call it a tech stack? Okay, but those two concepts of AI agents and personalized stack, personalized AI, huge, huge. That's the app. That's the killer app in my opinion of 2024. The emergence of the killer app in AI. And it's going to be personalized and agent-bedriven, meaning augmenting human intelligence, okay? Not AGI, augmenting human intelligence, A-H-I, okay? And that's going to be enabled by the companies and startups and engineers and entrepreneurs that build the intelligent data platforms and solve the governance problem. So I think he nailed the rising and I think he actually hits a home run on this because he hits the two critical things that have to enable the killer app. Because without governance being built in from day one, rethinking that, you're screwed because you can't actually scale the AI. You'll always have solutions and you'll have deficiencies. So, and the other prediction I'm going to make is that LLMs are going to be refactored in a way that people aren't seeing them. Two reasons, security. Insik Ray brought up the AI piece before. He kind of teased out LLMs, but you're going to see a lot of people protecting their LLMs more. And then LLM integration with other models, right? So the power law that we put out last year, 2023, that'll continue to happen and the emergence of some alchemy between LLM models. So integrations of fusion, the interaction of data, that will enable new use cases. That's a good word, alchemy. He's got it. And you can't do that, by the way, if you have old legacy data management. Okay, so here's the thing. So when I think of modern data stack, I think of, again, separating compute from storage and cloud native and SAS and everything we talked about before. And I think of Snowflake and Databricks. When you look at his diagram, it says figure two components of an intelligent data platform, you build it from bottom up as GPU, CPU, and then on top of that cloud platform. Then you got unified storage. And then above that, you've got this multi-engine orchestrator, which is the semantic layer, the vector indexing. You got other analytic engines. And above that, you got APIs and an SDK to support a data prep and AI governance. And above that, you got agents and co-pilots that are gonna be taking action for you, systems of agency. And then data products come out of that. So if I'm Snowflake, I'm saying, well, this is exactly what we're doing. This is what we're building on top of our clean rooms and our governed architecture. And Databricks is saying the same thing. And so what I think is that those modern data platforms, they're gonna be challenged, but I think they're well on their way of migrating to these intelligent data platforms. I do think what's gonna happen is the guys who really didn't hit escape velocity that got a lot of money from VCs to separate compute from storage and go to the cloud, they're gonna get bought or they're gonna, companies tend, I guess they do die sometimes, but they tend to get bought up or accu-hired. And I think that's what's gonna happen to many of them that didn't get through the knot hole. And I'm dying to see Databricks go public. I can't wait to see their numbers. Exactly, that was what they're hiding behind the curtain. What do you think they're hiding? I'll blame them. What do you think they're hiding behind the curtain? No, I don't think they're hiding. Well, I mean, they're private. No, I think they're hiding, but they're private. Yeah, but if you look at the IPOs in 2023, it was a really shitty time to do an IPO. And I was thinking, I think they said, we don't have to do an IPO. Why would we do an IPO now? And they can raise money in the private market. Right, and the old days, you went public to raise money. Now it's just like, you don't have to. Do you remember, I know you remember well, because you know these guys better than I do, but Cloudera, how they basically, Mike Olson's rap was, well, we don't need to raise money. They raised a boatload of money from Intel. And so they said, we don't need to go public, I meant to say, not raise money. They raised money on private markets, and they held off doing their IPO. There was a time when being a public company, you remember in that timeframe was just coming out of the downturn was a horrible time to be a public company. And so I think, I don't know, I think why would Databricks rush to go public? Look what happened to UiPath. UiPath had a $38 billion valuation before it went public. It's like last round, all the late investors got into $38 billion valuation. I think they're probably trading at 11 billion today, 12 billion maybe. And so, when you see companies touting their valuation, I mean, I get why like Vast was excited, $9 billion valuation, and there were only a couple of other companies like Anthropic and OpenAI who were able to sort of sustain higher evaluations that were in the AI space. But you don't want to be the next UiPath from that standpoint. I mean, they still made a lot of money and founders did great, but the late stage investors, they got 14 billion now. So they got a ways to go before those series F investors get their money back. So you don't want to be that, and I like UiPath, it's just it's got too hot too fast. So that's the risk of some of these AI, these new emerging AI unicorn pluses. So, yeah, a lot of M&A going on in terms of companies. Cisco just bought a company, Isovellant, it's just a company we interviewed. Tots like Tom Gillis did that deal. What do you make of that? What do you mean? That was a security deal, right? No, networking deal. It was basically, they were doing cloud native and networking. They did a lot of the kernel stuff for Kubernetes. Basically, they were the Kubernetes networking company. So Cisco bought them up, and then Scott Rainovich got a good post on this. He's a good guy. I saw him somewhere, it must have been a re-invent. He's a smart dude. We tried to work with him for a while. He's out in Montana, so we don't see him that much, but he does good work. He's got, his name of site's called Futurum. What is it called? Futurum, Futurum. F-U-T-U-R-I-O-M. Futurum? Futurum, yeah. Futurum. It's like, there's, it's like Futurum. Like Futurum, yeah. So, you know, he, I think he was like, well, that sounds like those guys. He's like, don't ever say I'm so mad at that name. So he had a little thing. But anyway, the Berkeley packet filter EBPF was the thing that was part of the Sillian, which is an open source project that this company did. Pretty deep tech chops, guys from Google, OpenStack community, we've known them for years. Very geeky company that Cisco bought. So again, good for Cisco getting back in the action. Alter, it's going private in the news. That's another big story. And again, mentioned that, you know, the Anthropics trying to raise $750 million on a $15 billion valuation. $750 million, so. I thought, you know, I saw that. I thought, and I wrote this in a breaking analysis last week when I did my open AI, my David versus Goliath. I thought Anthropic at one point had a $20 billion valuation. Now, I think so. Well, that's interesting. I remember they have a lot of incestuous activity with other clouds, right? So remember, Amazon's got a deal. Adam Sileski quoted on theCUBE saying that, you know, they can invest up to $4 billion in Anthropics. So I wonder if they're in the round. And what they're saying in the reports claiming that Menlo Ventures in this deal. So it's going to be very interesting to see how this all plays out. So this is a post from October. Anthropic got money from Sam Bankman-Free, too, back in the day in 2022. So look at this, Google put $300 million in. Prominent chat GPT rivalized $30 billion valuation with big, new big tech investments. This is in October. This is on thestreet.com. Anthropic, the startup behind GPT competitor, chat GPT competitor with its Claude is looking to raise an additional $2 billion in funding according to the information. Google, which acquired 10% stake in the startup, is expected to make additional funding around. The startup looks to expand its resource and pool investors. The company is seeking a valuation of between $20 and $30 billion. And then now it's, do we say $15 billion, right? So why would they have a- That's the valuation. That's what we're talking about. The latest valuation number, I read on the Silicon Angle, was $15 billion. Well, I mean, it's, again, this is weird, that all this fuzzy math going on. So Google had put a commitment up to $2 billion from Google. Amazon committed up to $4 billion. They raised $100 billion from South Korea and telecommunicationally, SK Telecom. Sapphire viches committed more than $1 billion. And then their last priced round, again, we don't know what these rounds are, Jay. Remember, we were speculating that it could be in kind from Amazon because of all the work they're doing over there. Their last round was a series C, $450 million led by Spark Capital, that included Zoom, Salesforce, Microsoft, HOF Capital, Menlo Ventures, Sound Ventures, Pioneer Fund, Wilkus, SV Angel, Alphabet. So that's- It's all over the place, I've seen. May of 2023. Now, we reported on Silicon Angle that $300 million from Google was on a $4 billion valuation in March of this year. $4 billion. So Silicon Angle, we break the whole post out, all of the trending. It says in May, they were valued at $5 billion. And then last week for my breaking analysis, I had them at $20 billion. I'm seeing $15 billion. I'm seeing $18.4 billion. And the information has a report saying it's between $20 and $30 billion. And Silicon Angle had it at also at $15 billion. Oh, here it is, here it is. Although there's a Silicon Angle, although a valuation of $15 billion is being discussed, the final valuation of deal could go above $18 billion. I think it's gonna be north of $20 billion. I really think so. I think the research I did last week and a half ago had them at $20 billion. Well, it says $15 billion on this report from Bloomberg. I know, and I saw that too. And that's why I was like, wait a minute. And now the latest is it could be $20 to $30 billion. But anyway, there it is all over the place. But you saw that, you saw the, I mean, there's the Gen AI haves and the Gen AI have nots. Like Jasper's getting crushed. Here, you got it right here. Last week's breaking analysis. Hugging phase, $5 billion on $823. They're obviously doing well. Databricks is really not Gen AI, but it's ML and AI. It's $43 billion. Open AI, $86 billion. Anthropic, I had a $20 billion as of 1023. I'll have to dig out that report. Go here, $3 billion as of last August. Jasper was $1.5 billion, and now it's down, in September, it was down to $1.2 billion because it's getting disrupted by chat GPT. Data robot was $6 billion. There's no way data robots, $6 billion value. That was $721. July of 2021, it's probably less than half of that now. Right, I mean, I think it's interesting time, Dave. I mean, look at the world of private equity. One of the big challenges right now is that the capital markets on the entrepreneurial side are tight. The bigger getting bigger, like Anthropic, and then private equity sideways because they've bet all those late stage and trying to do startup deals. So you have a weird world of private equity. M&A activities should be hot in 2024. Again, there's gonna be a lot of rollups. You're starting to see people do that. Again, if you're on the wrong side of a rollup, if you're a company and you're buying distressed assets, let's just say that you're a company, you're buying all this distressed assets. You're buying shit assets to begin with. Then you gotta make them work. If you don't have the cash position and then you can't generate revenues with the free cash flow, you gotta go to the capital markets. And the capital markets right now are very tight. So if you're gonna do M&A, you better have the right formula. Well, the way- It's cultural too, right? If you're buying a bunch of distressed assets, are you gonna be able to keep the people? Are they gonna fit in? I mean, that's always a hard thing. But John, look at the delta between open AI's valuation, 86 billion, and they almost fucked that up with this firing Sam Altman knuckleheads. But then let's call it anthropoc is the next closest. Let's call it 20 billion. Give me the benefit of the doubt on that research and people now saying 30. But look at the delta between that open AI and that. You remember last January, you, Sarbjeet and I and Palo Alto studio did a breaking analysis around, the title was open AI won't be able to maintain its first mover advantage. And you disagreed with that at the time. You were saying, yeah, they will. I mean, it looks like you- What did I say? What did I say? Yeah, you said they'll be able to maintain. They'll have a first mover advantage was your argument. And Sarbjeet and I were taking the counterpoint saying that other guys will catch up. Google is gonna get great AI. They're gonna be a bunch of VCs, but open AI is like there's a huge gap. Now the question that the context, first of all, I was right and I'm still right, but it's evolving answer. And at that time I said, the context of that question we were debating was, is there one rule we were kind of debating is there gonna be winner take most, maybe a handful of models or one model to rule the world? I was saying, no, there won't be one model to rule the world, there'll be an open source. And that's how we got to the long tail. That conversation was the impetus of us publishing that pioneering research around the power law for AI. Because at that time we were saying, the market's not just a chat GPT. The question was, is chat GPT have switching costs? Is it sustainable? That was our discussion. I was saying they're gonna have a first mover advantage and will maintain that for some time, if not for a long time. Again, I didn't even say at that time the Netscape moment which ended up becoming an issue when Altman had his problem with the board. But that's what we were saying. We were saying it's gonna be just like a Netscape moment and we were arguing with you that they were gonna not be able to maintain that first move advantage. It's still early days, but not only does their evaluation show a big gap, but if you look at the spending data and the momentum and the sentiment, they are far ahead of anybody. They blew through Databricks, which is extremely successful. When you look at the private company sentiment in the ETR data, nobody's even close. They're like up and to the right. And then you got this pack of people, Databricks is in there again, they're not Gen AI, but the Gen AI guys, which really hugging face is not Gen AI, it's really like a market, right? So, but anthropic, cohere, snorkel, character AI, there's many, many others, they're not even close in terms of the market sentiment. I mean, open AI is kicking ass. And we see it, we use open AI with the cube AI, it's dirt cheap, it's better than llama two, it's better than these other models. And I keep saying, I'm dying to get to bedrock. I know that the guys are playing around with it. I haven't seen a lot of results yet, but. Yeah, I mean, it's that open AI is keeping them in. The gap's closing though. It's not, I mean, open AI is easier to work with. It's gonna, it's back down to the power law from a big language model standpoint, open AI is number one, and chat GPT is the application. Anthropic with claw number two, llama probably pushing up there with number two hard because it got uptake with the developers. And that was a genius move by Meta because to your prediction, Dave, Metaverse was a missed Gen AI is the home run. And look at the stock price with MetaSense, they should have changed their name to AI, not Meta. Yeah, or Meta AI. Maybe they will. Facebook, this Meta AI. I'm looking for evidence that along that, I mean, I love the long tail and the power law of Gen AI. I'm looking for evidence that there's stuff happening on prem. Honestly, I don't see it yet. I don't see it yet. Even though we predicted it, it's gonna take some time and I do think it's gonna happen, but it's not obvious. We're not gonna have time to get to it, but we can maybe do a quasi rant and this, but the cloud repatriation done right by the base camp CTO co-founder, he posted another LinkedIn post and this time given an update on stats. So three days ago, he gave some stats. They're saving millions of dollars. He says, quote, they bought 4,000 CPUs, 6,000 Dell machines. No, 600,000 worth of Dell. They said they're saving right here. Bring it now, because they're use case, they said they overpaid for Amazon. They don't need that. The provisioning is great, but for their use case, they want it on prem and they save more money. It's not worth the premium. So Patrick Thornhill shared this with me. They said conservatively they'll save 7 million. See what he says here. Already in September, we've secured a million dollars in savings on the cloud bill, okay? So, and as the reserve instances were as, and as the reserved instances started to expire, the bill kept collapsing. So they put all the reserve instances, they did all that stuff on there, but they said since they've deployed, okay, they're up and running over a million and the savings are coming in. Well, that's where- 3.2 million dollar cloud budget for 2022 and they're saving money. That's where, okay, so what does that mean for top of the pod here? What does that mean for the market? And is that a good thing? There's six months into their exit off the cloud. So isn't that a good thing for the market? Yes. If you're going to be able to use AI to drive that kind of productivity and, I don't know. And here's why the repatriations in this kind of use case that's interesting. I think the net new use cases are going to be higher on premises. And I've said that and before I'll say it again, I'll stick to that position. But when you have a premium in the cloud now, the value is going to shift because Amazon's value purposes is that we'll help you with the undifferentiated heavy lifting. What this guy is basically saying is that that's not heavy lifting for me anymore. I'll make more money by hosting my app on premises and multiple geos because their app doesn't need more provisioning. The use case of provisioning on the cloud is over. The higher level services, they weren't getting value out of those high level services for the premium that they were paying. That's his premise. So I get it. However, Amazon has other things that are heavy lifting for companies. Exactly. Exactly, exactly. So I mean, look, right, like contact centers, like what they're doing in Telco, what they're doing in CDNs. Yeah, so the value shifts, it's just a shifting of the curve. So even though it's all raw, raw, people getting all wrapped around this, it's just good business because Amazon has essentially taught everyone how to do this. They've got Kubernetes, it's all cloud native now, so they can operate. Best of breed, open source, cluster management, great hardware from Dell, for instance, they have their good storage. They architect their own stuff, build their own gears, swap out broken stuff, a little SRE action, little DevOps on premise, no problem. Yeah, but it's a bit, but look it. They don't have a need for the cloud scale. Right, but if you need scale, right, and you don't want to load a storage array onto your loading dock, and then you're gonna be still doing cloud, it's not like repatriation. I mean, the repatriation talk is still, anybody who says repatriation is gonna start eating into cloud, I just, I don't buy that at all. Now, the interesting thing is, so Jassy uses this metric that only 10% of everybody repeats it, only 10% of IT is in the cloud. Now, that's true. I think it's factually correct that of the 4 trillion in IT, probably only 10% of that, 400 billion is in the public cloud, the big three or four. The question is, how does an Amazon go after that? Because they're really an infrastructure player. And if you narrow the market into infrastructure, the lot more of the spending is in the cloud. I would say 30 to 40% of the workloads are in the cloud and that's gonna continue to grow for sure. But so that begs the question is, okay, how does Amazon expand its TAM? Do they already have the strategy to do that by going after things like contact center and new industry transformation and telco and applying LLMs for the industry? I would have liked to heard more of that at re-invent. So is the market already so huge for their existing platform or do they have to really affect a TAM expansion strategy? Whether it's going up market with software or into, I think industry transformation is more likely identifying other heavy lifts that they can simplify where the industries just need streamlining. So it's a huge market. There's no lack of TAM. We're winding down on the pod time here, but one last story just popped across my phone here, breaking news. Apple explores AI deals with news publishers. The company has discussed multi-year deals with worth at least 50 million to train its generative AI systems on publishers news articles. Hmm, hmm, okay. No, I must have got the memo. Oh, so, you know, what does that mean? So Rich get richer. Apple's gonna use their AI and we saw an article where blue business insiders owner, Alex Springer, cut a deal with the open AI to get in their index. Here it is, Google News, Dave, it's coming. You know, AI. So the work we're doing with the CUBE AI is right in line with this. This is great for us. So, but it might hurt journalism. Again, the question I'm very skeptical on these big platforms with journalism that has not ended well for journalism companies to bet the ranch on the big tech platforms. Twitter screwed everybody over. Facebook never delivered any revenue. I don't even know what Apple does for revenue. It's just tiny mouse nuts and no user closed loop. They're sucking in the entire estate into their apps. I do think, having said that, I mean, I do think there's always gonna be a market for good journalism and good analysis, like real analysts and real journalists. I think the problem for the journalist is like it's hard to make money. I mean, I know our friend, my really good friend, and you know, I'm Charlie Senate. I mean, I think what he does and people like him are really unique, you know, bloggers don't do that. I mean, some- He's a non-profit, though. He's a non-profit. You know, I know, I know. I'm just saying that type of, you know, they just gotta knows for the story, you know, the big, big cost overrun, you know. It might be sad though that the journals has to be a non-profit. You gotta, it's a non-profit entities now that are, it's just, anyway, we'll pick that up later. Dave, it's end of the year. I want to just say thank you to everyone who's listening to the pod. We appreciate all the feedback. Go to SiliconANGLE.com. And also, Dave, congratulations to you on your breaking analysis podcast, which's got almost a million downloads this year. Congratulations. Not quite, but yeah, thank you. That's about three quarters of a million. Yeah. Great franchise, great brand. You wouldn't expand on that on the research side. And the CUBE research team is growing, expanding. And we're gonna see a lot more new things coming out of the CUBE. We're gonna see a lot of exciting 2024 for us. And our pod, we're gonna mix things up or 41 episodes in. And we're gonna do a review, a look back on what worked, what didn't work. As we said, we're gonna iterate this podcast, keep changing it. So again, keep DMing us and emailing us on what you think we should do or not do to make this podcast better. Better insights, better experience. You want more entertainment. I had someone this week tell me that they wanted more entertainment. Dave, they think you're funny. They think I'm funny? No, I'm not kidding. I'm always saying that. They want us comp, they want us... You're the life of the party. Okay, I think it's fun when we argue and people get to see us like going at each other in a friendly way. No, I think there was a section that we should do a meme section. Anyway, we'll get to that next time, but have a great Christmas. Hey, Merry Christmas, John. Merry Christmas to everyone out there and see you next time. Happy holidays. See you.