 Hello, welcome to theCUBE Pod episode 14. I'm John Furrier with Dave Vellante. This is our podcast. We look at all the stuff we're talking about every week, our events we're going to, top trends, what's in the news, what we're analyzing, and the new data that we're getting that doesn't hit the presses yet. We talk about it. They'll probably hit SiliconANGLE.com later in the week or weekend or next week. We take a look at what's happened and we look at what's happening now and what's going to happen next week. This is with theCUBE Pod. Dave, great to see you, episode 14. We said let's get the 20 episodes, see how it goes. Great to see you. I'm here in your home turf in Boston. We can't have been glad to see you again. It was awesome. You stayed after Red Hat Summit. I came in from Vegas. It's been a good week. Mongo. I mean, we're going to talk about that. MongoDB. I don't want to pop the stock and video. The stock was up 80 points today. 8-0? 8-0. 82 points, it's 28%. Wow, it shows the benefit of being conservative. Well, I think they probably tightened the expense belt to get the numbers, but there are the bellwether and the proxy for developers because MongoDB database also uses a lot of cloud compute, a lot of EC2 as we know. So if Mongo is doing well, that's the canary in the coal mine that the developers are continuing to drive the innovation cycle. And that's going to be very interesting to see the jobs numbers are out and Vidya hit a trillion dollar valuation. We did that last week. We did the breaking analysis on it. I think we were the first ones to report on that. But like I said, it's going to increase more opportunity. Barriers to entry, AMDs wants a piece of the action. I just interviewed Matt Garmin at Top Executive at AWS who ran Compute knows that Silicon game really well. So he's pumping up their GPUs. I mean, NVIDIA AI is happening. Again, more AI, Broadcom popping after hours. Their stock's up VMware, trading at a discount. Mongo popped, Zscale has shown acceleration. Dell beat, but they had their worst numbers and three straight numbers in history of BC sales. They declined, but they beat. Right? HP soft, HPQ super micro was soft. HPE cited softness for service. Broadcom was really interesting because after hours, it was all over the place. The company hit its numbers and then Hoctan gave some AI chip guidance. Basically said last year, it was 10% of their business. And I was 15%. It's going to be a 25% of their business. And people were like, oh, at first, they said that's going to detract from the existing business. So they sold the stock and pre-market. And then afterwards people were like, kind of got a reality check and said, no, no, that's a good thing. And then the stock went up today. So people are just confused. So of course, Charles Fitzgerald, Fitzy who we love always puts out great tweets. It's just his tweet. Dell quote, servers and networking unquote, revenue down 24% parentheses down more than their PC client business. But, but, but cloud repatriation. Smiley face, laughing, crying, smiling. I wrote, where are the repatriates? And he wrote, not buying their hardware at Dell, it seems. Dell numbers were down. What's the repatriation look like? I mean, does it have anything to do with repatriation? You covered Dell. Is their business down because no one's repatriating or is it cyclical? What's your take? Well, I mean, the business sucks right now. I mean, servers and storage are soft. I mean, that's really the bottom line is that people, I think what happened, John, is people came out of the pandemic and they bought servers and storage to replant and networking to replenish the headquarters. Now Cisco was interesting last month because Cisco was chewing through its supply chain. So it had unbelievable, it had great growth. Like top line growth at 12%, but that was because it was chewing through its backlog. And then it guided really conservatively. So, you know, it's really, really mixed right now. It's no longer a supply problem. I mean, Dell can get servers. They can build servers. Storage being done. Pure storage did okay. NetApp did okay, but not great. I mean, storage is not a great business right now. So it's soft. And I think that's again, what happened is replenishing the HQ and now people are taking a pause because, you know, they don't know what's going to happen. There's a lot of uncertainty out there. It's a weird time of the year, right? Everyone's graduated from high school, middle school, college. See a lot of graduations on Facebook and in social networks. It's summertime, pre-summer. June's going to be big month of the cube. We got massive events going on. We got Cisco Live, HPE Discover, Snowflake Summit, Databricks event, which is data plus AI. We got AWS Reinforcers, which is their security show. I just think that there's so much going on around AI. Everything's an AI show now. So it's really a weird time. The public's afraid of AI. And again, more news about banning AI. So there seems to be the two schools of thought going on right now about AI, anti-AI and then pro-AI. So I find that really fascinating that there's actually the emotional response of AI into the world right now. So it's just a very weird time. Well, Dell was interesting last week. So Dell had this announcement of Project Helix where they did a deal with NVIDIA. And it wasn't the first announcement that they made, but it was up there, it was prominent. Other people were saying, geez, they kind of buried the lead. At the same time, you saw a red hat, right? A red hat, right? With Ansible, it was all about AI. And that was pretty prominent. I don't know, were people accusing them of AI washing? Not really. Not really, right? So it seemed deemed as very positive. And then we just had Zia Scaravalla in here prepping for Cisco Live. And he's thinking, well, they're not going to do a big huge AI. And that's where just big AI is going to be everywhere. So they're not going to try to hook their wagon up to AI. It's very confusing right now. I think AI is going to actually expose our super cloud narrative is to be true. I think, I mean, basically my feed is like, AI is going to kill us. And then the next set of, how to integrate into your life. So infusing with AI and then it's going to kill you. Super cloud is definitely. I think AI is one of those things that's horizontal. It's going to be all developers toolkit. Software developers are going to use it. And I think it's going to get abstracted away as John Chambers said, voice activated, auto coding. Humans will be involved, but it will be a developer software thing. And I think that's going to be where the value is. It's going to be embedded in all apps. And the question is, how does an entrepreneur or how does that next Airbnb, that next Dropbox, these are companies that made their business successful because of Amazon and cloud. What's that next generation startup look like? Get the product market fit really quick with AI or whatever. Then they got, what do they do next? What's the path? Or is open AI that startup? The path is, I don't think that's, I think they're more like the browser. Not the website. You think they're Netscape. I think they're the Netscape, but it's not the startup that invents something. So say there's a healthcare app or say something in public policy. If you're the startup, it was like the old way is get product market fit, sell it and scale it. Get go to market. Now what do you do? You got to get into the cloud, do like a snowflake did, get on top of the CapEx, be a super cloud app or a super app and then double down on everything, right? Get the success and scale it. I think the old way of getting press releases and doing like startups is over. Well, I do think that the, as we talked about, incumbents are going to benefit from AI if they can apply it. And there's going to be some company that comes out of the woodworks with some new business model like Keyword Search that people will poo poo at first. The VCs will say, oh, that's a stupid idea. And then they'll come through the side door and blow people away. I think that's how it's going to go. I mean, I hope AI is interesting, right? I mean, it's really, you know, use chat GPT. I do too. We use BARD. And it's, you know, when you dig into it, and you research it, it's about predicting the next word and taking the internet and the human condition and figuring out, okay, how can I write sentences? But I think the next step there is going to be able to talk about guardrails. You and I had a conversation. Like, Dave, what does that even mean? I actually think it means to train the system to when it doesn't know, to not hallucinate and to ask questions back of the prompter when they don't know, right? So what do you mean by this? Or what do you mean by that? Or I don't know what that means. Or I don't have the answer as opposed to like Jason Calcanis started the cube with Dave Vellante. I mean, that's like, what? Well, I mean, it's all dismiss information at that point. I mean, technically. But data is the control point. I mean, you have your own proprietary data, you put it in. I mean, if you're a startup, what do you do? But seriously, you get product market fit. What happens next? Well, you go sell. I mean, that's... Where? Like hire salespeople? Go to the marketplace? Well, that's a good question. I mean, what happens to the BDR? Look at Datadog. No, seriously, what happens to the BDR? Does the BDR get disintermediated by AI? It's a big conversation. I think AI will change the lead generation, demand gen cycle. I think you look at what Amazon's doing, Azure's doing and Google, they all have marketplaces. If you have software and you're a software app, you want to be in these marketplaces because they have sales right there. They're embedding in contracts. You're part of the feature of the next thing. Or if you want to be monolithic or old school, you go standalone, you raise a bunch of VC money, and then you hire an army of salespeople, BDRs and inside salespeople, and you go. Is that, that was the fast way. Right, but so... No, is it now the slow way? What the BDR do? The BDR, he or she would send out emails, so troll LinkedIn, send out emails, right? Well-written emails, and then follow up. I mean, how much of that can AI do? Not zero. I think human plus AI is better than AI, as we've been saying. I would look at that as the more productive, I think the BDRs might be the new reps. But let's say, oh, interesting. But so, okay, let's say you were going to hire 10 BDRs over the course of a year. And now you've got AI, and you're actually using AI well, and you know how to prompt AI, and you can write good intro letters, and you can have software that trolls LinkedIn. How many BDRs are you going to need? Five? It could be a new dynamic. One, remember the 10? What's your guess? You think you can cut it in half? Yeah, definitely easily, easily. I mean, what percent of your time, when you're doing mundane tasks, does chat GPT save you? A lot. I would say it cuts mine down at 40%. Yeah, it's a productivity opportunity if you're editing. Well, that's not relying on it. Well, right. Well, in our world, we do a lot of editing, right? Especially with transcripts, you take the transcripts that we have, and they're auto-generated, right? It just makes you more productive. So that's easy though, right? That's not, we're going to look back and say, okay, this is like, we always joke about it, it's just like dial-up. You know, so there's CC mail, right? So news happening, Amazon workers walk out amid layoffs citing concerns for climate. Basically three days, they ask them to come back for three days a week and they're all walking out. Because of climate? I don't know, I don't know. No, no, they're coming back to the office. Right. So they're revolting. They're revolting because they need to come back to the office or they're revolting because it's bad for the environment to commute. Well, the concerns about rising carbon admissions, office policies, all that above, and you know, Amazon's going to be carbon neutral. They have a target date that's in visibility. So I think it's just a red herring of people are afraid of AI taking their jobs and then coming back to the office, I think people want to work at home. Yeah, I think that's really more the issue. Did you buy that it's sort of climate change? That's the issue? Or is it they just don't want to go back to the office? I don't know. I didn't read the article. So walk out of the collaboration between the Amazon employees for climate justice. First of all, justice is in the rats. That's definitely an interesting term. It's scary. Amazon employees for climate justice and an informal group of employees who oppose Amazon's mandated return to the office. Employees say recent layoffs in the mandate, which increases admissions as workers commute, have left them questioning whether Amazon executives are leading the company in the right. That is so laughable that I have to drive now. I'm going to be polluting the environment because my car's on the road. Do these people eat hamburgers? Amazon, do they understand that large language models that take up most carbon and Amazon's actually has carbon neutral footprint and that's actually good. Amazon employees for climate justice. When you see that, when you hear that word, what jumps into your head? Amazon employees for climate justice. Big red flag to me. And it's like, I think they just want to work at home. A group of people having beers at home just saying, hey, let's have some climate justice. You think they're shooting the breeze, having group meetings? Oh, this hurts. Do we have to talk about this anymore? I'll say that for my rant section, but I wanted to bring that up. It'll definitely be in my rant section, but the big issue is the AI and the GPUs. That's the real, I mean, meaty story right now is GPUs are tight, Nvidia stock is up. You and I were talking before we came in and yesterday that they got a hold on the GPUs. Yeah, well, I mean, you got Nvidia making GPUs. You got AMD's making GPUs and they're going to come out with a new line. You got Intel actually making GPUs, but like that one practitioner told us, they're subpar, but we use them anyway because we need competition to Nvidia who's gouging us. But I mean, you can't blame Jensen for gouging because they can't make enough of them. So it's supply and demand, right? They're going to take that, but they are. They do run hot. It was interesting at Dell Tech World, John. You've been to numerous Dell Tech worlds and EMC worlds. When you look at the ecosystem at Dell Tech World, you know, there were a lot of companies doing like cooling tech, like liquid cooling. We had some on theCUBE, it was actually quite amazing. It's back. It's like, remember the old mainframes? Thermal conduction modules, they had like, you know, pipes and they were pumping in, you know, cold liquid. So that's back as you got to cool these things down. The energy is a huge issue here. I mean, it's consuming a lot of energy and, you know, Bitcoin was mining was criticized, but, you know, I mean, you got to just focus on creating alternative forms of energy and that's what these big companies are doing. And so you're not going to stop progress. It's just going to keep beating, whether it's coal or nuclear or solar or wind into the system that's going to use energy. And so I don't see that as being a blocker. I see it's something that the industry's going to have to deal with. Okay, some news going on in SiliconANGLE.com. Of course, all that is hitting Cisco buys armor blocks to beef up cyber-screwed generative AI. So Cisco lives coming up again. This is an AI story. Will it be in there in security? Blink co-pilot shows how generative AI can supercharge cloud security operations. A first story from one of our new writers freelance of Tim Keary. Coming on Cisco, you know, hired Tom Gillis to work for G2 to run their security business. And I think what they need to do is to build a security super cloud. I do, I think they've got an opportunity to simplify across clouds and on-prem and create a common user experience. And they're in as good a position as anybody to do it. You know, Cisco better than I do. They're good engineering. And if they really put their resources on it, well, they're an engineering company. Well, the networking is really the important part right now. If you look at multicloud where that's going, I think I've always said Cisco is perfectly poised for multicloud. They are super poised for us. We define super cloud, which is the fabric between policies and tying that with software-defined networking and software-defined data center for software-defined cloud or cloud operations. If you look at Red Hat Summit, for instance, we were there, it was incredible. Look at the open source software work they're doing as well in the open community. Networking is prime for innovation with cloud operations, with edge, latencies involved. So you're going to see more and more networking action going on, both in the large cloud operations and on-premises and edge. It's just, it's a dream scenario for Cisco because with all the AI coming on board, there's going to be more need for these apps with low latency data transfer. So, you know, I think there could be a generational kind of renaissance for Cisco. Let me ask you a question, because we've had Steve Mulaney on a bunch of times, maybe Atrix, he talks about the networking super cloud and sort of, you know Cisco obviously very well, you know Cisco, is what he says, he implies that Cisco's, you know, the old guard, he kind of uses the anti-jassy depositioning, but I'm inferring from what you just said that Cisco's actually in a good position to build the networking super cloud because they got the install base and they got, if they can get their technology. If you look at Microsoft, when Microsoft, you and I were, I remember when you and I had a chat, like Dave, I can't believe Microsoft's at 26 bucks a share. What are they at now, 400, whatever the number is? Check the number, it's huge. 2.4 trillion. Their turnaround is to me the exact play that Cisco will and should and will do, most likely in my opinion, I don't know, haven't talked to Chuck Roberts about it, but if I'm the CEO advisor there, I would basically be advising him heavily to double down on their core enterprise install base, bring value immediately with cloud to the customers and nail multi-cloud in a way that is so epic that their customers won't even consider switching costs. So when you and I met, Microsoft was $26, $27 a share and then it peaked at whatever, 400 or something. It's still, now it's 335. Look what happened, bye-bye Balmer, bye-bye proprietary, open Brace Linux, open sourced, open compute, they started open compute, they started doing a lot more open stuff. They took their scale from MSN, made that their base cloud and changed the airplane out of 34,000 feet, as we say, and brought in cloud and they whitewashed everything as 365, which we're critical of, but that was a step in the right direction. I mean, they were, they're constantly upgrading and peddling as fast as they can to get to Amazon web services. And the difference between AWS and Microsoft is Microsoft has huge install base and generations of serving enterprise customers. And enterprise customers sometimes just want the numbness of what's comfortable, they're comfortable with. And so as a new trend comes in, if Microsoft can deliver good enough, then that's going to put pressure on Amazon web services to have a more compelling value proposition to handle the switching costs, which is uncomfortable or is price the issue. So I think Amazon has done an amazing job of being first with cloud to be the swipe to credit card shadow IT and then workloads in the cloud are amazing. So clearly Amazon's number one, they're not, they're not hurting, but Azure has the install base. So it's a race. It's a race for how fast can AWS take territory on Microsoft's customers and how Microsoft can hold the dam from breaking and keep the customers there. And that's the game that's going on right now. And we've got the history with Microsoft and AWS and the years in the industry. That's what's happening. And you can squint through all the PR that Frank Shaw's doing and like all the pomp and circus open AI, which is a $10 million marketing campaign to make Bing look like they have AI. So, you know, that's not real AI from Microsoft. That's just a great move, which by the way, I love that move just saying that's not what Amazon. I go back to you, Jerry Chen and I, I think the first reinvent we ever did on theCUBE, we had Jerry Chen on and we were talking about, you know, do you think Amazon's ever going to go up the stack and Jerry, I think nailed it. And you did too. Basically said, look, Amazon's play is to provide tools so that companies builders can compete with the SaaS players, not necessarily build their own SaaS. Now they will do that in certain, you know, examples like they do with, you know, Connect and Call Center and things like that. But their strategy is not to mimic what Microsoft is doing or what Google's doing with Workspace. Their strategy is to allow their ecosystem and partners and customers to actually build those. And so the reason I bring that up is because I think when you think about generative AI, that's sort of similar approach that they're taking. Right? I mean, Amazon and Microsoft being the Amazon because they have the most cloud customers, serves virtually every kind of company and industry. From start-ups they've evolved to be that big. And so when you look at an enterprise versus say an ISV, there's different use cases of consumption. So Amazon Web Services has that problem where every time, probably internally, I can imagine the internal dynamics where there's an opportunity there, there's an opportunity there. Everything's an opportunity because they're leading. They got to figure out how to do both. They got to provide SaaS products to enterprises that want turnkey. And they got to provide the building blocks to builders. Right? And they have everything up and down the stack. And I think the work to do and the Silicon, I've always said, has always been the most important, the physics. So, you know, Adrian Cockroach, he's come on, came on theCUBE last year and he said, you want to block the Legos, build your own and make it hardened, good platforms, a lot of work? Or do you want to have the toy? That's pre-built for you, the Legos are the pre-built. But if it breaks, you've got to throw the Lego away and get a new one. So, you build your own, make it tight, or you just get one. And that's when it breaks. What's his point? Is that Amazon's doing both or that Microsoft is? No, Amazon has both. Microsoft is the ultimate Lego prefabricated service. They're a software company. So they're used to serving, you know, prefabricated software, general purpose software to companies. What that means is, the building blocks allows a team saying, I want a real platform and cloud. I'm going to build it and it's going to cost more, take more time, but you can always fix it. That's Amazon. That's Amazon, but now the higher level services that we covered last year is a lot more, okay, I got the house ready to go. It's already built, but it breaks. It's still not SAS, right? I mean, it's still not like, I mean, I guess it is up to the data level. Well, right now, if you want to use Code Whisperer, that's a SAS product. Yeah, yeah, sorry. What I mean is it's not like productivity software. You know, they're not doing Office, right? Well, exactly. They're not doing Google Docs. Yeah, but they can make software like Call Center, right? It's like. That's what I'm saying. Certain verticals where it makes sense for them to build solutions, they will. But do you think they're going to start mimicking Microsoft's software portfolio? I think that would be a mistake, actually. Well, that's what they announced at Reinvent. I think that's going to be the challenge. Wait, wait, what do you mean? What did they announce at Reinvent? Well, Adam Slesky, when we did the exclusive interview, he was essentially saying, look it, we're going to be moving up the stack and providing higher level services. That's code words for things like Connect. But he couched it as very industry specific. Yes. Which makes a lot of sense to me. You know, you want to do is identify industries that are ripe for disruption that are very unproductive or have a lot of waste and then reinvent them. I mean, that's kind of what they did. Bedrock is like AI, open AI. That's a service that runs on AWS. Yeah, it's a managed service. Right, okay. It's got building blocks. You can take first-party models, build the third-party models. But they also integrate. Versus, say, Code Whisperer, which is helping a developer, that's that solution from Amazon. That's a SaaS service. But they also will integrate with third parties. You can use Hugging Face, for example, or other generative AI products. I asked Matt Garmin directly today in my interview with him, which will be published next week, and is, do you want to be offering open AI? On Bedrock, he said, we want to offer every model available on AWS. Yeah, why wouldn't they? But do you think Microsoft's going to allow that? Well, Microsoft doesn't own open AI. They only did a $10 million deal with it. According to some people, they basically, oh, dual, by virtue of that $10 billion, they essentially control them. Every customer that I talk to privately before I come on cubes, or a CISO or CIO, and the insiders of the VCs all say they want open. They want open AI. And remember, Andy Jess used to tell us that Windows workloads is one of the most popular workloads on AWS at some point. Remember, iPhone came out, you could only get it on AT&T, right? Mm-hmm. Right, so maybe it's like that. Maybe for a while, maybe that's, I don't know if that's part of the deal that Microsoft struck with open AI is. I mean, I think the big clouds like Amazon are going to look at the general AI apps and look at how they can differentiate on privacy and IP. Because remember the whole thing about if you put stuff into open AI or chat GPT because part of their corpus. Right, well, so you got IP leakage that you're concerned about. Also, costs that we were joking about repatriates, but costs are a factor too. I mean, if you're running supercomputers with half a million cores, you going to run that in the cloud? I mean, no, honest question. Are you going to? Is it cost effective to run a half a million cores in the cloud? Depends. I don't know if they're in a steady state, maybe yes, maybe no. I mean, some, one, I would say maybe no. Well, then you just, you don't have to over provision. Over provisioning was the big problem with data center capacity planning. Black Monday, Friday surge. These large language models are running essentially on supercomputers. And some of these supercomputers are massive, like physically massive. So if I'm a university or a national lab and I've got a supercomputer, I want to show it off so I can raise money. Come on in, look at my supercomputer. Look at the lights flashing. Right, so you can't do that in the cloud. So, I mean, Fitsi's right, the cloud growth still significantly outpacing that of on-prem, he's right about that. But I think there's, it's going to be interesting to see where these models actually run. OpenAI is running a lot of its training in Ohio and supercomputers. So here's a story. ChatGPT wins over doctors in answering medical questions. A study published in JAMA, Journal of American Medical Associates and compared two sets of written responses to real world patient questions. One set written by physicians. The other by ChatGPT. Panel preferred. ChatGPT, 79% of the time. Is that a blind study? It had to be, right? Well, we've asked many times in the cube over the last several years, when is it that AI will make better diagnoses than doctors? And I remember we were interviewing that guy from KPMG at IBM thinking. He's like, it's already there. I don't know if it was or not, but it's getting close. Well, I mean, I think you've always said it. You know, ChatGPT will make a great talent, greater, and a mediocre talent, great. And so it's all a function of the human. And this is one of the things we've been covering is that the humans plus AI is better than AI. Constantly the case. All right, Dave. Well, we've got a lot of gigs going on. What else is that you're covering? What's on your radar? I got Cisco Live coming up. We'll be there. Me and you, the pop-up cube. We got the new pop-up cube. Yeah, so Cisco Live, I'm excited about it, right? I think it's interesting to see what we're going to see next week. Again, I think Cisco's got an opportunity to simplify that. The problem that Cisco has is so complex. I mean, you know this. So they've got to simplify the portfolio so they can keep growing. Sales people have a hard time sort of absorbing all that. And then, yeah, in June, we're on the road every week. We got reinforce. I don't know what that's going to be like. I'm home for four days this month. Anyway, we have a lot of, we have super cloud. I think I can remember the more crowded cube month in 13 years. No, but the two months of May and June I've seen is more crowded. But as far as a single month, everybody's trying to get stuff in before the quarter. Every week is we got something going on. And doing a lot of work with George Gilbert around looking at the future of data and data platforms and Snowflake and just in prep for the Snowflake Summit, I'm excited to see what they're going to be announcing. We're doing Mongo. I mean, Mongo, like the hottest stock on the planet right now which is kind of awesome to see. I like those guys. Let's unpack that Mongo. We're going to be at their event this year. They've been kind of bunkered in. I felt like they've been kind of retooling, keeping their expenses tight. You know, you would have thought that the Atlas stuff which is their managed cloud service was what drove the quarter, but it was really a combination of conservatism. They had a decent traditional license growth, like really good actually, and good expense control. And so I think it's just, I think a lot of companies, I mean, even Snowflake has struggled with its visibility and its guide. And so when Snowflake's notoriously conservative and Scarpelli's really good at it, I think Mongo was so ultra-conservative and they blew that away. I guess the same within video, I feel like in video was just, maybe not just ultra-conservative, I think they might have been caught off guard. They beat by, you know, $600 million. I mean, that's just a massive beat. So it's very uncertain right now, John. I've never seen a time when visibility was this uncertain. I've seen times where visibility is really bad. Remember Ed Zander after the dot-com bubble said, does anybody want to buy a server? Nobody had any, you know, inkling that there was going to be sales. Everybody knew, they had good visibility that the market was terrible, right? Today, it's like, some people say, no, actually we got pretty good demand, right? We're going to beat and raise service now. Yeah, no, demand's really good. Others, you know, Dell's being super conservative. The repatriates are being very conservative right now. Software's mixed, right? Software's really mixed right now. Yeah, and I think you had a good breaking analysis on this where you looked at where people are they holding because of the spend? Are they pausing? Are they shifting? I think a lot of companies saw the headwinds and said, let's bunker in. Even the big companies like Mongo. But they, you can't stop, progress. And the AI wave is creating more action under developer front than ever before. We were reported, you and I reported, and I was at the KubeCon and then open source summit. In the past month and a half, the open source models, the lightweight third party long tail models have been booming with growth. Innovation, new projects, lightweight stuff too, not heavy models. So like, you're going to see this long tail power lock, the prominent models that the big guys have first party and you have third party. So Amazon's got, you know, first party models like Anthrop Topopic and they have then the third party through this bedrock application, which is their version of OpenAI. OpenAI is doing the same thing. They've got ChatGPT, which is their primary predominant model. But then you're going to have a long tail of models and like you put it up, how do you make put that together? So I think we're going to see, remember the web 2.0 mashups, you know, API maps, you know, overlay, you talk about Uber, all the time. Mashup camp, remember those guys, David Berlin? David Berlin. Doug Gold. Yeah, mashup, mashing up. I think you're going to see same exact thing happen with AI, you're going to see the confluence of the prominent third parties on the big clouds and then a long tail of open source and it's going to be a developer dream to mash these up and you'll see stars come out of the woodwork. It'll be easy to see if the big guys will enable it, whether Amazon will enable their tried and true fashion of putting apps at the top of the stack and not competing and somewhat friendly competing on the big stuff. Microsoft wants to own everything. So the question is, what's the better play? And some say it's the Apple versus Android. iPhone's doing great, proprietary, but well, you know, in periods that we've seen a lot of these disruptions and what often happens is, you know, the new stuff, the existing incumbents, it's relatively small in the grand scheme of their, you know, multi, multi tens of billions of dollars of business. So they can't pivot fast enough or take advantage of fast enough to move the needle. You got certain players that actually like Nvidia in this case, like Cisco back in the IP days, you know, Cisco still hasn't achieved its peak, gone back to its peak valuation of 2000. Remember, Cisco was, I think at one point, the most valuable company on the planet, right? And so what you typically see is in the early days of this disruption, a select group of companies really can take advantage or is perceived to be taking advantage and the stocks go crazy. The incumbents are trying to figure out, okay, how do we take advantage? And that always takes more time for them to reach an equilibrium. But explain this one, the Dow was up over 2% today. So the big thing was jobs report, right? Jobs report came in hot. So people are thinking, okay, you would think, oh, well, that means the Fed's going to raise rates. Okay, but the market loved the fact that, oh, maybe we're not in a recession. Okay, S&P was up, the NASDAQ was up, tech was up. You know, so I think the market's really confused. And then watch Monday, they'll realize, oh, wow, that means the Fed's going to raise rates again and then the market will go down. So it's very hard to predict right now. And I think, I do know this, I think you and I both agree, this AI wave is real and the people that figure out how to apply it to their business to cut their costs and maybe come up with new business models are going to thrive. Yeah, I mean, I think it's entrepreneurial dreams and the area, if you're a startup and you're an entrepreneur, everything you're reading online is going to be outdated. You're going to see new playbooks emerge. I think it's a lot like the web, there'll be a bubble, we're seeing it now. And the question is, because it's so software embedded with AI and it's so horizontally scalable and every vertical, what's going to be the playbook once they get the code, the product market fit? What happens next? Because there's a huge chasm that hasn't been there before. Before you'd have to go through, at least an accelerated compare to the old model. The old model was, you get a PowerPoint, show a VC, they give you some seed money, you buy a bunch of servers, put under your desk, put one in a data center or build your own and then you code away. Six months later, you get a prototype, get another round of funding. And then the web came on with the cloud and be like, code a prototype, get some customers, show some traction, get some VC money, upgrade the product, reduce the technical debt, hire a go-to-market team, get customers and run as hard as you can. That was the fast way, now it's faster. The question is, you get the product, you put models together, you jump in open source, you grab essentially open source software, which would have taken you years to develop, man years. Another bunch of man years over here, just like chunk it up and you have code. And then if you get customers, then you can just put it in the cloud and the cloud scale kicks in. That's why the super cloud narrative is really relevant because you can then leverage all the cap backs of the cloud providers to get scale. And once you're there, the money just falls out of the sky because you just skip A and B rounds, you go right to the C round, growth round. So I just think it's going to be a massively new, celeranted, new venture creations trajectory. It's going to be something we've never seen before on that product market fit, highly accelerated, almost anti-hockey stick, it's just straight up. There's no curve, it's just straight up. In terms of actual adoption, you're saying. Yeah, if you hit this wave hard, you're gone, you're good. You just don't hold on to the rocket ship, right? So it's like, don't fall off. So the only way, and that's the key. And again, the cloud is so successful. You know, you get something on Amazon that hits and if it goes into the marketplace, you just accelerated three years ago to market in a month. So, I mean, that's just a whole nother power dynamic. It's definitely more internet like than it is, you know, the PC. The internet, remember the buzzword was paradigm shift? I mean, for years after the dot-com boom, we didn't even use that term because people were like, oh, but it really was a business model shift. It was like, you know, Adam-based businesses getting totally disrupted, you know, sorry, bit-based business getting totally disruptive. Adam-based businesses sort of unclear, but they got disrupted too. This seems to me AI just applies everywhere, right? I mean, you can apply it to make your infrastructure run better. You can apply it to reduce your headcount. You can apply it to create new products faster as you were just saying. I mean, there's just like an infinite number of use cases. If I was a VC right now, I'd have my own scale lab, basically, because remember when the old days of software, when you get software put on a shrink wrap package, you put on shelves, people would buy it. The folks under the age of 40 might not know this, but you have to buy software load on your machine. They didn't know if it was successful until they had, well, we didn't sell all the inventory. Then the internet king had SaaS, where you had SaaS analytics, KAC and acquisition, long-term value of the customer, all those KPIs and all the SaaS gurus defining success, land, adopt, expand, you know, you had metrics and that took some time to squint through the numbers. But you started to know if it was successful if people used it, if it had recurring revenue. And you could see it right in front of you. And that would bake out over her time series horizon. Now it's like product market fit. Boom, you know right away, bang, much faster, much faster. The most adopted app in the history of humankind. It wasn't even close. And remember before that, some of the other apps that were successful, like Instagram, boom, once it gets traction, I think AI will have an Instagram, WhatsApp, chat GPT-like trajectory on adoption once it hits, because I think it's that compelling. So I think, you know, looking for startups that are going to come out of the woodwork for deal flow from a VC, you got to get in like really early, because the early stages where you can just really leapfrog everybody, because that means, if you believe what I said it's happening, that will then accelerate that whole A rounds controlled by the growth. So you think about, again, we like the students of history here, but you go back to when the PC came out. I remember actually when we were at AC, because we were very quantitative, we actually had to go to class to get trained on how to use Lotus, right? We had a trainer and it was like, okay, we had weeks and weeks of training, slash, bio, retrieve, paint, copies, and it was like, this is how you do it. And we had to go deep. Okay, internet. What was the killer app on the internet? It was email. Yahoo. Right, email and search. Yeah. You really didn't need- Yahoo was the killer app right out of the gate, because it was a search engine that allowed you to click on things. So it was email, not everybody had email, right? Hotmail came out, web-based email. You didn't really need training, right? So adoption took off. You had AOL, you had AT&T, WorldNet, you had Hotmail. But I interviewed at the Red Hat Summit, the guys at Microsoft now for like 25 years, he ran Hotmail when it was a startup. He stayed with Microsoft after the acquisition of one of the most successful companies in mail. I still have my Hotmail account by the way too. Oh yeah, it's like a vinyl record, I'm gonna keep, I have all my AOL, my AOL account, I got my Hotmail account. You get an email from Hotmail, it's like, whoa, what is this? Chat rooms were kind of clicking. Yeah, vinyl records are up now. Throwback is coming, Dave. Hotmail might be- You got the Mac in your office behind you, the original Mac. I had one of the first leases. I told you that. Steve Jobs shipped it to us and nobody wanted to use it. I was enthralled by it so I started playing with the user interface. But I guess my point is that those killer apps were relatively narrow, email, communication, okay, fine. And search, you can search for stuff and that actually kind of came later. And then, but you think about generative AI, I mean, the use cases are insane, right? I mean, so you don't need training, obviously. You don't need any instructions. All you need is creativity, right? So, fastest adoption error, what was it called? The AI heard round the world. That's so great. It was definitely fast. So the other thing I want to get your thoughts on is I'm looking at my LinkedIn and there was a meme going around that looked at news since 1920 and it shows that every single error, everybody was afraid of automation. So this whole AI automation thing is complete, you know, garbage in my mind. 1920s, oh my God, there's wash machines or whatever that instrument was at the time. And then through the 90s, everyone was always questioning automation and jobs. It just never happened. Jobs are gonna, they go away, new ones create. So I think that's going to happen here. I think chat GPT and open AI and Amazon stuff and as apps come in, we'll eliminate positions in favor of better questions from doctors that could be done by machine AI, but AI won't replace doctors. So doctors will get better with AI. So the question is, I agree, but I think you're right that machines have always replaced humans. We've talked about this in the course of time, always. But this is the first time ever that machines are replacing humans or doing human tasks that are cognitive, right? We've never had that before. I think that scares a lot of people. So then the question becomes, okay, will it create new jobs? I think yes, it'll create new jobs. What kind of jobs? I'm not really sure yet. Chat GPT prompter? Well, prompt engineering is just a signal that- Prompt engineering. It's just a signal that you're prompting something, you're calling a procedure. You're essentially voice activating or text activating. But is that gonna be a title that emerges? Prompt engineering probably already has. No, I think it's called the job. But everyone's job will have some sort of, some sort of prompt engineering. For example, my interview this afternoon with Matt Garmin, I had list the questions and I put it in the teleprompter. And I did the read, I had some just bullets and how I like to do the teleprompter. I don't like to read the whole thing as a bullet. But it's a nice little sentence paragraph for the meat of the topics I want. It was like five or six bullets, I had a couple extras in there. I cut and pasted the whole thing, I said put in the chat GPT, do a teaser, preview. It wrote the most kick-ass preview post. That if I just pasted the video, I wouldn't even have to write a blog post. It just nailed it. I didn't have to add it one word because my little cryptic text was feeding the AI. I was controlling the hallucination aspect of it. There was no hallucination, I just said write this and it made it good. And if I put the video there, you'd read the text and everything in the text would be like me teasing out my questions with his answers in the video. That means I don't have to write it, send the video, more people watch the video. I did that literally in 10 seconds. Yeah, you got a great product out of it. So productivity, are you kidding me? So why do we need you to do the interview? Because I wrote the questions. Okay, because you were a good prompt engineer. I prompted the teleprompter, which prompted me, which then fed the transcript. Now when I get the transcript, which is now online, because we have the Qubectivity.atalake in our generative QBAI software now. So what I'm going to do, what's going to happen now is that's going to spit out his answers. Then I take all his quotes and say, write me a story about this as this preview. I merged the two together and it writes like a New York time-style story. That's hours of content, savings from productivity. I merged the preview, I got the video, I got the text, I merged them together. Now I got the preview promotion for the video. Then I merged my preview with his answers and I say, write this together and it just kills it. I love ChatGPT. I mean, I'm going to start, should we start pulling a byline, John Furrier via ChatGPT? I've put that in some of my posts where I've extensively used ChatGPT. I mean, on LinkedIn, whenever I use ChatGPT, I just say via ChatGPT, I'm transparent about it. I don't care. Why not, why not? I think people should, in fact, I want to have a site on a silicon angle, maybe it's a good idea or not, maybe we can find the resources, just a section on AI content. You mean AI written content? From our stuff, yeah. If we have all this ability to create more derivative works, why wouldn't we have an AI section? I'd be curious actually. You don't mean to covering AI news. You're talking about AI written content. As a user, I might be curious, find out, I'd love to look at angle.com, they have all the great news stories and enterprise tech, extracting the signal from the noise. I'd love to see what the AI thinks of it. If we create derivative works and better user experience, more content, if it's on point, I'd be curious. I don't give a shit if it's AI or not, do you? If it sucks, then we take it down. It's useful, right? Oh, I mean, that's another great way to use AI is to summarize news, right? So you don't have to read the whole thing. Give me the key points. It's like cliff notes. All right, Dave, we've got to get in the rant section if we wind down our episode 14. Well. What's your rant for the week? Well, the house passed, they came to an agreement with the White House, with the president to deal with the debt ceiling. I guess my rant is, it's not dealing with the debt. All right, all it's doing is raising the ceiling I think I'd like to see them either raise the debt or just stop with the madness. Just maybe change the law so you don't even have to go through this anymore. But I think, you know, I've been of the thought, John, that we actually have to deal with entitlements and we have to deal with defense. And you see, as part of this deal, they're going to cut nickels and dimes here and there. Of course, we're talking billions, tens of billions, hundreds of billions, but it's still in the grand scheme of things, nickels and dimes. But defense, up 3%, okay? Because that's politically, you know, suicide to political suicide, to cut defense. And they're not going to touch any entitlements, i.e. social security and Medicare. Those can't touch because people have earned those. And the fact is, unless you deal with those, you're never going to deal with the debt. So I think they've got to find another way, right? I think it's maybe, you know, recapitalizing, you know, the debt portfolio and figuring out how to take a longer-term view. But I just don't see a day. I mean, they talk about the debt ceiling, but it has nothing to do with the debt. And it's really not doing much for the deficit. So I guess that's kind of my mini rant. I think I'm almost, I'm kind of giving up on the government's ability to attack this problem. The last president to attack it was Clinton, whether they're Republicans or Democrats, they don't seem to care anymore. So, you know, at some point it's got to give, you know? I think smart people like Warren Buffett, I do, I don't pay attention when he says about crypto, but I do pay attention when he says about this, but he just, I don't think he understands it. But so not that I'm an expert at it, but I think there's some merit to the whole concept. Especially when the government's just printing, printing paper like this. So I kind of give it up. I think that we've got to find another way or just resign ourselves to the fact that at some point, it might just all blow up. And so damn the torpedoes just keep printing money. Well, I mean, I think there's going to be a revolution. You know my rant on that. It's not my rant for the week, but I'll just double down on my previous rants. I think we need new leadership and the younger generation needs to step up and lead. I'm happy to be involved, but that, you know, we could be, we need someone in their 20s. So I was 25 again, which I want to be. I wish I was 25 again. I'd be out there. I think you can do more now. It seems with the productivity and technology. That's not, it's not just a tech industry. More tech is everything. Tech is involved in all of our fabric of life. And it's going to be viewed that way. And the nerd should have a lot of power. The tech leaders should have a lot of power and the commissioners have a lot of power. And they should bring that power to government to reset and refact their government and how it works. Because the government's supposed to serve the citizens to provide liberty and freedom. Now, liberty and freedom is not a surveillance culture. So I guess I'm going on a rant here, but you want to live in a surveillance society or a free society with liberty? Well, if you want a surveillance culture, that's China. That's okay too. If you want that, you got to be China. You got to beat them at their own game. A lot of surveillance going on in our country. Yeah, well, people don't want, some people want liberty and freedom, which means you're being in your own house. You can say whatever you want. Well. Or, you know, pursue your dreams without any government intervention. I mean, we're living in a society that's borderlining where you could say something that might be controversial and you're going to get canceled. And there's a lot of backlash around this new culture of not opening up free ideas and free form of communication. So freedom of speech is, people use it as a punchline, but ultimately, there should be civil discourse and people should be able to talk. And without fear of doing that, when they say, oh, don't worry about, you're not going to get punished, they still don't do it. And my rant is going to be on this. It's not like Elon. Free speech. I think every American would say, do you don't want free speech? Do you want liberty and freedom? I'm for free speech. Yeah, constitution, but country's all about. But it's kind of evolved. Evolution is like, that's changed the things. They make it work, you know? Saw weapons, those rifles. You don't need AR guns. Come on, give it to those immediately. Anyway. Actually, somebody on our team has an AR. I mean, if you like to shoot, get an AR range and go rent. I was surprised and he explained to me, no, it's like, you know, we keep it locked up. Very careful with it. Right until someone gets the key and then boom. But anyway, my rant is on this Amazon story. I was mentioning it earlier about this. Amazon workers walk out of mid-layoffs, citing concerns for the climate. It wasn't the climate. That was a red herring. But the word Amazon's climate, employees for climate justice. You know, Amazon does a lot for the country. They do a lot for the carbon footprint. They need a lot of work on that. This is concerning how low morale is in the Seattle-based employee who walked out and spoke on the condition of anonymity because of fear retaliation. Free speech. Okay. Washington Post not putting down their sources. Again, someone close to the situation. It's just, I don't know how this works. What's your issue? What's your issue with that? One is, there's no quotes for any employees. There's no, they're hiding behind a firm. Some probably asked for it from Amazon employees for climate justice. This was on Bezos publication that this came out? Yeah, this is in Washington Post. Democracy dies in darkness. So let me see if they do a disclosure on this thing. So. Yeah, they do. Amazon's founder and former chief of chief owns the Washington Post. They have to put that at the end. Yeah, Amazon's under a lot of heat, right? But you know, why the walk out? Because you're not gonna go, you gotta go to work three days a week. You know, Jassy's on the record. I interviewed him during the pandemic. I asked him a direct question. Do you prefer virtual or in-person or hybrid? And he said, for certain jobs, being in the office, riffing on the whiteboard, that's where some of the best ideas came from and that you just can't do that remotely. I agree with him on that. I think that's true. For certain jobs, being in a group environment, dynamic, there's all kinds of verbal, non-verbal action, brainstorming, iterations that go on when you're on a whiteboard or having dinner or you're ideating over on a project. If you're a builder, that's what you do. That's where inspiration comes from in teamwork. So I'm for that. I think some jobs can be done remotely. But some of the most progressive companies on this topic, I could think of two up hand, Dell and Snowflake. Snowflake actually moved its headquarters to Bozeman, Montana, and Slooman and Scarpelli live out there. How are they doing right now? They're doing great. Okay. And then Dell basically said- That's not what I said on LinkedIn yesterday until I changed it. What are you talking about? I had to change my link with them in the same time. Yeah, so anyway, and so Dell, I always have been very open to people working remotely. That said, both companies are basically saying, hey, we have a back to work framework. I think Dell's was something like, if you're within an hour of an office, you got to be in a couple of times a week. I always call it Taco Tuesday, right? The company provides some incentives for people to come in. I don't think that's unreasonable two to three days a week to ask people to come into the office if they're within some kind of reasonable radius. You know, I get it. I mean, what's the traffic like in California these days? It's getting worse. People have gone back to the office. I was out there and Boston's terrible. Boston's terrible too. Well, that's because the streets and our parking- So I get- They put the bike lanes in at Boston. But I get people not wanting to commute an hour and a half each way or even an hour each way starts to get tough. I mean, so I get that. And it's frustrating and you waste a lot of time and it's, you know, you lose some productivity. But I don't think it's unreasonable to say, hey, a couple of times a week, you got to be in the office. Or, you know, once a month, you got to be in the office three or four days a week. You know, and let's pick those times. I think it's a reasonable request slash demand. So. Well, I kind of agree with- What's our policy? We've been kind of loosey-goosey. I'm going to ask the Chief PT to put a policy together. Well, I mean, generally speaking, we've had, we've tried to put some, we've tried to entice people to come in. Well, we have a work at home policy. We don't have a formal policy. No, but people come into the office and they're company for the most part. Unless they work, you know- Well, they're there mostly every day, five days a week. But, you know, I think Monday and Friday and then three days a week is beautiful. I think that's a great benefit. And if some people can do the jobs remotely, I mean, if you're a blogger, you're a written reporter, you're on the plane mostly, you could be at home, no big deal. There's no, the creative jobs are really the key and the teamwork with his camaraderie is involved. I think it's important. And if you're going to do virtual, like HashiCorp or, things like Snowflake, they have to provide a mechanism for group face-to-face. I think if you're going to do virtual, only virtual first companies, which there's proven ways to do it, you got to have a face-to-face company meeting, you know, get together. And then at some point it becomes so big, maybe you can do it differently. But that's my rant. This whole climate change, you know, justice warrior kind of thing. It just seems- It sounds like a ruse. It sounds fake to me. Doesn't sound real. Doesn't sound legit. And so that doesn't pass the stink test at all. Well, Dave, we got a lot of Cube coming up. We got a massive calendar coming up in this month, June. We got a couple of weeks off in July. We got SuperCloud coming up June 7th, 18th. And let me tell you, SuperCloud is getting a lot of traction. I think the, and now that the RSA kind of blooms off the rose, we got action happening. We're going to be at Cisco Live in Vegas next week. There's AWS London Summit. We're going to cover some exclusive news from AWS. It's going to hit on the 7th. Hence my Matt Garmin interview today. Next week, reinforced is Anaheim, which is Amazon security event. That's one or two days I might pop in for only one. I got some other things going on. And then you got HPE Discover, MongoDB in New York, and just Google Nexus coming up, VMware Explorer, SuperCloud. Oh my God, Dave, it's going to be crazy. I know, and then, right. And then, yeah, Google Cloud Next is going to be great. I mean, we got back at Google Cloud, excited about that. Yeah, and anyone watching out there right now, we're covering startups in the AI area, very hot cloud, next-gen cloud edge, open source. These are our primary areas. We've got some great stories or you're launching. Let me know. We'll get a video feature on you. We'd love to get that AI news. There's tons of startups emerging. You have any questions, DM us. Go to siliconangle.com. That's where all the traffic is. That's where all of our content is. You want to find out about what's going on, go there. cube.net is where we can find out where the cube is, what events are happening, past events, future events, what's playing. That'll all end up on siliconangle.com. That's the main site. I'm John Furrier with Dave Vellante, episode 14. Thanks for watching. Of course, give us feedback. We're going to have some special guests come in. We hit 20 episodes. We're just going to get our groove. Let us know how we're doing. We'll see you next time.