 Okay, welcome back everyone. Cube's coverage of Remarch here in Las Vegas in person. I'm John Furrier, host of theCUBE. This is the analyst panel wrap-up analysis of the keynote, the show, past one and a half days. We've got two great guests here. We've got Andy Thry, Vice President, Principal Consultant, Constellation Research, Larry Carvello, Principal Consultant, Robust Cloud, LLC. Congratulations on going out on your own. Thank you. Andy, great to see you. Great to see you as well. Guys, thanks for coming on. This is the session where we kind of break down and analyze. You guys are analysts, industry analysts. You go to all the shows, we see each other. You guys are analyzed in the landscape. What does this show mean to you guys? Because this is not obvious to the normal tech follower. The insiders kind of see the confluence of robotics, space, automation, machine learning, obviously it's IoT, it's industry, it's a bunch of things. But there's some dots to connect. Let's start with you, Larry. What do you see here happening at this show? So you got to see how Amazon started when AWS started. When AWS started, it primarily took the compute, storage, networking of amazon.com and put it as a cloud service, as a service, and started selling the heck out of it. This is a stage later now that amazon.com has done a lot of physical activity, and using AIML and the robotics, et cetera. It's now the second phase of innovation, which is beyond digital transformation of back office processes, to the transformation of physical processes, that people are now actually delivering remotely. And it's an amazing area for us. So back office is IT data center kind of vibe. You're saying front end industrial life. Life as we know it. I mean, I just stopped at a booth here, and they have something that helps anybody who's stuck in the house, who cannot move around, but with Alexa, order some water to bring them wherever they are in the house, where they're stuck in their bed. But look at the innovation that's going on there, right at the edge. So I think those are- And you got the Luna, got the sex appeal of the space. Luna Rappos interviewed those guys. They got a rover on Mars. They're going to be colonizing the moon. I said, I made a joke. I'm like, well, I left a part back on earth. I'll be right back. Can't drive back to the office. So a lot of challenges. Andy, what's your take on the show? Take us to your analysis. What's the vibe? What's your analysis so far? It's a great show. So as Larry was saying, one of the things was that when Amazon started, right? So they were more about cloud computing. So which means is they try to commoditize more of data center components or compute components. So that was working really well for what I call it as a computer economy, right? And I call the newer economy as more of a AI ML based data economy. So when you move from a computer economy into a data economy, there are things that come into the forefront that never existed before, never popular before. Things like your AI ML model creation, model training, model movement, model inferencing, all of that, right? And then of course the robotics has come a long way since then. And then some of their, what they do at the store or the charging, the whole line. The whole concept of all of this components, when you put them on green men, such a big show, it was getting lost. So that's why, I mean this, they didn't have it for a couple of years. They had it one year. And now all of a sudden they woke up and say, you know what? We got to do this. To bring out this critical components that we have that's right mature for the world to next component. So that's why I think they are pretty good stuff. And some of the robotics things I saw in there, like one of them I posted on my Twitter, it's about the robot dog sniffing out the robot roer, which I thought was pretty hilarious. Yeah, I mean this is the thing, I mean, seeing, you seeing like the pandemic, but everything on hold on the last remars. And then the whole world was upside down. But a lot of stuff pulled forward. You saw the call center stuff booming. You saw the zoomification of our workplace. And I think a lot of people got to the, to the realization that this hybrid, steady states here. And so, okay, that settles that. but the digital transformation of actually physical work location. The walk in and out store right over here was seen, that's the go store in Seattle, you've all been there. In fact, I was kind of challenged, try to steal something, I'm like, okay, I'm pulling all my best New Jersey moves on everyone, you know. You'll get charged for it. I couldn't get away with it. Two of them, double packs, they're dropping. It's smart as hell, can't beat the system. But you bring that to where the AI machine learning and the robots meet, robots. I mean, we had robots here on theCUBE. So I think this robotics piece is a huge IOT because we've been covering industrial IOT for how many years, guys? And you can know what's going on there. Huge cyber threats, huge challenges, old antiquated OT technology. So I see a confluence and a collision between that. OT getting decimated to your point. And so, I mean, do you guys see that? I mean, am I just kind of seeing Mirage? I don't see it, it'll get decimated. It'll get replaced with a newer. Perfectly, David would call me out on that. I mean, Microsoft's going to get killed. I think it's going to have to be reworked. And just right now, you want to do anything in a shop floor. You have to have a physical wire connected to it. Now you think about 5G coming in. And without a wire, you get minute details. You get low latency, high bandwidth. And the possibilities are endless at the edge. And I think with AWS, they got outposts, they got snow cones. There's a threat to them at the edge. Outposts is not doing well. You talking to anyone out there, it's like, you can't find success stories. Now I'm going to get hammered by Amazon people. What are you saying that? You know, EKS, for example, with serverless is kicking ass too. So I mean, I'm not saying outposts was wrong answer. I mean, it was a right at the time. What, four years ago that came out? Yeah. Okay, so, but that doesn't mean it's just theirs. You got Dell Technologies wants some edge action. So there's HPE. Yes. So you got a competitive edge situation. I agree with that. And I think that's definitely not Amazon's strong point. But like everything, they try to make it easy to use. You know, you look at the AI ML and they got Canvas. You know, so Canvas says, hey, anybody can do AI ML. If they can do that for the physical robotic processes or even like with outposts and snow cone, that'll be good. I don't think they're there yet and they don't have the presence in the market. Like HPE and Dell. Let me ask you guys this question because I think this brings up the next point. Will the best technology win? Or will the best solution win? Because if clouds are platforming, all software is open source, which you can make those assumptions. You then say, hey, they got this killer robotics thing going on with Artemis, a moonshot. They're trying to colonize the moon, but oh, they discovered a killer way to like solve a big problem. Does something fall out of this, this kind of remarge environment that cracks the code and radically changes and disrupts the IoT game? That's my open question. I don't know the answer. I'd love to get your take on what might be possible. What wild cards out there around disrupting the edge? So one thing I see the way, so when the IoT came into the world of play, it's when you're digitizing the physical world, it's the IoT that does digitization part of that actually, right? But then it has its own set of problems. You're talking about installing sensor everywhere, right? And not only installing your own sensor, but also you're installing competitive sensors. So in a given square feet, how many sensors can you accommodate? So there are physical limitations on liabilities of bandwidth and networking all of that. And integration. As well, right? So when that became an issue, this is where I was talking to the robotic guys here, a couple of companies, and one of the use cases they were talking about, which I thought was pretty cool is, rather than going the sensor route, you go the robot route. So if you have either a factor that you want to map out, you put as many sensors in your robot, whatever that is, and then you make it go around, map the whole thing, and then you also do a surveillance on the whole nine yet. So you can either have a fixed sensors, or you can have a moving sensors, so you can have three or four robots. So initially, when I was asking them about the price of it, when they were saying about $100,000, I was like, who would buy that? When they explained that, this is the use case. Oh, that makes sense, because if you already install entire factory floor sensors, you're talking about millions of dollars. But if you do the movable sensors in this way, it's a lot cheaper. So it's, based on your use case, what are your use cases? What are you trying to achieve? The general purpose is over, which you're getting at. And that's the enablement, this is again, this is the cloud scale open question, is okay, the differentiations aren't going to, isn't going to be open source software. That's open. It's going to be how you configure it, what workflows you might have, the data streams, or. I think, John, you're bringing up a very good point about general purpose versus special purpose. Yesterday, Zooks was on the stage, and when they talked about their vehicle, it's made just for self-driving. You walk around in Vegas over here, you see a bunch of old fashioned cars, whether they're Ford or GM, and they put all these devices around it, but you're still driving the same car. So you can retrofit those, but I don't think that kind of IoT is going to work. But if you redo the whole thing, you're going to see a significant change in how IoT delivers value all the way from the industrial to home, to healthcare, mining, agriculture, it's going to have to redo. I'll go back to the OT question. There are some OT guys, I know Rockwell and Siemens. Some of them are innovating faster, the ones who innovate faster to keep up with the IT side and well as the MLA model are going to be the winners on that. Yeah, I throw your great, Andy, your thoughts on manufacturing, you brought up the sensor thing. Robotics ultimately is, end of the day, an opportunity there, obviously machine learning, we know what that does. As we move into these more autonomous builds, what does that look like? And is Amazon positioned well there? I'll say they have big manufacturers, some are saying that they might want to get out of that business too, that Jassy's evaluating that, some are saying. So where does this all lead for the robotics, manufacturing, lifestyle, walk in, grab my food? Cause it's all robotics and AI at the end of the day. I got sensors, I got cameras, I got non-humans moving heavy lifting stuff. Brake fixing the moon will be done by robots, not humans. So it's all coming. What's your analysis? Well, so the point about robotics is on how far it has come, it is unbelievable, right? Couple of examples, one was that I was just talking to somebody and I was explaining to them to see that robot dog over there, the Boston Dynamics one, climbing up and down the stairs, that's more like a, you know, the dinosaur movie opening the doors scene. It's like that for me because the coordinated things it is able to go walk up and down, that's unbelievable. But okay, it does that. And then there was also another video which is going on viral on the internet. This guy kicks the dog, robot dog, and then it falls down and gets back up and the sentiment that people who are feeling for the dog, you can't, it's a robot, but people, it just comes to that level. Empathy, for a non-human. But you're saying, hey, you, get off my lawn. You know, it's like, where are we? It has come to that level that people are able to kind of not look at that as a robot, but it's more like a functioning, almost like a pet level, human level being. And you saw that the human-like walking robot there as well. But to an extent, in my view, they're all still in an experimentation, innovation phase. It doesn't matter in the industrial terms yet. Yeah, not yet. It's coming fast. That's what I'm trying to figure out is where you guys see Amazon and the industry relative to what from the fantasy coming reality of space and Mars, which is, it's intoxicating, let's face it, people love this. The nerds are all here, the geeks are all here. It's a celebration on James Hamilton's here, coming up on theCUBE, and he's here as a civilian. Jeff Barr, same thing, I'm here not for Amazon. I bought a ticket, no, you didn't buy a ticket. Come on, man, check on that, but he's geeking out. They're there because they want to be here, not because they have to work here. Well, I mean, the thing is, the innovation velocity has increased because in the past, remember, the smaller companies couldn't innovate because they don't have the platform. Now, computers have a platform available at the scale you want. AI is available at the scale. Every one of them is available at the scale you want. So if you have an idea, it's easy to innovate. The innovation velocity is high, but where I see most of the companies failing, whether a startup or a big company, is that you don't find the appropriate use case to solve and then don't sell it to the right people to buy that. So if you don't find the right use case or don't sell the right value proposition to the actual buyer, then why are you here? What are you doing? I mean, you're not just an invention, like a telephone kind of thing, or, you know. So you- Okay, now let's get into the next talk track. I want to get your thoughts on the experience here at RIMARS. Obviously, AWS and the Amazon people kind of combined effort between their teams, the event team does a great job. I thought the event personally was first class. Coffee didn't come in late today. It's a complaint, a little complaint in and out there. You're going to keep interviews. But world-class, high bar on the quality of the event. But you guys were involved in the analyst program. You've been through some of the briefings. I couldn't do that because I'm doing the Cube interviews. What would you guys learn? What was some of the key walk-aways impressions? Amazon's putting a new teams together, it seems, on the analyst relations. They got their mojo booming. You know, they got three shows now, RIMARS, Reinforce, Reinvent, which will be at the Cube at all three. Now we've got that coverage going. What's it like? What was the experience like? Did you feel it was good? Did they really need to improve? How would you grade the Amazon team? I think they did a great job over here in just bringing all the physical elements of the show, even on the stage where they had robots in there. It made it real and it's not just fake stuff. And every, you know, or most of the booths out here are actually having high quality demos. Not exactly, not paperware. I won't say the name of the company. And even the sessions, you know, were very good. They went through details. One thing that stood out, which is good, and I cover low-code, no-code, and low-code, no-code goes across everything. You know, you've got DevOps, no-code, no-code, you've got AI, low-code, no-code, you've got application development, low-code, no-code. What they have done with AI, with low-code, no-code is very powerful with Canvas. And I think that has really grown the adoption of AI, because you don't have to go and train people what to do. And then, you know, people are just saying, hey, let me kick the tires, let me use it, let me try it, and go. It's going to be very interesting to see how Amazon, on that point, handles this AWS, handles this data tsunami. It's because of the snowflakes. Snowflakes, especially running the table on the old Hadoop world. I think Dave had a great analysis with some other colleagues last week at Snowflake Summit. But still, scratch on the surface. The question is, how shared that ecosystem, how will that morph? Because right now, you've got data bricks, you've got Snowflake and a handful of others, Teradata's got some new chops going on there, and a bunch of other folks, you know, some are going to win and lose in this downturn, but still, the scale that's needed is massive. So you've got data growing so much, you know, you were talking earlier about the growth of data, and you were talking about the growth. That is a big pie, and the pie can be shared by a lot of folks. I don't think- Snowflake pays AWS, remember that. I get it. I get it, but they got very unique capabilities, just like Netflix has very unique capabilities. They also pay AWS, right? But they're competing on Prime. So I really think the co-opetition is going to be there. The pie is so big that there's not going to be losers, but everybody could be winners. I'm going to obviously follow up with you guys after the next time we have an event together, we'll get you back on and figure out, how do you measure this transitions? You went to IDC, so they had all kinds of ways to measure shipments. Even Gartner had fumbled for years, the Magic Quadrant on IaaS and Paz when they had the Marcus Sharers book, and then they finally bundled Paz and asked together after years of my suggesting, thank you very much, Gartner. But that's just, as the landscape changes, so does the scoreboard. So how do you measure who's winning and who's losing? How can we be critical of Amazon so they can get better? I mean, Andy Jassy always has said to me in Adam Slessy's same way, we want to hear how bad we're doing so we can get better. So they're open minded to feedback. I mean, not shit-posting on them, but they're open to critical feedback. What do you guys, what feedback would you give Amazon? Are they winning? I see them, number one, clearly over Azure, like by miles, and even though Azure's ticking ass and taking names and getting back in the game, Microsoft's still behind by a long ways, in some areas. So the scoreboard's changing, what's your thoughts on that? So, look, I mean, at the end of the day, when it comes to compute, right? Amazon is a clear winner. I mean, there are others who are catching up to it, but still they are the established leader. And it comes with its own advantages because when you're trying to do innovation, when you're trying to do anything else, whether it's a data collection, we were talking about the data sensors, the amount of data they are collecting, whether it's the store, the self-serving store, or other innovation projects, what they have going on, the storage compute and process off that requires a ton of compute. And they have that advantage with them. And as I mentioned in my last article, one of my articles, when it comes to AI and ML and data programs, there is a rich and there's a poor. The rich always gets richer because they have one leg up already. You know, I mean, the amount of model training they have done, the billion, or trillion dollar trillion parameterization, fine tuning of the model training and everything, they could do it faster, which means they have a leg up to begin with. So unless you are given an opportunity as a smaller mid-size company to complete at the same level, you're going to start at the negative level to begin with, you have a lot of catch up to do. So I mean, and the other thing about Amazon is that when it comes to a lot of areas, they admit that they have to improve in certain areas and they're open and willing to listen to the people. Yeah, let's get critical. Let's do some critical analysis. Where does Amazon web stores need to get better in your opinion? What criticism would you in constructive way share? I think on the open source side, they need to be more proactive. They are already, but they got to get even better than what they are. They got to engage with the community. They got to be able to talk on the open source side. Hey, what are we doing? Maybe on the hardware side, can they do some open sourcing of that? They got to grab it on, they got a lot of stuff. Will they be able to share the wealth with other folks other than just being on an Amazon side, on the edge with their partners? Got it. If they can now take that, like you said, compute with what they have with a very end-to-end solution, the full stack, and if they can extend it, that's going to be really beneficial for us. Andy, final word here. One area where I think they could improve, which would be a game changer would be, right now, if you look at all of the solutions, if you look at the way they suggest implementation, the innovations, everything that comes out, comes out across very techy oriented. The personas very techy oriented. Very rarely their solutions appeal to the business audience or to the decision makers. So if I'm, say, an analyst, if I want to build a business analyst rather, if I want to build a model and then I want to deploy that, or do some sort of application, mobile application, what have you, it's a little bit hard. It's more techy oriented. So if they could appeal or build a higher level abstraction of how to build and deploy applications for business users, or even build something industry specific, this is where a lot of the legacy companies succeeded. Go after manufacturing specific or education. Well, we coined the term super cloud last reinvent and that's what we see. Jerry Chen at Greylock calls it castles in the cloud. They can create these motes on top of the cap X of Amazon. And right there back. And the difference in what you're paying and what you're charging, if you're good, like a snowflake or a mongo, I mean, mongo's, I mean, they're just as big, it's not bigger on Amazon than snowflake is. Cause they use a lot of compute. No one turns off their database. Snowflake a little bit different, a little nuanced point, but this is the new thing. You see Goldman Sachs, you got Capital One. They're building their own kind of, I call them sub-clouds, but Dave Laudek says it's a super cloud. And that essentially is the model. And then once you have a super cloud, you say, great, I'm going to make sure it works on Azure and Google. And I'll be Bob if I have to. So we're kind of seeing a playbook, but you can't get it wrong because it scales. You can't scale the wrong answer. So that seems to be what I'm watching is who gets it right, product market fit. Then if they roll it out to the cloud, then it becomes a super cloud. And that's pure product market fit. So I think that's something that I've seen some people trying to figure out. And then are you a supplier to the super clouds? Like Adele, are you becoming an enabler? You know, what's Dell technologies do? I mean, how do the box movers compete? You know, the whole thing is now hybrid and you're going to have to see just you said you said. Hibers a steady state. I don't need a, by the way, we're box movers. We can't get the ships because Broadcom and Apple bottom wall. I mean, there's a huge chip problem going on right now. I mean, all these problems, when you abstract to a much higher level, a lot of these problems go away because you don't care about what they're using underlying. As long as you deliver my solution, it could be significantly a little bit faster than what it used to be. But the end of the day, are you solving my specific use case? Then I'm willing to wait a little bit longer if you can. Times on our side. And now the right answer is good. Larry, Andy, thanks for coming on this great analyst session turned into more of a podcast vibe. But you know what? It's chill here at Remars. Thanks for coming on and we unpacked a lot. Thanks for sharing. Oh, thanks for having us. Appreciate it. We'll get you back on. We'll get you in the rotation. We'll take it virtual. Do a panel. Do some panels around this. Oh, this is not virtual. This is physical. No, we're live right now. We'll get back to Palo Alto. You guys are influencers. Thanks for coming on. You guys are moving the market. Congratulations. Take a minute, quick minute each to plug any work you're doing for the people watching. Larry, what are you working on? Andy, you have to go to Larry. What are you working on? So since I started my company Robust Cloud, since I left IDC about a year ago. I'm focused on edge computing, cloud native technologies, and local and local. And basically I help companies put their business value together. Andy, what are you working on? I do a lot of work on the AI ML areas. Particularly, a few of my reports are in the AI ops, incident management, and ML ops areas. How do you generally improve your operations in other words? How do you use AI ML to improve your IT operations? How do you use IT ops to improve your AI ML efficiency? So those are the real, real hardcore business transformation. Yep. All right, guys, thanks so much. Come on on the analyst session. We do keynote review, breaking down remars after day two. We've got a full day tomorrow. I'm John Furrier with theCUBE. See you next time.