 Okay, welcome back everyone. Keeps live coverage at Red Hat Summit, Boston, Massachusetts. We're here, day one, coming to an end. We're doing keynote reviews as well as insights from the day. I'm John Furrier, I host theCUBE. My co-host is Paul Gillan. Great to see you. Rob Streche, analyst, breaking down all the action. Rob, great to see you. Guys, thanks for being here in day one packed. We had the CEO came on twice. IBM Consulting came on. A lot of great guests and customers. Full day tomorrow. Thanks for coming on. Pleasure to be here, John. Great to see you. Great coverage. Great editor. I love your articles, Paul, on SiliconANGLE.com. I retweeted a bunch of them. I noticed that, thank you. Very good job. I'm still not growing my followers. We'll help you on that one. You're a legend. Rob, great analysis so far. Let's go start with the keynotes, Rob. You attended the keynotes in this entirety. You had to jump out early to get the first interview off. What's your take on the keynote analysis from Red Hat today? I think it's a big thing and a big theme. How to make it simpler for the end users. How to get people into cloud, into Kubernetes, into OpenShift, in particular, easier. And how they went through developer hub. So how do you bring people into Ansible? So it had a lot to do with automation. I think some of the guests even today, and Paul and I were talking about hyper-automation and number of things. And I think they're playing really well into that. Also, the Service Interconnect, networking's hard in cloud. Service Interconnect makes it easier at a layer seven view. And I think that Lightspeed and Event Driven Ansible, and being able to do that, really is to kind of unburden people as well. Yeah, I think if you look at the announcements of the show, you're right on the mark, Rob. It's all about simplification. I mean, Lightspeed is about that. Event-driven Ansible, the developer hub, this is all about making life easier for people, whether they be end users or developers. And of course, Red Hat has a developer focus. And a lot of the news that came out of here today was music to the ears of those who build software. You know, Paul, that's a great point. I mean, think about Ansible, small little company that they bought. They're whole culture's automation. They breathe automation. And then it's a small little niche, that whole configuration management was small. But now that's grown, they're shutting down and folding in Ansible Fest. I thought that was a real nuanced signal that that's coming into the fold. These are their own event, now it's here. That's a big, I think they were dominating most of the thematic content on the keynote. Quite different from minimizing Ansible. They made Ansible the star of the show today. And I think they're realizing that this is a, they have a prime opportunity because the IT landscape is getting so complex. And we have so many tools to manage it, but so few tools to actually automate it. They see Ansible as a strategic way to bring some, to simplify the lives of people who have to manage these extraordinarily complex networks. Rob and Paul, I'd love to get your thoughts on the guardrails coming. I was on the panel and, you know, I ranted about it a day because, you know, we're going to put down guardrails and regulate, but guardrails can be good. You can bounce around the guardrails. I mean, when I play bowling now, I'll put the, I put the gut, I don't want to throw a gutter ball. I put the rails up and the ball bounces around. I get a strike once in a while, but that's a good metaphor to think about how to accelerate AI with a partner. So this idea of open collaboration is interesting. Now, of course, open source and Red Hat go hand in hand, but the AI hype is actually legit in their minds. And the question is, what is a guardrail? How do customers navigate it? Because in these new markets, simplicity, reducing the time it takes to do something, and the steps it takes, that's a killer business model in these new markets. Yeah, and I think even, it ties even back to the whole wanting to own multi-cloud, hybrid cloud, every cloud, be everywhere, and open source winning and being there. I think that why it ties back with their OpenShift AI is the fact that they love to talk about the fact that there was no YAML used in the demo for that when they showed that tying back to the blockchain. Did they call that out? They did. And then the next exact demo actually had YAML in it, but it was showing how you could actually use prompt engineering to then, with their writing of it using AI. So I think they're also looking for it to be hybrid because nobody knows where is AI going to really live and all that data. Your point about guardrails, John, CIOs have been through this process before. They've seen the hype cycle, the client server, you know, the web 1.0 and the stuff that people get giddy about and then they are the ones who actually have to make it work and they know how hard that is. And guardrails help in that case. What we're at right now is there is just a frenzy around AI and chat GBT and whatever may come out of that. And I think that what I hear CIOs saying is they're asking tough questions. Can we rely on this stuff? How do we govern the data? How do we make sure it doesn't go off the rails? They want guardrails, they want guidelines and they want applications that are actually going to have business value to the organization. Otherwise, their risk is spinning off into science experiments. Paul, that's great insight. And the folks watching, Paul's seen these ways of innovation and I think that's, and he's covering the stories now. I want to ask you a question. I think you could really kind of unpack this in a way that's uniquely skilled to your experience. Yeah, we've both been around the block. I'm an old guy. I'm an old guy. When I hear words in these interviews like, okay, I'm going to build a horizontal layer between the Azure stack and Amazon. Reminds me of like the old gateways. Remember when IBM and DEC had proprietary protocol stacks? You'd have to kind of create a gateway to make apps run across them. So when I hear multi-cloud, I kind of get like a gut wrenching feeling like, is it going to be slow? Is that really viable? And to your point about CIO is being sold something. I want to buy the dream. Implementation is different than actually getting there. What's your vision of the experience and looking at this market, knowing what's happened in the past. What could we learn from to call out what's right and what's wrong, what's BS, what's legit? Well, I think it's telling. We recently ran a couple of articles on what Walmart is doing with its super cloud where it's tying these multiple public clouds together. And it went and built its own abstraction layer to do that. There were lots of vendors who said that they could deliver that functionality, but Walmart went and built it itself. And I think that says a lot about the state of the market. Again, there's a lot of hype, there's a lot of promises. When you get down to making this stuff work, it always falls back to a lot of homegrown code. And by the way, they have the problem. They're living the problem. The vendor selling the software probably doesn't live the problem. Rob, now conversely, Dave Vellante's looking at Uber. See, Walmart's a great example. We've written about that extensively. Dave's doing work with George Gilbert on Uber and looking at how the data met, how they handled the data layer. So a similar kind of dynamic. They've got a lot of stuff on open source, but now people are trying to figure out what the architecture is for data across clouds. And so this is really an architectural conversation. What's your take? It's a good governance thing and a compliance thing as well. And I think, and it was talked about, and I think the multi-cloud makes sense, especially for sovereignty. And you're seeing this, especially in Europe. I mean, we've talked about it before, but Google Analytics has been outlawed in five countries in the EU. So when you start to look at where data is, who's the right to be forgotten? GDPR, CCPA in California. And you start to look at all these data meshes and how things come together. It is data has weight and gravity, but it also has regulation. It brings with it. And I think AI is just going to make that such an almost untenable problem for people that they have to figure out, which is why people are looking into it and saying, okay, what data goes where? How do I do summarization? How do I do some inference at the edge? And I think edge is going to become that much more important. The question is, how and what should regulate? Michael Dell was on theCUBE today. I was watching Dave's interview while I was taking a break when you guys were carrying in some interviews. I, Michael Dell said, he thinks there'll be regulation around AI. Should AI be regulated? And when- Absolutely, it has to be. I mean, the potential disasters, we haven't begun to see the real AI disasters that are lurking in our future. You know, when you look at, again, chat GPT being the example everyone is using, the examples of that utility going off the rails and doing crazy stuff are- Pro-Legend, our legend at this point. And that's a fairly safe environment because it's not operating equipment in a hospital. It's not managing traffic lights. People's lives are not being affected by that yet. But when you look at the potential, AI is such a black box that we're going to need rules about the data that we put in there, about transparency, about how decisions are made. They're going to have to be audit trails. You bring up a good point, actually, on the whole hallucination thing and getting off the rails. Not only is that a problem, by the way, I have a different opinion on the regulation. I think it's how it's regulated. I think it's a standards issue, but we'll debate that. But I want to just say, I've noticed that people right now are actually having fun driving manipulation to get the hallucinations. Okay, that to me is a completely tell sign. That means it's like war games. So it's like, I'm going to go, I'm going to just hack that thing. I'm going to make that car crash. Oh, look what I just did. I just killed five people. But thank goodness they're doing it, right? Because that's how you find the bugs. So it is a black box, but the question is, who should regulate it? Should it be standards by? So again, it's not an IEEE standard. There's no, who does it? But I think what will be interesting. I think there's a lot of different bodies that are pushing open source with AI, right? And we talked to a few, just a couple of weeks ago at the open source summit. And I think when you start to look at where people are trying to put those guardrails up around AI, I don't know if it comes from regulation, from standards body, or from the fact that it becomes so pervasive and transparent in an open source way. I don't know what the right answer is, but I think it's going to go to open source route. You make a good point. I think, to John's point, the regulation is actually going to come from the companies that use it. They are going to establish their own ethical guidelines and we will see, they'll publish those guidelines, they'll be broadly adopted. I think we'll see good companies realize that they have their reputational disasters, their potential reputational disasters if their AI models go off the rails. And so they have every interest in doing this right. Yeah, it's almost like a hack almost. It could be company killing. And we were riffing Paul on theCUBE in Vancouver and a little bit of KubeCon in comparing the Twitter moment where the plane landed on the Hudson. And I think AI needs to have that moment where something happens, people go, wow. And I think what's scary about it is it's not like the plane landing on the Hudson, which actually was a safe. They actually saved people thanks to the pilot, but this could be a disaster. It could be a disaster with critical infrastructure. It could be some takeover, some cause some real harm. I think part of me is waiting for that moment and then everyone's going to wake up and go, okay, this is out of control. Because it is hot right now. People are pumping out code, they're testing stuff. I just read an article about test data. People are putting bad test data into the large language models. This is causing contamination. They used the word contamination, which we were saying cloud pollution. We were saying code pollution at KubeCon, which was like, what are you talking about? Contamination, it's like bad stuff can get in and that's only going to hurt. Yeah, and I think it's only going to get worse because I think people don't know if they put something in there and they're trying to use AI for customization of things that they want to sell you, like all the spooky ad stuff. When you start to look at that, well, in what country, where was it collected from? And when you're training that model, what data are you using to actually train it? And I think it becomes a, hey, well, if I make fake data and I put that fake data into the model, what happens when the real data goes into the model at that point? Well, I think what we'll see is we're going to see a lot of vertical use cases. An example with Lightspeed announced at this conference is an example of that where they've taken generative AI, they populated the model, trained the model on data that relates directly to systems management. And they can control that data set. They can control the output. They make the outputs explicit as YAML playbooks. And I think that's the way you go about it. And I think to your exact point of what you said about the company self-regulating, also at the bottom of the demo, what was really neat is they showed all of the licenses and where the data came from that came up with that YAML file. And I think it said, hey, this is actually using a GPL-3 license. Okay, well, GPL may or may not be good for me if it's not Apache versus MIT and all these licenses and having to know them all again is really going to be complicated for people. And I think that's, I think it's going to get way more complicated before it gets any easier. So just reading some of the news headlines today, Spotify is developing AI tools trained on its host voices to create targeted ads according to Bill Simmons, founder of Spotify's podcast, Network the Ringer. On SiliconANGLE, besides your awesome article, Paul, here's the headlines on SiliconANGLE. Days breaking analysis, the AI-powered hybrid multi-super cloud. Love that one. Michael Dell talks about- How many prefixes can we add to cloud? Michael Dell talks about the power of AI and Dell's role in it. Microsoft has their build conference going on this week in Seattle. Microsoft expands AI plugins and co-pilot ecosystem for developers. Google debuts new generative AI advertising tool. This is just what's on the front page, SiliconANGLE. Yeah, and Adobe is just a generative fill for Photoshop to add more fake information. Apple inks multi-billion dollars. People broke out for wireless components and then videos and news and this is more in there. So again, dominating the headlines is this AI. And it's going to get dangerous. I think misinformation, pollution, contamination. You're going to start to see some real fast side effects of even in the media business, Paul. I mean, AI is going to, you got voices for hosts. How's it joking the other day? Maybe I'm a panel. Someone says, John, ask a good question. I mean, do we even, like hologram. So this is interesting time right now. The good, bad, and the ugly is going to rear its head. Yet there'll be unicorns born here, right? There's going to be startups going to come out and saying, you know what? I'm going to solve these problems. So to me, that's kind of my optimistic side of me that in all these waves, there's a new brand. There's a young Steve Jobs out there, young Bill Gates dropping out of college and saying, you know what? I'm going to solve this problem. And it actually takes territory off an incumbent and it seems too slow. This is kind of an internet moment, you know? And I remember back to 1994, the first time I saw the World Wide Web and people's minds were blown. They just could see that this was going to be a game changer. And I think the same thing is happening with AI right now. Well, and you think about the web, remember the reaction? Because I was at Eulah Packard at the time and a guy that I became friends with but invented HTML and his partner Internet HTTP. That was Tim Berners-Lee and Dave Ragget. They were like guys coming out of CERN, they moved to MIT for the W3C. And at that time, even Eulah Packard had a scientist on loan and they still missed the web. They poo-pooed it. And other people were like, it's a toy, it's not really real, it's too slow. AOL, so there was massive criticism. It wasn't fear, it was more like it's not real. I mean, Ken also even said, well, DC was a toy. So in every shift, there's always going to be some detraction. Here, I think it's just fear. I think it's just general fear. And no one's saying it's not going to happen. They're more like, it's going to happen, we don't know what to do. Well, they knew the web was going to happen but they were- No, I was afraid of it. They were taking, it was a proprietary product. So you had Copy, Serve, and Prodigy, and Washington Post and all these guys who were building these closed walled gardens. What made the web work was that it was open. There you go. And now with open source, how would you describe the modern times now? I mean, it's a little bit different, different equation, you get open source, you get cloud scale. Yeah, I think open generally wins because it's transparent. And I think that with this and with open source AI coming, I think there will be some transparency. Like we said, we were talking about a couple weeks back with the meta leak and some of the stuff that Google's been saying. The Google Memo. Yeah, the Google Memo and meta leaking their algorithm, their models and stuff like that. Yeah, I mean, when they use the word moats, moat to me means proprietary, means sustainable competitive advantage, which means some sort of lock-in. And so like, Matt Hex, he's an open source guy, Red Hat's all based on open source. So if open source becomes its own ecosystem industry company, it's going to have walled gardens or like ecosystems. So the question is, I mean, that's possible, I'm not saying it is, but if that happens, the groups that could create islands. So I think the connective tissue between these ecosystems and open source is going to be, to me, the tell sign. I mean, I don't know what else to say, but I get scared when I hear moat. Moat means sustainable competitive advantage, which used to be proprietary lock-in, Paul. I mean, I just think that conversation is over. I mean, you made the point today, Rob. I mean, open source has won. That doesn't mean there won't be proprietary products and the proprietary will go out of existence, but the natural pull of the market now is toward open. And that seems to me to be an unstoppable force. Yeah, I mean, you can even see it with the announcements today, with, again, developer hub and looking at backstage as being that opening to take away the control plane out of the hyperscalers. I mean, that's why Spotify originally went that route and it's why Walmart is doing what they're doing and it's a huge way to untangle from those control planes. I think of these new dynamics that you were mentioning about Walmart and then we talked about Uber, we have end users who are actually part of, they're not vendors, they're contributing. So I think you're going to see cohorts emerge where the big companies who have the problem solve it that might change the game, but also Red Hat has changed. They're not just a software company in open source, they're developing an ecosystem. You guys had Stephanie on, they got partners now, they have a global network and they're going for the global distributed computing play which is they got a plug in other companies. I think Red Hat is having a huge impact on the industry writ large, but the cloud in general because they are out there, they are a force for openness and they are driving this with the cloud providers and they're driving, they won't say as much, but I think the fact that the cloud providers are warming to the idea of embracing multi-cloud standards and I think Red Hat has a lot to do with that. I thought that last panel, John Grange, the Senior Vice President was on with Matt Hicks and Reddy Goodlove from Anthem, the big customer speaks volumes of the impact to IBM as well and you look at what they're talking about here, talking about IBM and Red Hat, it's always going to be that spear for IBM and IBM and John Grange is like, look, we are on Amazon, we got an Amazon practice, we got an Azure practice, you can almost see the formation going where they're going to try to be the middleware for multi-cloud. Or multi-cloud. It even goes beyond the big three or big four hyperscalers because they also announced some stuff with Oracle recently as well. When you start to look at the fact that they have OVH and others and other clouds and BT and if you start to look around, there's still another group of cloud providers. Good point. That is 500, 600 strong below them. I asked A.J. Patel before he left VM where he was the Senior Vice President of the software group and he now left the company, went somewhere else and I said, hey, you know, the super cloud thing, across cloud, they're calling across cloud. And he's like, look at John, I'm a middleware guy. What do you think about the hyper clouds? Because Amazon is not promoting multi-cloud, because why would they promote the competition? He goes, they're just the hardware, they being the clouds. He called them hard. That's one interesting. So of course, why would Microsoft and AWS, why would they want to promote multi-cloud? That's like the PC promote in the Mac. Of course not, yeah. Of course not. All right guys, tomorrow, big day. We're going to break down all the action. Today, the big surprise was a lot of red hat and Ansible together. Tomorrow's going to be about scale and as Paul quoted, this is about learning. Doing more with less was the theme. They showed a lot of automation. They played some AI cards, was really strong. Tomorrow's going to be about scale and efficiency and edge. We'll be back tomorrow. Thanks for watching.