 Welcome to Amsterdam, and KubeCon, CloudNativeCon 2023. Join John Furrier, Savannah Peterson, Rob Streche, and UPSCOT as the Kube covers the largest conference on Kubernetes, CloudNative, and open source technologies together with developers, engineers, and IT leaders from around the globe. Live coverage of KubeCon, CloudNativeCon 2023 is made possible by the support of Red Hat, the CNCF, and its ecosystem partners. Good evening and welcome back to KubeCon Europe. We are live from Amsterdam. My name is Savannah Peterson, and I am sandwiched by brilliant men right now. I feel very good about this. I've got Rob on my left. I've got you on my right. And I've got one of our favorite Kube alumni, Basam, over here. Basam, welcome. Thank you. Thanks for having me again. I mean, you're obviously great. That's why we've had you back, what, seven times? Thank you. Yes, I appreciate it. I appreciate it. Every one of them, so. That's what we like here. Is that OK? Did you say that? What do you think of the show so far? It's great. It's really awesome to see the energy. It's the largest European open source conference, which is just awesome to see. That is cool, isn't it? Yeah. And like half the people are new. So it's like, you know, every year you come back to KubeCon and you think, wait, it's the same people, but it's not. It's actually new people, which is, you know, exciting. This community continues to grow. Well, and at such scale, I mean, you mentioned earlier, they said, what, 58% were first time attendees. It's exciting. You know, I think we're still at a real momentum buildup stage with all of this. This cloud native thing is actually picking up. You think we're going to stay, you know? Yeah. It might be real. Cloud might be real. Yeah. But I thought what was also interesting was they talked about over 54,000, you know, community members out there. So if you look at it that way and you've got, you know, 10,000 of them here today where how many of those, it would be interesting to know how much of those were overlapped. Is it people coming here first and then getting involved in the community or not? That would be interesting to understand. Yeah. I wonder how much of this is a gateway drug. Yes. Or is more, you know, or is established? Well, actually, you know, we had Red Hat on just a second ago and they were commenting on how the experts are here helping out the kind of first timers or folks getting started. And I do, I will say, manage a lot of communities. Open source community is one of the best at that. It is. It really is inclusive. They make you feel welcome. I'm a non-technical and even I feel welcome, which really says something. Which is a real accomplishment. Anyone who makes me feel welcome, that's a real accomplishment quite frankly. But Sam, what are some of the trends you've noticed? I mean, look, I think the thing that stands out to me is that I think about five years ago, this was all about Kubernetes, right? I mean, it is KubeCon. It is KubeCon. It's also CloudNativeCon. It is. Both. Don't forget both hashtags on everything. So in some ways, I think, you know, if you look back five years ago, this was all about containers and microservices and all that stuff. And now we're actually seeing a branch out, right? We're seeing a very clear branch out, right? Yeah. Now it's like Kube Kubernetes managing everything, right? Like it's literally becoming the way that everyone's doing it. That everyone's doing automation or infrastructure and everything else. And it looks like it's completely branching out, right? And not, you can see it from the vendors. You can see it from the open source projects. But you're getting into policy and cost and everything else starting to build around this ecosystem. And it's just visible in everywhere. I think that's, yeah, I heard Kubernetes native as a term today for the first time. And I think, I love that you just said that. It's so, when I, well, when most people, I think, were exposed to Kubernetes, it was all about containers. And now I feel like this whole conversation has expanded. Right. It's almost like containers that... We almost haven't even really talked about them that much today. Do your containers much? I don't know. It's almost taken a backseat. It's almost taken a backseat to the whole ecosystem maturing beyond just looking at compute, beyond just looking at scheduling. It's so much more, you know, there's vendors out here that do cost management, which wasn't a thing. Right. Or not in any meaningful sense for the longest time because... That's a great point. We had to innovate. We had to move forward. And now we're getting into this age where maturity is actually, you know, becoming a differentiator. It's not just about getting into containers. It's actually being successful in production at scale with containers and not just containers. Right. And I think that level of maturities is fun to see because it's now no longer just that one single thing. It's a whole ecosystem and it's grown way beyond Kubernetes. And I think that's one of my key takeaways for this event and for today is seeing that it's no longer just Kubernetes. I think that's a really... Yeah, it's such a good point. And it's no longer just containers. There's so many different businesses. There's all this stuff going on. Speaking of conversations that have heated up and matured, AI, chat, GPT, all the whisper, all the buzz around here. Basam, what's your hot take? Very interesting topic and a trending topic, of course, but I don't know, it's interesting. I think we're barely scratching the surface on how AI intersects this cloud native world. Totally agree with you. Like the finger nail on the ice. Yeah. And I think there's a lot of people here are trying to really quickly differentiate with integrations, chat, GPT. We're doing this. Yeah. And everything has AI acronyms on it, right? I went from IO to AI real quick there. I'm just imagining that people are like literally changing booth design last minute to add the word AI. Or put a little chat, GPT. Exactly. No, I think you're actually spot on with that. Yeah. It's not clear exactly how it's going to intersect. And so I think we have a lot to learn in terms of what's actually differentiated and what actually has value, what's actually going to work, what's not, right? There's probably a lot of sorts of legal and other implications of all this AI and chat GPT stuff. A lot. So I think early days. It's super early days. Very early days. I think it's changed. Like last year it was like all about AI ops. And like everybody, like you're saying was like AI-ifying, you know, from an operations perspective. It was more about automation. And this year, you spin around and it's not, there's a few booths that have chat GTP. You spelled out right on them. But it's like, okay, well, what about AI ops and ML ops? And I think to your point, it was the fact that it's those day one, day two things that we're focusing on. Again, it's not about the containers and the infrastructure as much as it is about, hey, we got day zero done, but now we're moving on to day one and day two, which is pretty interesting. Things are moving so quickly. I mean, one measure of that is how many talks actually have chat GPT in them. And then if you think about submissions, there's actually, I looked, very little mention of chat GTP in the entire schedule. That's how recent this is. The fastest adoption of a new technology we've ever seen, if I'm not mistaken. I bet you next coupon in Chicago It'll be inundated. Every talk is going to be on this. I think you're absolutely right. And I think there's going to be a lot of armchair experts where we're sitting here in such an early stage. I've had people talking to me about it. They're like, oh yeah, I'm a chat GPT expert. I'm like, it's been 90 days. Nobody's an expert in anything in 90 days, really. I mean, I guess, yeah. Do you think, and you touched on the legal side of this, but we're in a real tenuous stage within this. Do you think that chat GPT and the AI ML advancements we're seeing right now are ethical? Do you think they need to be governed? I mean, from my perspective, we're pushing the boundaries on what's ethical and what's not at this point. I'm assuming there's a bit of give that has to happen on the legal side and there's probably a bunch of things that have to be revisited. Is it open source if I was writing the code and I was assisted by a co-pilot of some sort that brought in proprietary code? Is it still open source? What about the reverse side of this? Great question. Is it proprietary if it actually came from code that was open or et cetera? There's a lot of things that need to be revisited or folks that are putting customer data and open LLM models. There's some pretty wild stuff going on that is a little privacy breaching. It feels like the gold rush right now so people are doing whatever is needed. Will they regret them? Will it stick? It's not clear at this point. It's so early. Yeah. No, it's definitely... Yeah, it's vague and it is dicey. In fact, you talked about the customer privacy that I've been watching people or even some of the videos of the prompts that people are putting in. I'm like, yo, you are exposing all your stuff right now on a Twitter thread that you're doing a freaking LUME video. It's like... The one that gets me is putting private customer data. This is what I mean. You can use the HAT GPT so that you can get a feature... Insights or whatever. You're literally actually giving away your own customer data, your private data, the thing that actually is a competitive advantage. Yeah. Giving and making part of... That's one of the most competitive advantages you have. It's one of the things you can sell all the way to the bitter end or all the way to the top depending on what you do. You actually said something earlier to John that I think is really interesting that proprietary data is the new IP. Tell me about that. I mean, look, if you assume that everybody is able to actually build models and train models and create... That's going to be a commodity. The data that trains the models is really the new IP. If you look at a number of companies that have private data that is not accessible on the Internet in whatever domain, whether it's financial or infrastructure, automation, et cetera, that data is actually more valuable than the result of training a model. The actual data is more interesting. Again, going back to... Why are you putting your private data in public models? Well, and knowingly... Knowingly, to get a feature so somebody can actually give you a prompt and that's your feature, the prompt, right? I love that you said that. Because one of my lines... We're all community people here. I mean, this is a celebration of community, but one of my lines is community is your first defensible asset. And I love thinking about that as the IP within that. I mean, you're just nailing it. Glad we have you on today at the end of the day. There must be a reason for that. There must be a reason we keep inviting you back. There must be, yeah. There's something. Maybe it's just to keep us entertained at 5.15 p.m. on day one. Which can also be the case. What do you think about chat GPT? I think it's that and I think that the fact that it could also open you up from a security risk perspective. And I think that worries me to no end. You still have to go through and actually review it. And you have to be smarter than the code it produces. It's very much. I mean, it kind of reminds... Like I have one of those autonomous driving vehicles, right? And it's cool. Like, you know, it drives over the Tesla. Yeah. And it's a cool car. It drives for a while. You know, I have to... It asks me to put my hands back on the wheel for every while. Right. Would I pick my children in and ask them to drive across country? Absolutely not. Right? Is it assisting me in driving? That makes sense, right? Maybe we get to the point where it's completely autonomous and drives, but we're far away from my perspective, right? And so, it feels like the same thing with this all as co-pilot and chat GPT star. Would I have it define a security policy for my cloud and then walk away and just trust it? Absolutely not. I have to be smarter than it to review what it came up with before I can trust it, right? But there is utility in having it help. There's utility in having it assist a human in coming up with stuff. But you still have to be smarter than it, because you're liable. You're the one that's actually going to deploy it. Just like there's human error, there has to be human oversight. Yeah, go for it. So, that kind of brings the future aspect of this into the conversation, like a year from now. We're sitting in Amsterdam again. It's KubeCon again. How many of these vendors have actually integrated features that's like this and have done the smart thinking around the checking and the fact checking. Everyone. Everyone will have done that. And so what's interesting to me is yes, we're absolutely scratching the surface, but I genuinely wonder how far we can go by having the smart people go and build solutions based on this. Well, that's the thing that's interesting. Like if you assume having chat GPT integration it's kind of like having a website. How differentiated is having a website, right? Like, okay. Yeah. If you assume that that's the case, then how do you differentiate then? If it's actually such a how do you differentiate? I think it comes back to private data. Who has the data set that's private that can actually use it to differentiate on models? Well, exactly. And the real winners are going to be those that can use private data sets in a self-hosted manner, create value out of that and sell it to customers. Yeah. Like that's going to be the number one thing. So right now we're experimenting. We're literally just looking at a toy saying, hey, this is cool. Yeah. But we're not going to use a toy for anything substantial. And so it's waiting until someone comes up with, you know, the self-hosted ideas based on our chat GPT. Yeah. Yeah. I think the thing that was super interesting about chat GPT is the fact that it's really a UX. It's the user experience that made it accessible to other people. It's not like, you know, LLMs haven't been around for a while, but I think that's the key. I think what will be interesting is will it help people go through and do troubleshooting or do tracing or things that maybe aren't so proprietary for that where it's looking for patterns and it goes, well, after this happens this happens. That to me seems like a normal thing that would happen. Or recognizing an abnormality or something like that. Yeah. I mean there are already people in the security space that are doing integrations that are here. And I think that's the thing. But where does that data live and is it now public or is it learning on those anomalies? That I think is that'll be interesting. And there was, I mean, some of the people who are on the floor in the observability, one of the billions out there doing observability, they have, I was having a chat with them last year and they got angry at me because I was like, you know, I don't like to term AI ops and they're like, why? I go because nobody really has AI. It's not taking an action or something of that nature. It's ML. I'll give you that. And I think that's the interesting thing is where the AI washing of everything next year is going to be, you know, in Paris is going to be just crazy. I'm like, it'll be interesting and fun to see. We should do an AI bingo next time. Oh my gosh, we should definitely do a buzzword bingo. I'm so here for that, especially in this space. And I'm curious to see how it feels like a lot of people are very all in on what this means, which I think is kind of an interesting, I feel like we've been, you know, there's always some sort of hype curve dragon that we're all chasing. But I think in this case it's because of the functionality that you pointed out. And I actually find the UX a little lackluster in my personal opinion, but that's more of a design perspective than a utility perspective. But I think what you're pointing out is it put the tool in the hands of people to see how we're actually going to use it. Which before, it's really just been big companies secretly building these models and we're not even quite sure what the AI side of it is or like with facial recognition and then it turns out to be racist or it turns into, you know, there's all this stuff that's happening. I think actually getting this out into the world and letting the open source community mess with it a little bit more it's going to be our best defense against something nefarious happening in this space is having more hands on deck and a much more diverse group of people interacting with the tool. Yeah, that's much better than, you know, friends or Italy forbidding using this stuff. Like, that's not a solution. We have to get our hands dirty. We need to play with it. We need to make mistakes. Oh, yeah. Do stupid things and then, you know, gradually we'll get to a point where this is actually going to be useful to us. Like to your point, observability a year from now is going to be completely changed. Yeah. It's going to be completely changed. We're literally scratching the surface. It's such a 90 days in, right? Yeah. I mean, something like that. Yeah. I mean, it's crazy to just think about what a conversation it's become, how many tech bros feel like they're an expert, you know, like a pro when it comes to chat, GBT and then, you know, we're seeing everything from students doing their homework, leveraging it to, you know, to your point, people putting proprietary and private data just straight up in there like it doesn't matter. Yeah. And I mean, yeah, it's just, it's wild. It's really wild. And we're gonna have to do this. You're a veteran, you guys are obviously going to be around. What are we going to be saying at the next Q-Con? What are we going to be saying in Chicago or in Paris? Any predictions? AI. Yeah. So, you mean we're going to be doing this panel again? Yeah. But with a little, with a slightly more, you know, 180 days in? Yeah, yeah. We'll have to Bingo cards all set up. Maybe we can... I actually really want and I feel like we just should. Yeah, I think we're going to be also talking about platform engineering and how there is still continued consolidation and cost reductions happening, because I think, you know. Oh, my favorite topic, left of engineering, yes. Oh, yeah, well, I think, you know, when I worked with the devs, they said, just don't make me an SRE. That's what they always said, is like, that's the last thing I want to do is get woken up at three in the morning, because, you know, something would bump in the middle of the night, but I think it's going to be an interesting year for platform engineering. Let's make the SREs go away and let Chad GPT take care of it. Yes, on that note. There you go. Basam, Rob, you know, thanks for being here. I'm so excited. Thank all of you for tuning in live or in the recorded version of this at home. We are here in beautiful Amsterdam at KubeCon Europe. This is our last segment of day one, but don't worry, we'll be back for day two and day three. I'm Savannah Peterson. You're watching theCUBE, the leading source for high tech coverage.