AI Disrupts Everything but VMware

Well, good morning everyone, and welcome to Infrastructure Matters, episode 56. We are here again. Good morning folks.

Good Morning. And we've got one face that you guys all know. My, my, my dear friend and, um, technical brother is Steve Dickens and, um, from New York City or New York State, I should say.

Not the city. Yeah, a Little bit up the river, uh, up the Hudson, Up up the Hudson. And I've got two other faces that you're gonna start to see on a regular basis on here, because we're kind of broadening where top and infrastructure matters, and it's still about infrastructure matters, but we wanna bring some really cool people onto our, our broadcast.

And that is Keith Townsend welcome here up in Chicago or Illinois someplace. Someplace, yeah. Just, Just south of Chicago.

Great to be on. Can tell 'em a little bit about what you spend your time doing? Well, right now, I'm, I downloaded all of my Twitter data, uh, and I'm trying to run, create some python madness to understand how people engage with me.

So I'm a geek just like everyone else on the, the call. Uh, and I'm a CTA over at Future Group, so I get to, uh, cover a lot of, uh, the great stuff that we're gonna talk about specifically infrastructure. There we go.

And Diane, hitch Cliff, who is our, who is our CTO Elite, or CIO lead, I should say. Hit welcome. I'm not even sure where you're located.

I'm based, uh, just outside of Washington DC Oh, wow. I love that place. And why don't you do a little intro about what you do and what you cover and what Sure.

Yeah. I, I'm well known for covering strategic IT topics, things like digital transformation now, things like ai, uh, cloud, cloud repatriation, um, but basically everything to do with, uh, it, uh, it service management, things like that. I work with our CIO audience.

I have been working with that, that audience for, for over 20 years. Um, so I'm the head of head of practice, uh, for, uh, for CIO at RUM Group. And it's great to be on here.

Thanks. Great. And welcome.

And you just got back from, uh, New York and the Gemini launch? Is that what I hear? We had a little jealousy going on between the four of us.

Yeah, I got, I had a, when you posted your video, Diana, I had a real case of fomo. I'm like, also, he got the invite to that event. Yeah, it's pretty, pretty funny.

Um, they're, uh, it was, it was kind of super secret 'cause Google did the big announcement, uh, with Curian and everything earlier this week. But they kinda had this, this, uh, underground event and really wanted to bring in more customers and things like that. Uh, it wasn't like you, we would say our traditional tech event or tech conference.

It was kinda like, you know, now, now let's get hands on with all these new Gemini, uh, capabilities. Uh, and so yeah, it was really fun event. Mark Cuban was there.

We got to hang out with him. That was, uh, that was, it was really well done on Google's part, I thought. And that specifically, because you kind of like, let me go.

Oops. Oops. Specifically, that's Thomas Curian, as opposed to, Keith and I were with George Curian on the other side of the country, um, at, uh, in Las Vegas at NetApp Insights.

So, Yeah, hold on. TK and GK There. There you go.

There you go. Okay. They've Actually ever done both of them on the main stage west At the being time.

Yes, Yes. They have This earlier today. And my favorite story that they've always, George has told, I'm sure Thomas has told this, is that their mother, you know how mothers are, right?

Yep. Basically told 'em to leave their business cards at the door when they come home. Oh, that's gotta be hard.

I love it. Okay, so let's get into the news. Um, first of all, I think what I'm gonna just, since we are already talking about NetApp, we're just gonna jump into the NetApp Insight Conference.

Um, Keith, you were there. What did you hear? And, uh, what was the important takeaways?

Well, one important thing you folks talked about last week, but there was a lot of AI coverage with, like, with every modern conference. And what I was most interested in was all of the cloud conversation. You know, I, I'm, I'm going to make a somewhat controversial, but I think accurate statement.

NetApp leads hybrid storage. The, their ONTAP everywhere is, uh, extremely effective. Customers that I talk to really, really do like that operational consistency.

NetApp leaned into it. Uh, they've struggled with in the past telling a cohesive story across their portfolio. They've been kind of a bits and pieces.

They have some cloud stuff, they have some storage stuff, and they have some data stuff. The message is getting better. And the hallway track was what you folks talked about last week, which is VMware migration.

Customers are still trying to figure out how to not be locked into Broadcom. And that, that really came out when you were, you walk in the halls and talking to the people there, because while Green Stage was a lot on the ai, the inside of the IT organizations are grappling with this entire VM where Broadcom licensing problem, um, on the big announcements coming from NetApp, they did, did announce a new aggregated disaggregated system and with a direction to be shipping that in 2025, which will also have vector vector database capability, fully ontap. Um, it's according to the one oh one-on-one discussions that I've had, this is already in the hands, early stage already in the hands of their customers or a few customers.

Um, so we'll be looking at seeing it going into early stage testing the beginning of the year with shipments coming into the summer. So very aggressive, very, very aggressive process. We having a research note that will come out, um, next week on this, um, with some more details.

And then of course, some big updates. I kinda love Sandeep Singh, who is, who heads up there, their enterprise storage piece was up there hugging their, hugging their storage system. You had to laugh at the enthusiasm around storage.

Okay. That's enough of that. Kimberly, just, I didn't get to go to the event.

Some demos stand out for you. When you get to these shows, you get shown a lot of things at Google Cloud next. Last year, probably the best demo I saw was the demo with NetApp connecting in their filers into Vertex.

I think that's a really, really fascinating kind of private enterprise. All the data sits on these filers, you know, people have all their sort of corporate files is sitting on these NetApp filers and then being able to interrogate that with a sort of Google sort of vertex interface. Did they, that was kind of previewing and they had some early beaters at Google CloudNet.

Did they update that? Because that, that for me is a perfect use case of NetApp. Well, And that's the really, that's the big step that it and most organizations has to make, is they have everything in relational databases and graph databases to a little extent and a little bit of NoSQL.

Um, but most enterprise data has to make into these vector databases. If you wanna really make it actionable, you wanna plug it into large language models and action models. Uh, and I, I think people are underestimating the amount of work they have to do to vectorize their data and the fact that they really have to do it.

That one Sense. Yeah. And Kimberly's more of an expert on the actual announcements.

I look, love to hear from her. But one of the things that stood out for me was their data explorer. The NetApp is embracing this concept of being a data company and being able to, uh, they brought up a stat that I've shared quite a bit, which is CDOs chief data officers, uh, complaining that their data analyst spent 80% of their time normalizing data, you know, setting up environments, et cetera.

And NetApp is going directly at that problem. When you think about a 80% of my time or my smart people's time are spent doing anything, that's the infrastructure problem or opportunity. And NetApp is going after that opportunity.

So yes, d Diane, they're looking at both your structured and your unstructured data. Feeding that in the xp um, blue XP that Steve talked about is this other layer that has to do with classification policy. They demoed that, um, there, this is part of what's gonna roll out this next year in terms of all of those pieces.

The thing that you saw, the Vertex piece is real. Um, it's out there. The, um, they have, since they are first party storage in all three of the big hyperscalers, which is fabulous.

Um, they're also have had, um, a lead in tying into each of the hyperscalers key data database technologies such as SageMaker, bedrock, um, vertex, um, and two terms. The other, so there is, they are definitely streamlining up in the cloud, but they also recognize that this is gonna have to be on-prem because of privacy, et cetera. Okay.

Which I'll kind of go into the next one, which is interesting. Um, and then we'll dive into the, the Google piece of it, of the announcements. Um, the DOJ um, rolled out some something called evaluation of corporate compliance program, um, in regards to ai.

And, um, within that paper, that paper, there is questions that are directed at AI compliance. And, um, I'm gonna read a couple of them. How does the company assess the potential impact of new technologies such as AI on its ability to comply with criminal laws?

How is the company curbing any potential negative or unintended consequences for the use of these technologies? Um, how is the company mitigating the potential from deliberate and reckless use of these technologies, including company insiders? And this is coming from them, you know, they're inspecting the companies to make sure this is all, you know, based on stuff that DOJ would go after the companies with.

IT has, you know, we know that the companies are taking their time to look at what are the rules and the guidelines around using ai. Um, maybe Diane, you could talk about what you're talking with, with your CIOs and how they're, they're looking at these type of things. Well, the, um, the compliance regime around AI is gonna be difficult because you have the EU with the, the, um, their AI act, um, which I think they learned a lot from GDPR because, you know, there's 20 or 30 million companies that are out of compliance with GDPR.

There's no way they can do the enforcement on it. And I worry the DOJ is gonna put themselves in that same position that there's just, everyone's using ai and there's no way they can police everyone or monitor everyone. Uh, and it's, it's gonna be a significant challenge, but CIOs are very concerned about if, especially if they're their global or multinational organizations, they have to have all, uh, comply with what's going on in California, what's going on in Europe, what, you know, what's going on nationally in the us.

Uh, and, uh, and EAs got, I mean, they're down in the Middle East, they have their own regulations. So it's, it's a challenge and it's a headwind because, um, it doesn't matter how good your ar AI is, but if you can't make it compliant, they can't use it. Mm-Hmm.

Uh, and that, that was one of the interesting pieces where we we're gonna get into Google Gemini, but they showed that they're by far the most compliant AI on the market right now because of that reason. Otherwise, you can't, the companies just can't use it. So that's, that's kind of where we are.

Yeah, and that was one of the things that I got pointed out with NetApp, is that when you bring it into that database, you're putting your policies in, and that those policies and those privacy issues and compliance issues continue with that as you move forward through the, the pipeline. So, um, you've Seen IBM on the same thread from the get go with granite. I think, you know, if you think IBM I'd have, maybe you don't think, as, you know, on the cutting edge of innovation in ai, what you probably do think of them on is trusted, compliant and sort of a trusted partner for a lot of enterprises.

So they've been leading in with that sort of certification compliance trust message from the get go. I, I think it's going, I mean, Diane, to your point, the technology can be as fascinating as you like and as innovative, busy, like, but if you can't use it, then kind of what's the value? Yeah, I, I, I, I push back in an interesting way, Steven.

I, I think I wouldn't call IBM not cutting edge in, uh, ai, especially large language models. If you're building a large language model for general purpose use, like granted you're cutting edge. There's only a handful of companies that have the financial and technical withall to do that.

I would also, uh, highlight that, you know, Diane pay homage to what's going on in California. SB 10 47 is now on Newsom's desk, the governor of California to sign. And the intensity a against the measure has been really intense from the VCs.

Can I Interrupt you before you keep on going? What, for our listeners, what is the, the legislation that, Um, uh, get to the legislation? Uh, you know, there's this desire to protect consumers.

It's California. It's all about protecting consumers, protecting consumers. It's Fascinating hearing you guys talk about the US and California, by the way you did it earlier, Dan.

It's, it's like us. I'm in California, but that's just me. And, uh, you know what the, uh, California effectively, uh, sets these regulations for the US because these companies are based in California.

Yeah, yeah. This is, uh, the power of the, uh, California legislature in government, this ability to shape our technology regulations here in the US effectively. So this idea that you want to protect consumers well against what just happened with, uh, this AI company that made these claims that the FTC put a, uh, lay down on for $193,000 fine of, of making these claims, of being able to create a lawyer level legal documents without actually backing up or doing any of the testing.

So the legislation is intent is to ensure that these AI companies that are big enough to impact the overall industry have these controls in place. So this, you know, regulation, um, enforcement compliance is coming from several different layers. The DOJ is as you as we've highlighted, uh, with companies in compliance and then the actual creation of these documents.

So back to you, Steven. I think IBM more than probably any company is pretty well positioned to meet the needs of, of companies because they are all about compliance, bureaucracy, and all the things that, you know, we kind of tease IBM about in the tech media is, you know, kind of their strength and why, why customers look towards, yeah. Being compliant can be, and being sort of a good and connected to the government is probably not a bad thing right now with the winds that we're seeing.

And, and IBM's been doing that for over a hundred years, being connected to these government agencies and, and being sort of compliant. So yes, maybe their superpower going forward. So since we're on the legal eagle kind of stuff, we're kinda go to the Google versus Microsoft, um, gig that's over there in Europe.

I guess Google is, is, um, suing Microsoft or non-competitive practices or something like that, which is kind of like, gosh, you know, it's kind of like having, you know, your kids in the middle of the house fighting And you don't really care who started the fight. You just want 'em to end. He just is like, really, guys?

Well, it's just, it's just so unusual for me. Google kind of prides itself on being a really ethical competitor in the market, is the way I perceive them. They don't really do the competitive knocking copy.

They don't go after their competition. They don't really mention their competition ever. You know, it's all about what they do.

My interactions, you know, whenever there's an opportunity that you can say, Hey, you can swing a few punches here competitively. They go, yeah, we don't do that. And then they go and file a lawsuit in the EU basically.

Well, I, I think they sense an opening. Yeah. The, um, Microsoft's been an incumbent for so long on the desktop, uh, on server inside, inside IT departments as, you know, the top vendor, you know, the kind of the infrastructure vendor.

And, uh, with CrowdStrike, everyone realizes this, it monoculture might not be healthy. Mm-Hmm. Needs to be more heterogeneous, uh, more resilient that way.

Uh, and, and also Google's done some really smart things in terms of, they now have, uh, go Chrome OS on a, uh, USB stick. They, they were actually, they recovered a lot of computers and got people back up and running during ransomware incidents during the CrowdStrike incident, because you can split their operating system from a USB stick on any locked up computer and get back to work, uh, in the cloud in just, uh, in just a few minutes. So they have a, they have a, uh, I've been tracking, they have a variety of advantages over Microsoft in certain ways they've never had before.

Chrome is the definitive browser out there. It's where everyone spends their time and, and runs all their applications. Uh, and, and they're, they're now pushing Chrome Enterprise, uh, to really, um, press the lead in terms of their, their own a big piece of the desktop for the first time.

And they, and now they really wanna expand that. Uh, but they're still having a really hard time. This the, the sheer gravity, uh, uh, that Microsoft has inside of the enterprise.

Uh, and so I imagine this is how they, they're figuring in, in a a, in a regulatory friendly part of the world, they might be able to, uh, gain just a little bit more advantage and, and, and get over the top. You mentioned an interesting point there, Diane. The, the monoculture.

It's interesting for me, I hang around with the Lennox community quite a bit. I just literally recorded a podcast this morning for the Linux Foundations. I'm kind of tuned into that.

There's more Linux running on Azure than there is Windows server, you know, which we, and that's a start from a few years ago, which kind of always blew my mind. I wonder whether we're just that, my, my point on sort of what Google's doing, I understand why they're doing it. The Windows server licensing on AWS and, uh, Google, it should be the same price that just the, the, I mean, that makes no sense to me, but I wonder whether that trend is normalizing anyway because of Linux adoption.

I mean, I don't know if, and Keith, this is maybe a place to bring you in if you are an enterprise architect and you're looking at a brand new workload, you're putting that on Windows server, or you're putting it on Red Hat, or slays or Debian or what, what's the sort of rubric? I don't, I think, I don't think I'd as an enterprise architect with a new, I don't know, system, I mean system of engagement, you know, whatever it is. I don't think I'm thinking, oh, I know what I'll do.

I'll put this on Windows Server. I'm thinking I'll put this on Red Tie or slides or, or something Else. Yeah.

Uh, Microsoft, What whatcha are you hearing, Keith? Yeah, Microsoft would love for you to, uh, consider Windows Server 2025, which is gonna be released shortly with Nanos. Really, you know, it's a respectable os but I guess people still on XP out there.

Okay. There's people still on xp. I'm going through a migration in the, in the, uh, six five, uh, data center from VPE to whatever, and I got a bunch of Windows servers I have to upgrade, grade lifecycle management, whole different conversation.

We'll have one day. But I, I agree with you that if we're going to build a new app, I don't think anybody seriously considers Windows as the default. It'll have to be that there are some capabilities, some piece of software that's needed, that you're going to use Windows Server that is less and less the case.

I can run SQL, uh, Microsoft, SPL on Linux now for the past few years. The one thing that I would, you know, kind of, uh, push back on is this idea that pricing should be the same Oracle Cloud, I mean, Oracle Cloud, OCI, and the price of, uh, using Oracle and what's supported in Oracle server versus what's supported in, uh, um, Oracle Cloud and what's supported on Oracle Cloud versus what's supported on the other major cloud providers. And, you know, AWS and Oracle just recently announced real support for, uh, Oracle database on, uh, AWS just a few weeks ago.

So vendors do this, this is how, you know, this is their can. To your point, Oracle will really explicit in their briefing with me. The Oracle is priced exactly the same Now.

It is now, yes. AWS Google, Microsoft, and on OCI, it is completely transparent, completely the same. And I think that's a customer friendly policy.

Google obviously is the hybrid cloud, uh, cloud provider. They want complete portability because that is their strength. So this argument makes sense from a customer perspective.

I don't care where my Microsoft workloads run. I need them to run what I need them to run. So if I have to consider price as a consideration, that is not customer friendly.

Uh, so I hope Google, you know, that I, I, I don't like to pick sides in these things, but I hope Google does win, uh, or at least get Microsoft to push more towards where the, I think the industry is going, which is, you know, equivalent pricing. You know, there's this concept, um, and I'll use the term that Pure just announced, uh, pure Storage Universal credits that, you know, they're selling their software universally at this price actually, and they're even selling it across all of their different platforms. Um, whether or not you are in AWS on-prem sta or whatever, and, you know, that is becoming more common, um, because of exactly what you guys, so hopefully we'll see if, if Mr.

Larry Ellison can do this. Surely. I mean, Keith, you may, you, may you frame this the right way, but I'd maybe pivot it a little bit.

I don't see it as a who wins the case. Uh, we, we, we work with both. I'm not certainly gonna pick favorites between Microsoft and Google.

I think what we need to see happen is if you want to use Windows Server, it is the same experience. So part of the claim is that there's different features and different access to patching depending on which cloud you're on. That makes no sense.

And pricing needs to be consistent. I think if what happens out of this case is that it's Windows server and Windows Server 25 when it launches is the same price on every cloud and the patching updates and sort of software life cycle for it is exactly the same regardless of whichever cloud. That's good news.

The other piece we've not covered is the, the bundling of teams. I'd love to get your perspective down on the bundling of teams and kind of what that does to startups. You know, is that good For Slack?

Is it good for Zoom? WebEx? Definitely good.

No, no. But Before we get into that, I did wanna address the differentiator for cloud providers. I think you do have to allow cloud providers to differentiate.

We see this in with Broadcom and uh, VMware, Broadcom now allows you, which is a welcome change for a change, allows you to take your VCF licenses and go across the cloud providers. But Google has vastly different network capabilities than a AWS. And, uh, OCI has vastly different network capabilities than either platform and that, that differentiation should be allowed in the marketplace.

So patching and updating and et cetera, those are value added pieces. All I care about as a architect is that if I want to run my base foundation code on one server to one platform to another, that pricing should be the same. The value add patching, I consider a value add.

'cause it's even a value add in Linux, then sure, I'll consider buying a higher level of, of lifecycle management if that's available. And that should be a differentiator. Yeah.

So we could go on for a long time here, but we're gonna run out of time and we definitely wanna get to the Gemini launch, um, and what happened up in New York or the big launch. So Diane, I'm gonna turn it over to you and, and let us know what happened. Yeah, You bet.

Uh, so, uh, Google has, is basically putting their ai, uh, models, uh, under the Gemini rubric throughout Google Workspace. So that's, that's their big competitor to Microsoft Office. Uh, and you're gonna see in everything, if you have a Gemini license, if you, if you, you have a commercial license, if you have the, using the free version, you, you won't get very much, but you're gonna see it in simply everything to automate all your tasks.

Like if, uh, if it sees receipts coming into your email, it'll say, do you want me to, you know, put this in a, in a spreadsheet for you? Do you want me to file, you know, put this into an, an expense report application. It'll help you do your work for you, and it'll constantly suggest things that can be done based on what's coming in.

Uh, and out of the various applications that you have. If you're inside a meeting, it'll, it'll live, transcribe it, and then suggest, uh, you know, uh, pieces of information that could help support and enrich the meeting. Uh, it's, it's gonna be ambient is basically what ambient AI is really the, the message there.

Uh, but they, they're really gonna put AI in everything. Uh, I, I've found out that inside every, uh, modern version of Chrome is, um, Gemini nano, and it's also that same model is also on their phones already. And so they, their, their newest phones, so they're gonna put, AI is simply everywhere inside, uh, you know, the, the, the pane of glass that all people are using.

If you're using a Google product, you'll see Gemini in there and you can customize it. Uh, they have these things called gems. Um, and now, now end users can customize them.

You can, you can can provide let's say a bunch of product information and create a a a a Gemini or, uh, a based, uh, AI model called a gem that's been tuned. Anyone can do it self-service. It's, you can kind of, it's kind of DIY uh, you know, anyone can build a rag model and then the gem is done.

And, uh, you can, you can attach agents to it, applications to it. You can use like a regular AI model, but it's tuned with your information for a specific task, like answering customer questions about a product or a new launch or, uh, whatever it's you want. And some, they also has a, a lot of coding support as well.

Uh, but the message is, is they're, they're all in on ai. They're using AI all the time everywhere on what you do. And they have Mark Cuban on stage saying it's how it's supercharged his business and how he works every day.

And it makes him 10 times more productive. All these agents running off doing everything that he needs to do, uh, automatically answering emails and his tone. It knows what it's learned, how he responds to emails and, and does it for him.

Uh, and he loves it. He's answering a thousand emails a day now, you know. Um, Donald, Do you have to just so we're a workspace customer here.

A future. Yeah. Do you just have to, is there a turn it on and pay an extra?

Is it from a pricing point of view? Is it flick the switch to get Gemini and does a different payment? Obviously we have a, a low level of that, but, uh, yeah, there's another, there's another tier.

You have to get to be able to get all the bells and whistles that are coming. And this is just to start. They're going to be providing some more, but every, everything's gonna be basically automated for you all road tasks.

So yeah. So if I go into my little nine block thing that's over there on my Gmail, um, and I look at the, I don't, am I gonna see it come up in there? Yeah.

So this, this actually happens to, Here it is. I see it now. Okay.

Gemini is right there. Okay. Yeah.

And I, okay, Who is a live demo? Keith, come on. 'cause I want turn.

I wanna, Diane sold me on it. I want to bunny on. So I bought the early version and it was not, uh, it was not there, uh, from a, from what all the features they released and talked about yesterday.

It is super exciting what they talked about yesterday, the earlier versions I've been working, I've been messing around with the Gemini code assist within cs, uh, and within VS code it is, um, we probably need to dedicate a whole show to like AI productivity and uh, like what, what, what's real and what's not. What, what can you, uh, realistically expected to do and what, uh, what what, what it won't do. And it is a fascinating like, exercise in what AI is, uh, capable of and not The other thing, Don, and I'd be interested to see what I said in the room.

I went down the rabbit hole around the contact center announcements. That is fascinating. So the traditional aid by eights, the genesis, the sort of Vons of the world, these people are in this traditional contact center space.

I know you track a bit of this as well. Fascinating. I I see that as literally a bomb going off in the contact center as a service space, you know, seek compression from a licensing point of view.

I, I just see it really disrupting those vendors have got a real choice of be the wrong side of the innovation and get crushed or lean all the way into AI over the next 18 months. Radically transform that, transform their licensing models, their business structure, how their tech stack works. I'd really love to get your perspective, 'cause I know you track that sector as well.

Yeah, well let's just do a, a, a quick soundbite on that. Yes. It's, uh, that they're entering a space that, uh, has had a moat around it with these, these, these companies have been willing to, to integrate a lot of the more legacy PBX and, and other pieces.

So you can say no matter what, you have, you, you just have to work with us and we'll connect all your different pieces. And Microsoft and Google is the big guys have been reluctant to get into that full kind of support, that full stack. So that's given those companies that gate by eight and others, uh, you know, uh, a, a, a breath of life, um, a room to breathe if you, if you, if you will.

Uh, and, and I'll even see Zoom, this is survival you were talking about with teams. Uh, you know, uh, giving away teams has been really making it difficult for the o the other, um, web meeting solutions. So they're going into that space too.

Contact center is a high value space. People want, wanna spend a lot. So yeah.

Now I think with uh, uh, Google bringing best in class AI to that whole space, uh, you're gonna see a real bruising competition over the next year. Two years will. Well guys, thank you very, very much Dion Keith, for jumping.

Steve and I, um, and you're gonna see it. Thanks For having us. Yeah, it was great.

Yeah. You're gonna see this four block here on a regular basis. Um, and, uh, talk a little about infrastructure matters.

Thanks for tuning in. Don't forget to say to like us, share us. Let everybody know we exist here.

Um, and we will see you next week.