Managed Platforms and AI with The Futurum Group’s Steven Dickens at KubeCon Paris 2024

This is Textron tv. Hey everyone. Welcome back here to Paris.

We're live at Q Cod. It's getting late in the day. It's after four 30 Paris time.

For those of you on the West coast at the us you're just getting up. Enjoy your day. I'm really happy to have my next, uh, guest on here with us.

It's kind of a reverse. He's usually on this side of the house, so I don't know. We don't know how well it's gonna, you know, is it's going to break him or what.

I wanna introduce you to Steven Dickens. Steven is an analyst with a, a company called the Futurum Group. We've never featured any fu folks on our, uh, broadcast before, so we're delighted to have you on.

Steven. Actually, we should start right off for people who want, who may not be familiar with Fu Futurum group or futurum, where, where would they go? com is where we go.

Um, I'm feeling underdressed, man. Well, I didn't get the Memo. This one didn't have any sequence or anything, so you're okay.

But Steven, well, let's first, before we talk about future ur, let's talk about Steven. Yeah. So give us a, your background.

Great. To be on the show. Thank you.

Thank you for having me. As I say, it's weird. It's normally do Me Looking down the camera doing the intro.

Yeah. I'll do this piece. Yeah.

So, and they normally sit me on the right, on our show. So I'm, I'm, I'm comfortable. But now Well, this way it hides the mates.

Yeah. So Stephen Dickens, despite this English accent living New York, spent the last, well, what is it now? Uh, 29 years in the tech industry working for Broadcom.

Well, what was Ca Sure. Worked for Xerox Straight outta college. Broadcom, HPE, most laterally, before joining the Futureum group as a product manager in various other roles at IBM And IBM, there's always a little IBM in There, 10 years at IBM.

Yeah, no, you find that, you know, I've been in the tech industry also 30 something years. There's, there's always a little IBM touch in there. There's Such a, I mean, that business is exploding.

vin's doing a great job of that company. They really have, you know what? They're back.

I feel, Well, for the first time you can ask an I IBM er and say, what are you guys doing with the cloud? And they actually have a coherent answer. They do compared to what they've had, let's say the last five years.

But I think they've got two clear messages. It's hybrid cloud. Yes.

And it's ai. And ai. The whole company's rallying around those two things.

When you talk to them as an analyst, it's really clear what they do and why they do it. It's hybrid cloud and it's ai. Agreed.

You know, and you would think, I mean, with Watson, right? We were all, for many of us, I think Watson was probably our first, uh, exposure to ai, though, I don't know if we called it ai, but it was Watson. I mean, they, they've been doing It for a long Time.

I it's been in this game, you know, they've been beating Casper off at chess. Uh, then they did Jeopardy uhhuh. So they brought natural language to that with Jeopardy.

And now with what's next, they're doing a great job. I think my conversations with a lot of the vendors is there's lots of cool people out there. There's lots of people that are coming up with cool technology from the IBM perspective.

They're trusted. Yes. If you want to now do indemnification of a model because you want to put it into production, and you don't want a New York Times, Don't wanna be The headline open AI kind of lawsuit situation.

That's the key piece of their story for me. Indemnification, they've got the tools, they've got the stack, they've Got all the Capability. But for me, the key takeaway is indemnification.

Yeah. Well, for the AI piece of it, for sure. Yeah, for sure.

com days, I, I helped take a company public. We were the second largest two IBM Lotus notes back in the day, if you remember the day. Yeah.

Back in the, I've used loads of notes for 10 years. Well, It was a great IBMs even Off. I News for you was better than Exchange.

It's still better than exchanges today, but I'm not here digressing anyone. Yeah, we Digress. We digs.

But the thing about when we went into enterprises with, to host notes, and this is before there's a cloud multi-tenancy, none of this, but the fact that you had big blue behind you, They trusted. Absolutely. I Mean, the same.

And there was an IBM rep at every one of the companies we went into. They're saying Nobody gets fired for IBM. I think that was obviously the industry Manager.

That was very true. I think they've drifted away from that, maybe. But I think that's back.

Yeah, I think that's back. I, I, I don't disagree at all. Um, so a great background for someone who wants to be an analyst and tremendous, I mean, you know, household names in tech.

Yeah. What, is there a specific, uh, specialty you're covering a future Group? Yes.

So we, we've structure our recent advisory business along, uh, what we call practices. That's pretty much a vertical stack by technology. I cover what we call hybrid cloud infrastructure and ops.

That's everything from a public cloud ias. So working with the Google guys, AWS, Microsoft, O-C-I-I-B-M Cloud, working with those guys right through to I cover the mainframe from a technology. Do you?

Yeah, I'm that. So that's near and dear to my heart. Throw, go to It, steer to a lot of people's hearts.

And then it's all the supporting ops. So observability from an operations, you know, we've got another guy who covers day zero, day one, day two. I, I'd like to say I'm day 300.

You Your second week. Yeah. Second week.

It's stable. It's steady. It's monitoring, it's log monitoring.

It's make, I describe my practice as the bits that keep the lights running. Absolutely. Very cool.

I don't get any of the sexy stuff. Well, you know what, but sexy always doesn't win the race. So it's, it's hilarious.

I get involved in a lot of the conversations around availability, performance, scalability, architecting for security, all those types of conversations. And that's fascinating to me. I gee count on that stuff.

Yeah, no, I, look, we all do. Yeah. Quite frankly, to be in this business as long as we've been in this business, if you don't geek out on it, oh yeah.

It's gotta be. That's work. Well, I mean, I think for me, the bit and why I'm enjoying almost three years now at future employee number four, Uhhuh, um, why I'm join it.

It feeds my inner nerd. I get to geek out, be intellectually curious. I get downtime at an event like this, and I'm like, can I go get a demo?

Can I learn? Right. Absolutely.

Which is, you would think I'd want to sit in a corner after this event, but I want, I'm key to learn. You know what I, I think there's a little bit of that in all of us here, and that's why we're here. And actually, that's a great segue.

Why we, why we're here. Keep coming, right? Yeah, It is.

We didn't just come on our holidays to Paris. No. Well, no.

Next week is my Paris holiday. Enjoy. I'm staying enjoy.

Um, actually, my wife's probably getting on a plane right about now in Florida behind the JFK. Your fun Stops and then your fun with your wife starts. We're live on tv, so I'm gonna say yes in case he, In case he's watching Right?

In case he's watching. Exactly. He said it.

I was here, Uhhuh. Um, that took me off track. But Cube Con.

Yeah, cube con. So, you know, I've been coming a cube con a long time. Yep.

Docker Con and Cube Con. And look this to me. Well, not just, to me it's the biggest cube con they had so far.

It's, I think they, what I've heard, 13,013, I, we, I just interviewed Arun Gupta about, I dunno, 45 minutes ago. He's the chairman of CNCF. Yeah.

And he, he said he thinks they're just under 13,000. 12,000 plus. Just under 13.

Well, They had the biggest one in Amsterdam a year ago, and now they're big beating that again. Well, you know, I was talking to Arun about this. I don't know, did you go to the one in San Diego?

So I've done Seattle. That was 2015. Right.

I launched a product there, so, okay. Then I had a bunch of fun. Uh, then I've done probably the last four.

Okay. Seattle was the last one before covid. Yeah.

And it was, it was not as big as this. I think it was 10,000 people, but it had this buzz, this energy, And we're back. It took us this long to recover from Covid.

We're, we're back, but we're back. Because if you look around and you could see at home, the next booth is right here. And they're two at a time all the way down.

And it's a huge block. That's probably a hundred vendors here, I think. I think there's more.

I think Arun said something. No, there's 180 projects in CNCF. Yeah.

And there's, I'd say a hundred vendor booth. At least there's a hundred booth. But then you've got the, the projects each have a booth.

Yeah. And, you know, all the various communities have a booth. Yeah.

So this place, it's huge. The CN CF events team do a great job With a relatively small amount of people. Mm-Hmm mm-Hmm.

And volunteers for the most Part. Yeah. A lot of it Really.

You did a great job. Yeah. No heads off to them.

So look, AI's the story of the year, I'm sure we're hearing a lot of, you're hearing we Went, we went a good 10 minutes without mentioning ai. I tried. I tried.

You did a good job. Well, so, and let me, let me keep that in the same vein. Besides ai, what else are you hearing from, from your contacts and your interviews?

So I'm, I'm almost coining the term and I'm coining it here on live tv. Okay. Jacket to hoodie ratio.

Okay. It Sport, the sports coat. I'm gonna coin this and I'm gonna track it.

It's the sport coat to hoodie ratio. Let me explain what I mean. If you've been tracking Kubernetes over the last 10 years, it's been tin, it's A hoodie crowd.

It's a hoodie crowd. It's tinkerers. It's how do we get this stood up?

How do we get something quick into production? Hey, I want to play, I think that was probably the first five or six years of the 10, last 10. I look at it now, and I, as I say, the jacket, it's A lot more jackets.

The sports coat. I think the, the companies that are on the stages doing the sessions, it's big household names. I had the chance to interview dominoes around what they're doing.

Yeah. With solo and service mesh, you know, complete rearchitecting of a monolithic application to Kubernetes and Cloud Native. So I think for me, the grownups are in the room.

No doubt. We were, we, Kimberly Bates and I have been tracking that dynamic over probably the last three cube comm. It feels like platform engineering's kind of kicking in.

We're starting to see this in production guys, guys like me, availability, performance, security, scalability, all those ities. And now starting to come cost optimization. You know, It's getting real.

This is part of the maturation of the whole thing. And That's a good platform. Engineering is a good thing for this community to get rid of.

It's not the death nail of DevOps or, or cloud native or anything. It's absolutely, first of all, I've got a confession to make. Okay.

Had I known you were gonna bring this up in my hotel room, I actually brought a, a sports jacket, a blazer, and the cashmere caught. You can't hide that skirt. No, no.

You can't hide that shirt. Well, today it was this, but tomorrow I have a Brooks Brothers blue and white pin strike looking good with a blue blazer and a blue cashmere sweater. So Maybe the jacket's a hoodie ratio kicking.

I'm single handedly gonna raise that ratio. But, but you're right. It, it's part of this maturation.

You know, we're doing a, we're doing a research report. Mitchell Ashley, our, uh, tech strong research principle. We call it DevOps Next.

Because I think when you look at DevOps and cloud native today, it, it's no longer, as you call them, tinkerers, that are cobbling together a couple of open source tools and some command line interfaces and, you know, trying to make a go of it. It It's production. It's in production At an enterprise level.

Yeah. So I, I mean, I'm hearing words like opinionated, curated, managed. When I hear, when I'm chatting to some of the vendors, that's because this community is moving past the point of I want to hand stitch this together.

Yeah. I'm hearing the transition from pets to cattle. Right.

Yeah. You know, I'm hearing all these phrases. No, no.

Triangular We probably have heard before, by the way. Right. Yeah.

This, this is a normal cycle. I I back to COVID, I think COVID might have of, you know, delayed it, if you will. Yeah.

But this is the normal ma maturation cycle that we see in technology. I mean, I talked about pets versus cattle. Cattle looking after your Kubernetes clusters and being sort of so hands on and having to hands end everything about them, or having so many of them that you Can't, they're ephemeral And you, and you've just gotta roll them out.

You gotta have automation and tooling and management. I think that comes, that transition from pets to capital has happened. Yep.

So I remember my, well he's at ACORN now, but he was the guy who founded Rancher. Shank. Shing.

Yeah. Shank. He was the one who, that was the whole thing behind rancher.

Yeah. Right. Yeah.

It was cattle not pets. Yeah. And um, But I think we've got there around the room, the SUSE guys with Rancher, Which Rancher.

Yeah. 0 of Rancher Prime. Yep.

Yep. 0 OpenShift on four point 15. Yeah.

We're now at the point where this is not new. No mature Products and mature solutions. It's not bleeding edge.

It's great to be on the bleeding edge, don't get me wrong. Oh yeah. Absolutely.

But now enterprises don't love bleeding edge. No. And and I think that's a, a huge here.

You gotta Gotta run the bank or run the telco or run the retailer. You want to know it works. It's funny, I was, uh, I was at lunch on the street out here.

I listen to two kids. They were kids. They were probably in their thirties with they're kids.

You can Claim them as kids. Yeah. I've got a 20-year-old daughter.

How did that happen? Me too. I am 24 and 22.

But these were two young salespeople at one of the open source, open core providers here. And I was listening to them talk about the difference in selling to the enterprise now versus when they first started with the company two or three years ago. And it was just this point, selling to the enterprise.

You, you can't be fly by night. You can't be seat of your pants. Solid, Dependable.

And this is where the likes of the IBMs with what they're doing with Red Hat suse, some of these more established players who've been in this game, they're the anchor points of this room. You know, what AWS are doing, what Google are doing, what Oracle Oracle is doing. Yeah.

Those anchor points, the traditional enterprise players. There's still gonna be anchor points I think, going forward. Absolutely.

Well, again, if it follows and there's no reason it won't, if it follows our normal kind of maturation process, you will see some of those anchor points kind of swallowing up. We've seen it already with Cisco and ICE surveillance. Yeah, yeah.

You know, my hypothesis is that it may be not good for the CNCF events team, but there'll be 30% less vendors here within three years because there'll be concentration. So I asked Arun Gupter about that. Okay.

He said, look, we have 180 projects right now. Most of those projects have companies attached to them. Do we, in three years, do I have 300, 500, a thousand projects?

The CNCF motto is led every flower bloom. Yeah. Yeah.

Now my comeback was when is enough enough? When is enough? When is too much Project fatigue?

Right. How do you support all of that? I mean, for me, that's almost counter to this management curation, opinionated movement that I'm seeing.

We are we at too many Sprawl, project sprawl. Yeah. I've gotta deploy, I've gotta manage all that.

People are coming and saying, obs abstract that away from me. I want consistency. I want it pulled together.

Bring me an opinionated view based on open source. Right. Yeah.

I know. I don't want you to give, give me source. No, we're not Killing open source.

We don't want to, we don't want you to come up with, you know, the management stacks of 20 years ago, which is homegrown. I want it based on open source, but I don't want to be thinking about 180 different projects. Right.

I want you to bring me the 10 packages I need and give me a stack in a platform. You know, when I speak to enterprise customers, one of the things I still hear from them is, Kubernetes is still too hard. It's hard.

I Think, I think we're on the cusp of that not being the case. I don't think we're there. I hope so.

I don't think we're there. But my conversations with AWS, my conversations with Google, Microsoft, red Hat, Suze, all of those guys, it's starting to get to the point where opinionated, curated, managed stacks are coming together and people are starting to go, I've picked my, picked my winner horses And I'm riding it and now I'm gonna ride it. Right.

You made my bets. I think we're at that point. Yeah.

I mean, and again, that that goes right with this DevOps next kind of research we're doing in that, that we're moving from those cobbled points into platforms. Exactly. That that work seamlessly.

Exactly. And that's what the market wants. That's what the market wants.

I think. So they Don't want necessarily a hundred gonna be, this is That startup that wants to tinker and they're a software company and they want to build it themselves. And they got a two pizza team that can figure it out.

They're actually using the pure open source and making their own modifications to It. Yeah. Yeah.

There's always gonna be those guys. But I think if you are a bank at Telco, a retailer, a government department, you want no part of that. No way.

Well goes back to what we said about IBM and ai. Right. Trusted.

Trusted, trusted. And, and indemnification and, and everything else. When it goes bump in the middle of the night, you want one, one throat Person to choke or one butt to kick, or One person to praise, But to answer.

Right. You, You want somebody at the end of the phone is gonna go, it broke. I need you to help me fix It.

And I'm here and I'm here. Yep. Yep.

Let's talk a little bit about AI though. 'cause we purposely shunned it, but let's come back. We done Wel and only bring it up once previously.

Another. Well, That's why 20 Minutes. We've gotta be 20 minutes.

Yeah. So what about AI here? So I think this crowd, obviously a lot of AI workloads are gonna get deployed in the next 18 months.

Mm-Hmm. So they're gonna be containerized. You're not gonna go grab something from hugging face and put it in without putting it In.

They, they've got a bit of a black eye lately. But yeah. I mean, you're gonna want to containerize that workload.

Yes. And you're gonna want to put it into an orchestration layer that's trusted. So I think the buzz for this room is, that's just yet another workload explosion.

No, you're not gonna put that on bare metal. No. Maybe you're not gonna put it in a virtual machine.

You're gonna put it in a container. It's a cloud native workload. But you might run the container coob on a bare metal platform.

You might, but Kubernetes is gonna be part of that AI Stack. Oh, no doubt. Yeah, absolutely.

I Didn't, I don't see many cloud native workloads that are gonna go out that I'm gonna be in a container. Don't tell the guy at Broadcom that, uh, But as you say, where that Kubernetes stack fits Could be on a vm. And the guy at Broadcom's also got tons who are here For now.

For now. I'm, I'm pro that acquisition. I'm very pro the acquisition, because I think vm, I don't know what it was about VMware, but they couldn't stay single.

You know, they, they just, they needed a big brother kind of thing. But I Think a newer streamlined Broadcom slash VMware is gonna be good for the industry. Oh, absolutely.

Get Those guys back to innovation. Get them back to doing r and d, simplify the product. Slowed down, Simplify the, I mean that 120,000 skews, Could you imagine hundred And 20,000, there's no way you need it.

120,000, the price book must have been this big. Who the Hell would That? But now they're simplified V czu.

It's really simple for channel partners to understand. There's some bumps in the road. There will be, but there always is.

You gotta break a few eggs to make the omelet. Well, Hulk does a really good job of everybody. They've had a big choppy three months.

Yeah. It's execution time. Yeah.

And there's probably nobody better in the industry than Hulk than than doing that execution. Absolutely. Lemme tell you another term that I've been starting to hear around here a lot in the last two days that I used to hear and then I stopped hearing about it.

Infrastructure is code with ai, right? So generative infrastructure is code automated infrastructure is code. Put an adjective in front of infrastructure is code.

Yeah. I mean, I think this is the abstraction. People don't want to be, devs don't want to be dealing with infrastructure.

They don't wanna be anywhere near that. They want to promote their code and be done. If they can do that with an open stack layer, with a, an interface where they can do it with infrastructure as code where they can abstract the complexity of Kubernetes, looking to do all of those things and just do not want that complexity.

Yeah. And I think that's it. But again, I think that's another place where AI can help.

Right. I don't know whether at an infrastructure level, certainly from an observability point of view Yeah. That, that those just alert is on already.

Yeah. I mean, I've been briefed by the Dynatrace guys. You know, listen, Splunk Open Telemetry here as a project.

I think bringing natural language interpretation and a large language model to a huge log monitoring corpus of data. That's huge. I I tell you something, you're gonna laugh, but it, it's also made a, a a, an opening for traditional IT service management.

ITSM. Yeah. It's cool.

Almost again. Well, I Mean, ServiceNow, BMC, We, we had them here earlier. Yeah.

Those Guys aren't going away. No. You're still gonna need trouble tickets.

You're, something's always gonna break. You need to improve that workflow. I mean, that's kind of my more my space, right.

Than the pure ops side that goes with dev. It's gonna break. How are you gonna remediate it?

Agreed. The bottom line is, I mean, first of all, it's good job security for people like you and me, but the other thing is You just gotta adapt. Right.

But, but the laws of physics always apply. There's, there's always a, a path. Yeah.

And, and what we're seeing here, like C'S really in the rear view mirror, this is a return to a normal Right. A normalcy. And I'm excited to see what comes out of it.

I think it's gonna be fascinating to see Keep con in North America, Just that transition. Salt Lake cities. I'll be there.

Me too. Maybe we do this again. It's been fun, right?

Yes. I don't wanna say anything that's gonna get us in trouble, but yes, we'll be there. Yeah.

A lot of people like us will be there. Um, but I heard London next year for Europe. London Water City.

I can't wait. I I Water City. I haven't been to London since Covid either.

I was there a few months back. Were you still London? I lived on the outskirts there for a while.

Did you? It's, did you Well, you live in, well you live in the New York area now, so you're used to it. I New York, London is New York with history.

That was the older buildings. That's how I describe it. Older pubs.

I mine, I'm craving a fish and chips. Yeah. I can feel it.

I was at a restaurant last night. They had fish and chips on the menu. Not last night.

That's not the same. French don't do fish chips. I said, I'm not gonna order fish and chips in Paris.

Alright, before we get into a culinary thing, let's, let's end it here. com. Yes.

You write on there, you're blogging on there. Your reports are there at Steven Dickens three on every platform. Very cool.

Alright. Hey, that's, I think that wraps up our coverage here. Day two CubeCon, uh, here in Paris.

We're here tomorrow at, I guess around 10 or 10 30 Paris time. So if you're in Europe or Asia, tune in then. If not, you can catch the replay on Tech Drunk tv.

This is Allen Shimmel for Stephen Dickens. We're out.