Techstrong Gang – October 17, 2024

Hey, everybody. I'm Mike Ard. Today we're gonna be talking about, well, Intel and a MD are strange bedfellows.

Lenovo seems to think AI needs a drink of water and BMCs at a crossroads. You're watching Textron. All right, folks, we're back in our guest today.

Of course, our start with our Mitch Ashley, who is not in Colorado for the, I don't know, for the fourth Street Week. I mean, do you have a home? Um, I don't Know, but I hear this is what sea level is like.

Interesting. I'm, it's, it's nice to experience this. All right.

And you're in the area today, right? Yeah, I'm in San Francisco Bay Area. All right.

And John Schwartz is also in the Bay Area, and I, and I got a nickel that says neither one of you is gonna meet each Other. Well, I'm not. I am, I am not in the Bay Area.

Mike, I am in Seattle. You were Seattle, okay. We're playing musical.

That's Right. You heard I was gonna be in the Bay Area. Yeah.

So I left town. I skipped town. I owe you some money.

So I'm out. And then we have Hope Lynch once again. Hope I think, you know, are you home or you're the only one home?

I'm at home. If, if, if we'd done this another day, uh, coming up, I would've been in Atlanta. But I am at home today.

All right. And, and I am outside of my normal New York environment. I'm here in Las Vegas for this BMC event, and we'll explain what that's all about in a minute.

But first, let's jump into this whole a MD and Intel have gotten together for a consortium that's gonna promote X 86 together. I mean, John, you were at this event, but my first reaction was, wow. Hey.

Hell did freeze over. Yeah. Pigs flew.

Who knew? Um, so it was pretty strange. So, so yeah, last week I did a briefing with Lenovo.

So this was, uh, at the Tech World's 2024 conference that they do every year now. And they told me that Pat Gelsinger and Lisa Sue are gonna be at this event. And I was like, no way.

I said, well, maybe they're gonna appear separately and there will be, they will be, uh, sandwiched with the other folks in the tech industry speaking. They said, no, they're gonna be talking back to back, and they're gonna be talking about something they agree on. And it's, it's this thing called the X 86 Ecosystem Advisory Group, which is basically a consortium that's designed to push forward the ecosystem around X 86 processors as we go into the AI age.

And so it's interesting. So Pat Gelsinger walks out and he says, Lisa and Pat finally agree on something that he mentions the group, and he says, who would've thunk it? And he proceeded to be very diplomatic.

And then she came out and did the same thing. And I think what they both wanted us to know, and they made it very clear, was how important it is in the development of this, of the processors as we go into the AI age, and how they want it to remain a bedrock of modern computing and kind of a defacto architecture. Uh, what was also interesting, everyone was that micro Microsoft, which rarely doesn't really talk so much about this technology, was part of this consortium.

While the founding members did not include Amazon and Nvidia. Oh, I just mentioned some of the members of this group. There's Broadcom, Dell, Google, H-P-E-H-P, Lenovo Meta, Oracle, IBM plus, Linus Tour Vaults, and, um, Tim Sweeney of Epic Games.

So it was, it was a very interesting event. And, and unbeknownst to me, and this wasn't to say about me as a reporter, I'm covering this, and meanwhile, the, the, the two, Lisa and PA go backstage and they end up being interviewed by our own Daniel Newman and Patrick Moorhead, who put up a video that I, I linked in the story. Uh, it is a very, very, very interesting, and, you know, it shows that they can't agree on something, but I think they're agreeing on something as kind of a defensive measure against perhaps ARM or, or others.

So, um, again, uh, something I didn't expect to see it at a Lenovo conference, to be quite honest. You know, it's, it's very, I think you can look at the AI chip side of it, but very clearly ARM has made huge in inroads, you know, kind of from the edge and into the data center. So that clearly has to be a big part of it.

Not surprisingly, it's, it's interesting to see what this consortium or what, what will come of it. You know, sometimes these things are, are mostly look and not much output, and sometimes they really do some interesting things. So I'm interested to see where this will go.

I'm wondering the what, when I looked at it, I was like, the fear factor here is huge. I mean, for these two to get together and do anything of this, like means that they are, I'm gonna use the word terrified. It's Qualcomm and Arm on one side and Nvidia on the other.

And these guys are, you know, walking around going, Hey, the 6 86 thing might just wind up being, you know, a nice legacy architecture that we finally remember at this point. Cobo go way, the, the, uh, way of the Motorola processor or something. Yeah.

Where, where's those, where are those, where are those sun risk processors when you need me? And I think I've got some of that talking. Yeah.

You know? Yeah. It's, it's interesting that Mitch said, you know, he's absolutely right dead on.

How many consortiums do we all remember are these initiatives that were announced that amounted to nothing? They were all ceremonial. The, the in, the difference between that and what's going on with Intel and a MD was, uh, very specific.

I've looked at press releases from both companies, very specific ideas of what they want to do. So they've been thinking about this. And Mike, you're right, they are probably terrified.

Mm-hmm. Hope, you know, from the software community, when you look at these little hardware parties, you know, does anybody in the software side notice care? Are they, or is it just kinda like something that happens, you know, deep down in the hardware that they're not really paying much attention?

Um, notice and do care wants, there are some standards and, and, and something you can develop against, right? So if you have a team that is developing for iOS, they would be interested if there was some announcement in this consortium that had something to do with Apple, but as far as X 86, uh, yeah. If there are new standards that, that they provide, I, I think there will be a lot of interest.

But one of the other things that does come to mind, and, and this bears on a conversation we had a few weeks ago, we were saying, you know, what's Intel gonna do? How are they gonna solve this problem? Is it, is it going to be, uh, a takeover or, or what other solution is going to come about?

And, uh, this is one option that did not come to mind for me, that they would, uh, do an enemy of my enemy as my friend tactic and, and, and try to find a way out. The part that leaves me scratching my head a little bit is both a MD and Intel have other processors besides X 86. And so now they're standing around touting X 86 is like a platform for running your AI inference engine, but are you got other processors that arguably are gonna run those inference engines better?

I mean, X 86 may be less costly, but you make a trade off between Yeah. The capabilities of the platform and the accuracy of the LLM when you use an X 86 processor. So it strikes me as kinda like, well, um, why did we need to tie this to ai?

Because ai, well, I think The, the, the intel, intel chip train, I mean, we all have been on this, this, uh, this transportation line for a number of years, right? Every generation of Intel that comes out, and they come out with faster and faster and faster. And the next generation comes out with a different, you know, footprint for the Intel, uh, chip line, uh, for, for, for processors.

And, you know, everything from low power to, you know, very high, high-end server kind of things. So I, I don't know, I, I'm not saying it's signaling an end to that, but I think it's signaling an end to the, that's the reliable path, right? We can't just stay on Intel, can't just stay on that path alone.

A MD has certainly become much more competitive and, uh, doing well in, in its technology that it's delivering. You've got Arm who's also, you know, now competing on the scene. You've got Nvidia and everybody who's making HHA AI chip or as well as Intel and everybody making their own.

So it, it's become, instead of Intel and everybody else, it's become everybody. And yes, Intel is still a big gorilla, but there's so is Nvidia, and I think it's just a, a, to me it's more of a sign of a receding of dominance in the market and less dominance by Intel and more, I don't wanna say equity, but more strength by multiple players. John, you know, Lenovo, you know, I was gonna tell you before Mitch brought up a good point, you know, um, before I forget, Lenovo was, was hedging its bet, by the way.

So later in the program, which is about a two hour keynote in the morning, they brought out a video of the Qualcomm, CEO Kristen, and then later, guess who appeared on stage? You only get one guess, guess who appeared at an AI event on stage as the headliner to announce a partnership with Lenovo. Yes.

Jensen, did He have a fancy leather jacket? Of course, the same one, right? Steve Jobs.

Steve Jobs would wear the turtleneck he wears the, wasn't Steve Jobs the black Jacket? Yeah. Steve Jobs with a jacket.

The tough guy. Look, but he came later. We'll talk about what they, In the age of ai, it could be Steve Jobs one of these days, so Oh, yeah, exactly.

True. Yeah, I'm waiting for that. Um, uh, but, uh, it's interesting.

So in a sense, yeah, so they they crossed all paths. Yeah, well, I, we'll talk about that in, in the next segment about what Nvidia and Lenovo did, which is kind of interesting. It Almost feels like a Game of Thrones episode.

It's like nobody dies, you know? It's just the seats are changing and, and a lot of things, you know, games within games, plans within plans. It's just really, it's really fascinating to me.

'cause the market has really changed substantially. Uh, That's That, yeah. That's interesting.

Yeah, it's interesting because Lisa Sue referenced that. They talked about how, how quickly things are moving and how the import of what they were doing, just, it's just in general with this consortium. And it, and it, in a sense, if you really read into it, it's a fear of the unknown.

She talked about short gain versus long vision. That was a theme of what she was discussing, and it's the fact that they just really don't know where this is going, and they're trying to hedge their bets. John, a couple episodes ago we were talking about, uh, Qualcomm and maybe launching a bid to acquire Intel, and we had Wall Street investors and financial services firms were talking about that.

And, you know, we were likening it to, uh, I'm sure you remember the book and the HBO series barbarians at the gate. Um, you know, was there any talk of that at this event on the sidelines where people going, Hey, you know, who's gonna be on? No, no.

Um, there wasn't, uh, uh, no, not really. Um, that's what happens. I guess I remember that story.

You know, these, these, these companies all talk to each other, right? In one form or another to some varying degrees of, of seriousness. That one had some traction for a bit of time.

I, I wonder if it was a trial balloon in a certain sense. Uh, but no. Oh, by the way, I forgot to mention Motorola also had an executive there actually, before the event started.

I didn't know who this lady was, but I was sitting there talking to her. She ended up presenting on stage. So they, again, speaking of, of companies, you know, in the, in this realm, they were, they were everywhere.

But, um, no, that, that, that rumor, I don't know. I, I don't know how much credence there is to it. I mean, I'm sure there was discussion, but I, I know I I would bet on that Qualcomm intel, nah.

And did they mention anything from this? Um, you know, the United States government, um, shall we say, uh, offering perhaps, uh, helping hand, I'm trying to avoid the, using the word bailout, but No, but it is what it is. Uh, hand Yeah, yeah.

The, the, the Pat Handout handout tour, um, no, they did not mention that. Uh, his, he kept his, I gotta be honest, they had so many tech, they had Adela, Zuckerberg, those are both by video. They're obviously, they were obviously taped beforehand.

They had almost, they had the FIFA president there for god's sake. They had so many moving parts that everything was at the most superficial level. They had the, the major talking points, and they shuffled them on and off stage.

Uh, but the, yeah, the government thing is, is a, is a kind of a point of, um, sensitivity for Intel, right? Given where they used to be and where they are now. So, So It was, it was, was Do there, you know, I mean, was, Okay, so yeah, this was, this was the big reveal.

The big reveal. So Apple has their one more thing. They literally ripped off that idea.

The, uh, yy, who's the CEO of, of, um, Lenovo said, we have one more thing we can't tell you about. It's state secret. And it was the FIFA president, uh, Mr.

Infantino announced that they are the technology partner for fifa. I know they have a partnership on with F1, and they've worked with Lotus, this is Lenovo. They're going to be, uh, the, the official tech sponsor or tech partner for the Men and Women's World Cups in 2026 and 2027.

I asked that executive later, you know, if you could tell me, well, what does that mean? Is it you're gonna enhance the fans' experience? Or, well, what are you gonna do?

And then he, he kind of shut me down because I think they don't know what they're gonna do yet. They're, they're gonna formulate ideas of what they might Yeah, well, they're very superimposed. Should be over By then, right?

Yeah, yeah. Well, let, let's, so let me, let me take back what I said. Let gi, given the F1 relationship, Lenovo has a very long standing, um, relationship and demonstrated practice of capturing telemetry.

That's what's happening in racing. Um, now I'm not saying yes, you know, maybe we're putting chips in shoes or, you know, armbands with chips in them or something with, you know, on FIFA players. I, I don't know that that's true.

But there, there could be some very interesting statistical analysis, AI applications, things that they could do to really enhance the game, um, either for the viewers or possibly, you know, for coaching and playing too. So there, there could be some really interesting things that come out of that. Yeah, Well, the stuff, The, that, Yeah.

I'm gonna end this conversation, but I'm looking forward to the eu, what it has to say about collecting data off of soccer bubbles. I'm sure I'll be interesting. Oh God, Ation next year.

EU AI feedback. You'll all be fine. There you go.

I, unless my team wins. Alright, folks, we're gonna be, we're gonna be back in a minute to talk more about Lenovo. Accelerate your DevOps journey at DevOps experience 2024.

Join us for an exhilarating experience that will transform your DevOps practices. Explore emerging trends in DevOps and advanced platform engineering techniques. Learn how to integrate security while leveraging AI to drive efficiency and innovation.

Mark your calendars for October 16th, 2024. Discover how to improve operational velocity and foster a culture of continuous improvement, ready to elevate your DevOps strategy. See you there at DevOps experience 2024.

All right, folks, we're back here. We're continuing to talk about Lenovo, which in addition to hanging out with Intel, was also getting warm and fuzzy with Nvidia, as is everybody else these days. And John, you were there, but apparently they're talking about water cooled systems to run these AI things that we've got going.

Yeah, they, so they, uh, they, they, they saved the best for last Jensen came on late Tuesday afternoon, and, uh, they, they talked about, um, expanding an existing partnership with something called Lenovo Hybrid AI Advantage with nvidia. And they introduced something called the Think System, SC seven, is it 7 7 7 v four Neptune. So it's a system that leverages, um, uh, Lenovo's Neptune product to bring, uh, Blackwell and NVIDIA's GB to hundreds of the enterprise markets.

So anyway, what they're trying to do, um, is let organizations build and run accelerated computing for generative AI while reducing the data center power consumption by up to 40%. So it's a, basically a vertical liquid cooling system that maximizes data center power for AI and research. That's, that's the bottom line of what they're doing.

There are a lot of verbiage, but in a sense, it's a collaboration. And, um, it was interesting. It was, to me, it was, it was kind of a change of pace from what Nvidia usually does, which is, um, you know, it's kind of this highfaluting partnership with the ServiceNow or Adobe or whomever, Salesforce, I'm losing track.

Uh, Jensen came out, he didn't talk so much about this initiative, but he talked about the importance of the AI agents. And again, he, he started talking about this idea of unleashing billions of them. He calls 'em Little Jensen, right?

And they would basically be glorified workers or coworkers and environments in, in, in raise efficiency. But, um, one of the, one of the takeaways from the Lenovo event was two things. A star studded array of CEOs to, to Pat Lenovo on the back.

Plus they did, they did make a big play about sustainability. Um, and I was gonna mention this to Bonnie. Uh, this, this was a very hardcore theme throughout the day.

So, again, Jensen shows up, they do this announcement, they present him with some hardware. He genu flex in the glory of Nvidia being the king maker in ai, which it continues to be. And, um, this was the highlights of a day that was devoted to kind of a review, a refresh of the Lenovo product line with the, the, the Greek chorus of support from Zuckerberg via llama, uh, Microsoft via Azure, uh, Qualcomm, et cetera.

Um, it was very interesting. It was quite a stark contrast from last week in the Tesla announcement, which was one of the weirdest, uh, I I would almost, it was almost kind of like a bad LSD trip, right? This long 40 minute video then followed by absolutely very little superficial data, or a numbers here with Le Lenovo is just the opposite.

Every tech company imaginable is involved with, uh, specific numbers, specific statistics, specific topic talking points. So it's quite a, it was quite a contrast comparing the two. Mitch, I'm gonna go with that theme that John brought up.

'cause I am having a flashback here. I'm back in the day when we used to visit data centers, right? They would give us the tour and they were all, you know, so proud of the fact that they had these water cool capabilities and there was pipes everywhere.

And then we spent about the next 10 years ripping out those pipes. 'cause we had finally made the chips cool enough that we could actually run them without all this water, and we could actually just get out, run them in a lower cost cage. So, um, here we are talking about water cool systems again.

Are we coming full circle on this? Well, I think it was once the hearing loss kicked in from those high speed, very loud bands in those data centers, I mean, little, you can't decide. We had to be this close to each other, to talk to each other.

Um, I it's may, maybe it's a pendulum. I, you know, it's, I think it's very much representative of how much power, um, how much GE has generated from GPUs. And as we continue to accelerate what they can do, you can't air cool 'em, it's just, you know, there is a, you know, a ceiling of which that goes up against which, so, so if I kind of turn it on its head, let me look at the software side of this.

To me, every announcement we hear from Nvidia partnering with somebody else, we talk about the hardware side. The part we don't talk about is Cuda. It's another place where Cuda is the software framework, right?

That's use for AI applications, strengthens its whole, or strengthens gives us another foothold. Strengthens a foothold it already has in the software world of interfacing to ai. And we talk about Nvidia and its strength and hardware in the end, you know, others may catch up in, in the hardware side of it or become more competitive with Nvidia.

The software is hard to displace. You don't rip that out like you do the next generation servers. We're gonna go with somebody else's chip, right?

Well, no, I built all these apps on Cuda and, uh, that's gonna stay. So we're gonna stick with Nvidia. So I think there's some real, you know, you talk about short term and long term, John, kind of what the, the horizons of these things are.

I, I think it's a huge win for Cuda. Um, and, uh, you know, CUDA is a, as barracuda, if that's what it's named after as a waterborne fish. So maybe that's why we're seeing all this water cooled stuff, Mike.

I don't know. I don't know. But hope, um, you know, in my experience, every time the processors get too hot and we can't afford to run the systems, it's not too long before we start blaming the developers and the software people for writing crappy, inefficient coat.

Yes. So I knew we were gonna get to that. You helped yourself, Mike, Unfortunately.

But, but I, I think I, I, I think you have, um, there, you know, there, there are boundaries, you know, that, that you can work in, uh, as a developer, the code you write, um, and, and so much of what your boundary is, is based on, on, on the hardware you have. There have been breakthroughs over the years where developers have been giddy because instead of now the system, uh, being overclocked because of, uh, just how hard they're having to make it run, um, it, it can handle it. But if this is now saying, um, developers can get less, say, higher performance per wat right out of this.

So, um, they can ha harness more of the parallel processing power of the GPUs. This gives more efficiency for some of the large scale computations than you would get on traditional CPUs. Um, this can reduce overall energy consumption, but also, yes, it, it can help developers, um, when they're trying to build higher performance, uh, systems doing higher performance computations.

This, this definitely can be a win. Will it solve the problem? No, because, uh, developers are gonna still continue to scale up.

Once they figure out what the edge is. They're, they're still gonna push it to the edge. But I think this does give some head, Well hope, you know, every, every time a developer adds a new API call in to the Cuda framework.

Another developer gets their wings. Yes, I believe it. Absolutely.

What's up? Uh, One thing we're coming up on the holiday season. I thought I'd throw that in there, Mike.

One, One question I have, or, or I'm looking for opinions here. One of the things that also I thought about, um, at least in North Carolina, there are a lot of data centers being built, a lot of data centers that already exist. Is this going to be a way that, I don't know, some of, some of these organizations that are building these big data centers, they can come in and say, now, hey, you have these concerns about water usage and power consumption and all of this.

Look at this advancement we have, right? That's gonna let us do more with less in this space with this footprint. The one thing I am looking for, and, and maybe I missed it, is what, what are the numbers?

What is the percentage of reduction, right? In power consumption? Is there a reduction in water use, or are they gonna use more water?

Those questions are gonna be really important in the communities where these data centers are, are built and exist. They're gonna use more, and they're gonna, it's not like they're gonna use lut. They're just gonna say, Hey, now that we got these water cold systems, we can use even more stuff.

And then they're gonna wrap themselves in that sustainability flag that John was talking about and say what great citizens they are. Oh, yeah, John, I mean, you, they were talking sustainability. I mean, what did you hear that at the event?

Well, um, they were throwing out a lot of, a lot of percentages. I think what, one of the things, one of the themes I think, and it just doesn't apply just to Lenovo, but I think they almost feel a responsibility to throw out safeguards, guardrails. These are words that I consistently hear.

Sustainability. Um, we will adhere to regulations, we will, we will gain your trust. These are almost things that they're almost thinking, uh, kind of six to nine months beyond the announcements.

And once the product is out there, they wanna, in a, in a sense, kind of safeguard themselves against backlash, right? So sustainability is something that is, it's like a, in Silicon Valley, it's, it's something that's being overused, I think as a talking point to show that you're a good corporate citizen. Because I think a lot of these companies are a little bit on their heels, a little bit touchy, uh, around their relationship with the government, the, just this trust around ai, which continues now.

So I think they're checking a box. I I, I don't mean to be too cynical, but I do really do think they're checking a box. I do think they also, I mean, Lenovo claims and I, I don't know the history of Lenovo to either agree with them or disagree that they've always thought about sustainability.

I mean, this is the sixth generation of the Neptune product. So I, I'm gonna, I'm gonna give them the benefit of the doubts. Maybe that's part of their DNA.

But, um, I, my suspicious gene kind of jumps my, the spidey sense goes off a little bit when I, when I hear these things because they're almost, you know, part of the script in a lot of these events, Mike, as you can probably vouch over the years, they, they all tend to follow a certain type of, a type of script in a certain type of talking points. So that's kind of my cynical take. Sorry to That's cool to feel that way, but I really do, This is an imperfect comparison, but we talk about net carbon neutral, right?

We take carbon out of the, and we expend it, right? We almost cut. We need something similar to that to say we have reduced power consumption.

We have reduced, you know, water consumption, whatever it might be this much while we're gain, while we're using more to do more power, but we're using it at a less increasing rate. You need to, you need to know that differential, right? It isn't that just that it uses 20% or less power while compared to what, right?

Right. We've increased the overall Power consumption just hard. Yeah.

Yeah. All right. So I, I have two cynical thoughts on this.

All. We might be surprised, I Knew You did, but, but here's my first thing, right? If there's serious about it, I would like to see a response back from the occasional prompt that says, based on the amount of carbon that you're about to kick off on this puppy, we're not gonna run it.

'cause it's just not worth the trouble, right? We have no, like, metrics in here to say this thing is, you know, outta control. Secondarily, it would be nice, maybe if we had a pricing mechanism that said, Hey, based on the amount of carbon that you're about to kick off for this prop, you are gonna pay more for the privilege.

Right? And maybe we're gonna use that money to upset some of the costs that we are, uh, incurring as we, uh, destroy the landscape in North Carolina. Follow me.

Sorry. Hope, You know, it's, it's the price. Okay.

I guess. But yeah. How do you respond to that?

I don't know. Yeah, it's there. You know, I was gonna, I, I was gonna mention that years ago I went to, there was a, at and d had this massive data center somewhere in Virginia.

I remember I had to literally had to keep my eyes closed for a while while they were driving me there. It was surrounded by a moat. It had armed guards.

They, one of their prize possessions that they wanted to show me was the water tower in the back. And I think about that as the days of those might be going away. Um, I don't know.

I I, it, it'll take, it, it, I don't think this, this announcements is, is as impressive as it may be to some people, is gonna have that significant and impact on the data centers as a, there's a, there's, there's a rush of building them ice. I used to do data center tours in my first job. I love college with EDS.

And besides the big control room, you know, the, be able to open the curtain, that was the big wow. The second biggest impressive thing. The tank barrier at the entrance to the, that raised stuff outta the ground.

Everybody envision like, wow, I'd hate to run into that. You just never know what's gonna impress us. Was this before or after you oohed and awe over all the racks observers.

I mean, where was that in that, That was in the IBM days. They're all enclosed. Cabinets.

You couldn't see anything. It was just a big cabinets. They could have Been closed, stored in there.

You don't know. They used to have labels from all the vendors that were in there. Like they were badies of honor.

They did? Yes. Now I just think they all, I think they just get 'em all now from like super micro and there's no labels anywhere.

So Yeah. Blinky lights, blinky lights, Blinky light. All right.

Well, we're gonna end this chat about Lenovo and water cold systems and AI and sustainability because, well, it's bigger than all of us. So hopefully we'll figure out something to do about this soon. But, uh, Come true Right now.

There's a data center coming near you. Stay tuned. All right folks.

And now we're gonna talk about what's going on with BMC there, here in Las Vegas for their BMC Connect event this week. And, well, it's kind of a tale of two cities. If you haven't been paying attention.

BMC was announced that they're gonna separate into two independent companies. One is called BMC Helix, which is their ITSM platform. And the other is, you know, what we might call classic BMC, where it's basically the automation frameworks and the AI engines that get wrapped around that.

Now, it's interesting 'cause one of the things they were talking about here in terms of Helix, is that they're adding AI agents to the platform. And the, at least the theory is that these AI agents will take on more tasks and they have reasoning capabilities. Now, uh, this is a kind of a separate conversation, but we'll jump into this.

'cause I know John has done some work in this, uh, AI agents and orchestration conversation. And one of the things that I'm hearing from folks is that the, so-called reasoning engines that are in these LLMs are roughly equivalent to, you know, a five-year-old child that burns his fingers, lows that the stove is hot. And we are just over hyping the crap out of these reasoning engines.

And one of the things that BMC is saying is you're gonna need these orchestration frameworks to actually tell these various agents, along with all the humans involved, what to do when. So, um, what is your sense of what's going on with these reasoning engines that everybody's touting and agentic ai? I mean, is this, you know, the second wave of a hype cycle Sounds like it to me.

Um, so was there a story? Wasn't there like a story or a research that was done? Uh, uh, I don't have it in front of me.

It was just, it just happened. I just wrote about it. It a blur after a while, Mike, there's somebody, a agentic related type announcements.

There was this swarm. We, we, we, we don't have to talk about swarm from open ai, but that whole concept of things working in coordination with one another. Yeah.

Um, but then there's result, there's lab results or tests that show that these agents aren't nearly as intelligent as we thought they were. So you juxtapose that versus the hype cycle where we have Benioff and Jensen w at Dreamforce telling us that as we arrive at work, these agents will be doing all the drudgery that we dread doing. We first get into work like all the, the, the ditch digging, you know, so to speak.

They're gonna take care of all that. I, that, that seems to me like very optimistic when they, they say, oh, it's right around the corner. Well, what do you define as the corner?

Is that five years, three years? Um, and so the more I think about it, I, I, I wonder, wonder what hope thinks about this? Should we be as afraid or fearful of these age ages genic AI taking our jobs?

Or is this just kind of part of, of a hype, uh, utopian view that is, that is far away and is not gonna happen? I feel I, I agree with you. There is a lot of hype right now because it's, it's in the realm of imagination, right?

Everyone imagines and, and, and thinks and dreams about all of the things that can happen before. Um, people on a larger scale, companies on a larger scale have actually tried these things and put them into practice. And one of my amazing cress skin predictions I will make right now is, it won't be as great as some people may believe.

Because imagine you're in an organization and you have all of these AI agents deployed and they are taking care of a lot of the lower level tasks. That is great. There are some things that should be automated away, but if it is some of the thinking tasks that you would do in your role, now you are, to me, building even bigger silos.

You're creating bigger knowledge gaps, you're increasing the opportunity to build tribal knowledge because there are so many things in your role, you will not see touch you, don't, you, you, you don't know how to do and you don't understand because the agent has always been there, you know, just taking care of it, you know, look out for the day that the agents break and you're just sitting there trying to figure out how to get your work done. But I think there's a lot of hype right now. I think there are useful, useful places for it.

Like, you know, repetitive documentation, um, repetitive intake types of tasks. But I think, I think some of the others, we should, we should wait a little bit longer before we jump aboard and, and just try to throw everything at agents. Yeah.

They had this, this weird hypothetical situation at the Lenovo conference near the end. They have a, uh, futurist Ali Miller who, um, is, has quite a following Mm-Hmm. Sheet.

He was talking with the CTO of Lenovo about, uh, car buying agents in a scenario where your, your agent, uh, knows the type of car you want, all the, the a additions you want, your personal financial situations. So how you would pay for it. And this is all gonna happen very conveniently without you in involved at all.

And I was just, I was thinking, come on, this, this is, this is like pie in the sky stuff. Um, I, I'm, I'm on the other side of this. I think, you know, we're gonna have all these agents running around doing stuff and they're gonna drive us crazy.

'cause they're gonna be like, you know, they're gonna be overly helpful all the time. And they're gonna be like, can you do this for you? And you're gonna go, no.

And then like two minutes later, you're gonna be like, put that down. And then about half an hour in, you're gonna, you do that. I will smack you upside in your head.

Yeah. I love it. Hey, can I, can I mention something?

Totally off topic. I have two fun facts about the amazing Kreskin. Oh yes.

Number one, he was on this tonight. You've gone to tonight's show more than any other guest in the sh in the history of the show, evidently. Really?

Yes. Number two, I went to a tech party, a Christmas tech party years ago with, I took my daughter, who is, who's a reporter, he was performing. He, no one knew who he was, number one.

Number two, he bombed on everything. He, he screwed up everything. And later I went, that makes it even funnier.

And it was so bad. It was so embarrassing. And it was just, it, it, I'm watching it.

And it was, it was just so heartbreaking to me in a certain sense. And I woke up to him later and I said, I'm, I'm sorry, what happened? He goes, you know, that's showbiz.

Anyway, that's something that was an insight to life to me, right? And I thought that was so great. I i it of even more.

Sorry, I had to share that. Oh no, that's, that's great. Great though.

Now I have visions of Johnny Carson and the amazing carnet going through my mind too. So thanks for Oh, Mike. But you know what, Mike, it's also that the whole, the whole thing with Yeah, you're right.

It's gonna be, uh, when you talk about swarm, swarm is a good word to use. 'cause these things are gonna be pinging you from every conceivable angle. It's like the over zealous GGPS system that says, oh, there's a, there's a matrix.

A little bit of a backup there. Why don't you go off the side of the road for a while I've done that, and then I turned it off. It's like, you know, do this now, do this.

I have another suggestion here. This will help you. It's like, please stop it.

Just leave me alone. Mm-Hmm. And that, that's another, another consequence.

Right? And, you know, and I love the, what John's talking about is open AI has this, uh, swarm project, which is gonna be this orchestration engine for these agents that they're talking about. But it doesn't seem to me they have a lot of confidence in the thing.

'cause the first thing they said is, A, it's experimental, and b, there's No product. There's no, yeah. It's not a product.

Instead of the blog thing. Yeah. Basically.

Here it is. But, but don't hold us to anything. Just, just, just, it's just a concept.

Don't worry about it. I think what they're saying is, this is something we built. We're not sure if we can make it work.

And we're hoping somebody else can figure this out, but here's where we are. Let's, let's, let's put it out there and, and see what, what happens. But, uh, yeah.

Can I ask you, Mike, about, uh, bm? Was it, they, they talked about generative ai, uh, in the mainframe? What's that?

Yeah. What was that about? This is a follow up.

They were in talking about this as a statement of direction for a few months now, and they are now have an actual AI assistant. And, you know, it's called Amy actually had to explain to them that, uh, you know, there was actually a song called Amy and, you know, basically had to refrain, Hey, Amy, what you wanna do? But, you know, they didn't, they were totally like way over their heads, but I think they're all looking it up now.

Um, but the, the point here is It Explains the code on the mainframe, which is an issue because, you know, large numbers of main framers are now sitting on a beach drinking mi ties and being semi-retired. At least that's my vision for them. I hope that's how it all worked out.

Um, and there's all this code that nobody documented back in the day. And now we need to use AI to go sort this thing and say, Hey, um, can we fix this code? Can we modernize it?

Do we need to rewrite it? And I think AI will play a huge role in that. So B C's on the right track for that.

And now they've come up with an assistant that helps you sort through that. And they too will then probably extend that to include agents for mainframes and maybe they'll do the actual work of rewriting that code. We'll see.

Um, that's, you know, clearly IBM and everybody else is kind of talking about the same path, but it's interesting. I've seen and hope, I don't know, maybe, uh, I've seen a pullback in the language used for these tools, right? It was kinda like, we're gonna convert co ball to Java using this.

And now it's kinda like, Nope, we're gonna help you upgrade from Java eight to Java 17. And that's about as far as we're getting at the moment. So, you know, have we kinda like pulled in our horns a little bit on what these code explanation writing tools can actually do for us?

I think so because it's been in more hands and people have actually seen not only what it does, but what it does well and, and not so well. Right? So I think it's tempered the expectations and just the, the use cases that people are applying it to.

Uh, if, if we think about though, when, when you're talking about the mainframes, um, my extent tingled a little bit because I'm thinking, you know, there's, there's gotta be a solution. But a lot of, a lot of the systems that are still using mainframes today are the heavy hitting systems that are running the financial backbone, at least in the United States, right? I know the IRS still has a lot of mainframe systems.

There are other companies that still have a lot of mainframe systems. So if you, you bring agents in and now the agents are are, you know, making updates, they're working autonomously. Um, yeah, I'm gonna say, you know, my brain starts to think about, you know, the bad actors are already there, but, but wow.

If they can put these agents to work for them, and by the time we figure it out, you know, we might be a long way down the road. Yeah. Who exactly has control of what agent's gonna be an issue, right?

Because all I gotta do is fish your credentials and then I gotta control of your agents and party's not. Yes. So yeah, that will definitely be a interesting, um, story to report about another than nine months.

Or, or I think, I think you're right. I think you're right there. You kind of, what, you kind of wonder if the, if these companies are, uh, like, uh, uh, anticipating these things happening, right?

So they put in like little fine prints when they talk about security and they're about trust with their products, right? Like they're anticipating there will be a problem down the line. Mm-Hmm.

So they're trying to get a little bit ahead of it. And I always get that feeling at these things like, it's too good to be true. So, but you need to read the fine prints, you know, if you're patient enough, nobody is.

And, um, that's, that's the other thing that always lends a little bit of suspicion to me at these events. Um, they just, they're just presented as like so clean and, and, and so easy. And it's just, there's, it's not, that's not the case.

It is not, It is not. A lot of effort was made to ensure that the mainframe is still hanging out with the cold kids and, you know, those, it, it's part of the mainstream platform. And honestly, you know, I hear all the vendors, you know, doing this kind of thing and I scratch my head sometimes.

I'm like going, you know, I think you guys are the only ones who think the mainframe isn't cold. 'cause everybody else keeps using it and they're work doing work on it every day. And you're only walking around going, oh my God, is this thing, you know, not happen?

And it's like, no, actually the platform's plenty it, everybody hanging around it maybe not so much. We'll see. Yeah.

Mm-Hmm. All right. Well folks, that's where we're gonna end it.

Uh, it's been an interesting week so far and we got more to come tomorrow. So by all means, stay tuned and we have an awesome lineup of shows right behind us. Hope John, thanks for being on the show.

Mm-Hmm. Thank you. And, and, and if you hadn't noticed Mitch had to leave early, but hey, we'll be back tomorrow.

Thanks for watch Discover Textron Group, the epicenter of tech innovation. We are your go-to for reaching IT leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us.

You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more. Join our satisfied clients. Let's revolutionize your tech journey.

Contact us today and tell your story to the world in the most powerful way with Techron Group.