End of Windows 10 Support, OpenAI’s Code Fixes, and AI Agents in the Channel | TSG Ep. 960

Hey, there's a lot of you buying PCs out there. Again, go figure you're watching Textron Gang. Hey, good, good morning everyone.

Well, it may not be morning, could be afternoon, evening. I don't know when you're watching this, but whenever it is, I hope it's a good part of the day or evening for you. It is Wednesday here on Techstrong Gang, and we are happy to have you join us.

Let me quickly introduce you to our panel today. We have, uh, Chris Blas, who is also drowning in his beers over the blue. Jace, Kate Scarsella, Mitch Ashley, and he's, he's looking up towards next year already.

Cole's throwing off the mound. I heard Mike Ard. Hey guys, Ladies, gentlemen, welcome to Wednesday, text on gangs.

Good to see you all. Good to see you. Good to be here.

Good morning. Good day. All right.

All right. All right. That was my way of doing a mic check.

I see. Um, I'm glad it's all working. So, Mike, as I led off with, it seems people are buying PCs again, go figure, uh, PC sales jump, and we've got a bunch of different potential reasons for it.

Um, I'm, I'm looking forward to getting into it, but let me kick it over to you. What do you think? Yeah, well, it seems like, you know, Microsoft is pulling support and people are finally hearing that at least.

I don't know how much of this is consumers versus corporate, but we'll jump into that in a minute. But there's a couple of interesting facts in here. So, PC sales are up.

It's not clear that people are buying those newfangled AI PCs, which are still relatively expensive. And if you look into the data points, the two big winners here seem to be, one is Lenovo, and then the second one, surprisingly, is Apple is benefiting very handsomely from the fact that people are revisiting what PC that they wanna get. And many of them are switching out from Windows to Apple, haven't forbid, but that seems to be occurring.

Um, so when you look at all this stuff, it seems like well, uh, we're seeing a, a big boost in these systems. I'm assuming small businesses are involved in this as well. And what I can't quite figure out is, you know, are corporations still holding on to Windows 10?

'cause they have all this software that's tied to it, or are they really gonna migrate to Windows 11? And we have a lot of, you know, if you've ever done a operating system migration, it's not easy in a corporate environment. So it's not a whole lot of fun.

But Alan, what's your take? So, I, I, you know, I think, as you've said, Mike, there's several factors of play here. One, obviously is the, you know, the sunset of Windows 10 support all those Windows XP and 95 users are gonna have to do something.

Um, but I, you know, I, I think the fact of the matter is, in the corporate world, it's huge. I think in the individual world, people don't necessarily run out and get a new Windows PC because Windows 10 support or Windows xp, as we've seen, unfortunately, it makes all kinds of security problems. But there is that normal circadian rhythm, if you will, of corporate PC upgrades.

And, you know, sunset gets dark and, and we buy new, new PCs. What I find interesting is how many people are still using their PCs versus tablets, phones, or other computing devices? How many people are using laptops versus desktops, right?

With no one, you know, everyone being hybrid, you probably need a more of a portable solution, though, I gotta tell you, I love coming into the office and working on my iMac, which is an all-in-one. And there are Windows versions of that all in one as well. I, I, I still prefer that over a laptop when I use my laptop in an office.

I still like to have external monitors, but, you know, I'm not your average user. I, I get that. So I, I think that's part of it.

The, the A IPC is attracting, I think that early adopter cool kid crowd who wants to say they're doing AI on their pc, of course the real problem is, what the hell are you gonna do AI wise on that AI pc there, there is no, there's no compelling software, there's no compelling killer app. I think FU has a, a decent report out on that you can go check out. But until there is, again, people like me will buy it just 'cause I can say I have an A IPC, but, you know, big deal.

What are you doing with it? Um, and then lastly, we're looking at new distribution models for pc. How many people are buying PCs in Costco?

I was in Costco the other day. I was blown away by the selection of PC of laptops. They have actually, they had all in ones too, but they've got a great selection there from the likely the people you mentioned Lenovo, apple, But others, As we noted in the, uh, lead into this story with Lisa in the show notes is, and Intel's got pop-up stores now that you can go pop in and buy whatever you want.

So that got me by surprise though, with everything Intel has on its plate, does it really need to get into popup stores? Didn't they try this before or was that the Microsoft store? That it was the Microsoft stores, but yeah, but Intel was like a big investor in or something.

Weren't they like MDF funds or something? What I thought funny about the Intel story was, is this what they did with the US investment? And there's only one store.

So basically in the us I mean, you know, so Well, you know, maybe because the US owns them, we could start selling them on military bases or something. Well, you know, Alan, Alan, this maybe Intel just doesn't like the way Dell and HP and Lenovo are marketing thinks. I don't know, Maybe You know, John Schwartz story.

Um, I think it's on, is it a techstrong it, if I remember right. Um, is is talking about, you know, is there a compelling reason in terms of AI for doing updates? That's a lot of what, um, alleviate Bouchard's, uh, report was about from rum.

It's a great guy. He really, really up on what's happening on the intelligent devices market. And you know, my experience is enterprises don't rush out and say, oh, there's a new chip.

Let's go buy new computers. Oh, there's a new app. Let's go buy new computers.

Most of the time they've got a planned refresh cycle. You know, we think these will last two to three years, three to four years they budget. 'cause this is a capital expenditure, right?

And so they've got a budget for that, and they have programs. And the, the next year is this a refresh year? And here's the three models you can choose from Dell or HP or Lenovo or Apple or whatever you're providing from maybe the, maybe the AI boom prompted some of that to happen a little bit earlier.

Um, I'm not really sure if that's true, but you have multiple factors. You have AI happening. Bigger thing I think is, you know, end of life or into support for Windows 10.

So you're getting pushed off of that and push you onto window 11 and just where spending is at, what co what companies are doing to conserve cash given, uh, economic conditions. Or is it time they want to, don't wanna miss the yay boat, so let's do our refresh, You know? And it, it, it's not just you, you know, to be clear, this is the Windows 10 farewell tour.

That's what this story is about, right? But it's embedded in everything else. You know, we all, you know, in our daily lives, in our daily professional lives and, and in this show, talk about it over time.

So you put it all together, right? So this is a bump for that reason. And as you say, Alan, there is no killer AI app, but you know, if you're a person, you know, you look at the demographic of buyers, yeah, there's geeks like us who will buy them because, and there's people a couple of quarters behind.

And as it flattens out, you know, we'll get, we will get AI apps and applications into next year and so forth. If you're a corporation organization, maybe now is the time to up some hardware for whatever reasons. But looking out into next year and the year beyond and the year beyond, when we're really wanting to use and capitalize, particularly that new hardware, how are we gonna be using it?

Right? And yeah, as you know, I have my own, we, we use the word mesh a lot right now. When you touched on this, like the last decade has been, which, what do you mean my computer?

It's like I have my laptop that I use a lot, and I got a desktop and I have 17 different, you know, devices, you know, and, and it's all in the cloud. Well, maybe, you know, uh, at home I actually need a certain amount of com compute. 9% of the time?

Or do do I need a household mesh of devices to do things? So anyways, whatever that killer AI app is, you know, there's lots of opportunity space in my world. It's that this hardware is being better used in complex environments at small risk scales in its hardware lifetime.

You know, even, you know, new, new, new equipment today in the 1, 2, 3 years, you really expect to get value out of it. You'll do different things with it. You know, Mike, this sits in contrast to that Wharton study we talked about yesterday, right?

Where executives, you know, the bloom hasn't fallen off the roads quite yet for the ai they see, they're real positive about it. Middle management, less so. But even though there may not be a killer app yet, doesn't seem like at least executives haven't pulled back the reins quite yet on ai.

Are we just waiting for some sort of security event to kick the remaining people who will not give up Windows 10 off of that platform because, well, there's no going to be no free updates. You gotta pay for those updates from here on out. But I feel like there's gonna be some sort of event in the next couple of months that's just gonna get everybody to go, all right, I finally give up.

I'm getting rid of this thing. You know, if that was the case, I think we would, we would've seen less Windows 95 in our critical infrastructure. So I don't think that's a, but, but let's go back to a simpler time when the world was a, a simpler place, but in some ways more fun.

Let's have a good old fashioned Mac versus PC discussion. You know, the knock on max is the backward compatibility of them, right? Apple didn't give a crap about backward compatibility where Microsoft did, but here we are with Microsoft saying, look, windows 10 and, and look, windows ten's long in the tooth, right?

Windows eleven's been on a while, but Windows ten's no longer supported. You gotta kind of upgrade that. Where here you go, talk to Mac users.

You are gonna have people using MacBooks, whether it's MacBook Air, MacBook Pro, MacBook, what just plain MacBook, but they're gonna have M fives, M fours, M three, M two, M1. You may even have some that are still running Intel processors. Now, the Intel processors can't, Mitch, what's the latest Mac Os?

Was it? Um, a Tahoe. A Tahoe?

The Intels, I don't believe run Tahoe. I don't think so, But they're all running Mac Os and from M1 on can all run Tahoe, right? Right.

They're all backward compatible. And so you have five generations of processors there that are all still viable and out in the market. You know, it's almost like they've switched roles between the Mac and the PC here, though.

Look, you have Windows that runs on Qualcomm Snapdragon though, that that has compatibility issues. My my son found out taking the bar exam this year. Uh, the, the Qualcomm Snapdragon didn't pass.

The wasn't, you couldn't use it. Um, you have it on Intel, you have it on Arm, you have it on on a MD, you know, on the X 86 s. But, you know, windows has become slightly more versatile on the amount of platforms, the amount of processors it plays on, but it, it, some of the backwards compatibility is, is waning.

So a real change back from the old days In, uh, windows on Arm, how that, who would've guessed, you know? Yeah. Well, there you go.

I'm, I'm gonna hold out for Windows 12, man, I don't know what it is, and I don't know when it's coming, but By God, I, I thought they weren doing Windows 12. I thought they were gonna do like a Mass Mac, OSX and then just have, you know, like that kind of thing. But Ma, I look, I, who knows, I, I think there's an, we're talking about an agentic operating system at one of their latest conferences.

So I think that's in the offering. So whatever that is, I wanna wait for that. I, I'll, I'll throw down on that one because I've been a, a Xenex, SEO xenex, you know, I'll take my Unix creds, you know, uh, back to the early days and, and we build everything on Linux now.

Everything right? You know, I, I, I don't wanna jump out here I am and say, you know, you know, there os wars are dead, but it used to, I, I keep thinking that Unix, you know, Linux, wait, When you second Chris, are you trying to tell me this is the year of Linux on the desktop? I don't know.

I I don't know if I'll go that far, right? You know, because we're talking Windows. This is the Windows 10 Fairwell tour.

Let's be respectful of the dead, right? Mm-hmm. But, but it's still there.

No people are, you say, no one's gonna stop using their existing computers for a long time. And I don't really care. The desktop's a harder thing, but I've used it on the desktop onboard a bunch, two Linux os you know, cap Server right now.

And it's so much easier for me now, even You dozen in other people. But anyway, my point is though, I think you put these things together again, you know, the vibe, coding, speed of, of, of AI today, much less. 12, 24, 36 months from now, maybe we can stop having to have some corporation print us an operating system that more or less does kind of what everybody mostly wants to do.

And the individual nodes on your network will just configure themselves to do what they need to. Yeah. Because that's the way we're Building 'em now.

And I am with, with Chris on this. I mean, you know, Ubuntu, and it really, you know, if we're not gonna do anything new with the, with the hardware tech, you know, technology that is coming, then it's, you know, what's the point? So you buntu for the win, right?

And spread things out. Make, I'm gonna end this late. No, I gotta end it right here, Mike.

You know what? We're gonna end we'll. Take a break.

2026, the year of Linux coming to your pc, it's an electric. Let's come a lot of promises. Let's come back and talk about, hey, if we could do that, why can't open AI solve all our security problems?

That's next here on Tech Drunk Gang. You've earned it. The spotlight, the responsibility, the weight of teams, companies, and entire industries fall on your shoulders, lives depend on your decisions, your home life included that work.

You are protected physically and digitally. Nothing gets through your team without a fight. But in a globally connected world, everyone sees you, including those who mean to cause you and your organization harm.

And now home your sanctuary attackers see an opportunity. Your digital front door is wide open. And what compromises your home can breach your boardroom.

Because the devil's greatest trick isn't targeting your workplace firewall. It's convincing you that your personal life isn't at risk. Black cloak, digital executive protection, defending the new attack surface, your personal life, Folks, we are back and we're talking about open AI has got this new tool they're calling Avar, and it promises to go in and identify, uh, vulnerability issues in your code and remediate it.

It sounds like the, you know, magic silver bullet that we've been looking for all these years. But Kate, what's your take on this? Is this possible and a, do I really wanna let the AI fix things without me checking on it first?

How's that gonna play out, Right? That's the question. Can we trust AI to secure the code?

It's now writing. So open AI's release of Aardvark and AI security agent designed to detect and remediate code vulnerabilities is more than a feature drop, a signals a pivot in how software will be secured going forward. I personally, as you know, um, think that we have to get rid of this whole, you know, I'm for deep bloating and, you know, getting rid of code that we're not using anymore, just turning it off and, you know, just getting rid of it.

So I think when I look at this, I think, um, specifically that this will help go towards that model of, uh, you know, uh, we have over 40,000, um, secure, um, coding vulnerabilities. We need something, we as humans are, are not able to do it. If we were, you know, we would be doing a heck of a better job than we are now.

So, so I think personally, from a cybersecurity point of view, that this is, this is good. I mean, we need something that will help us with speed and, and to detect vulnerabilities. And I'm for, you know, DevSecOps, ai, this is it, DevSecOps with an attitude.

And, and, and, and again, on this bloody show early this year, we start, you know, we we're laughing about it, what you gonna use? You know, ai aye yes, yes. Go Kate.

Yes. Right. You know, and there's a yes but to that, right?

You know, but, but you know, this is security. If I'm the adversary, and maybe I am, who knows, who's the adversary, right? I'm staring at you and using these tools to find the vulnerabilities, I will find all those vulnerabilities.

So you should probably look too now, Mike, right? You know, who's auditing the auditor? Who's auditing the auditor, right?

That's our world, right? You know, at tested systems where you're actually tracking things and verifying things, not even just as we do as humans, but we realize as we look at human structures, we kind of, you know, skip steps sometimes we're, you know, we're coding those things in so you can do them. And as you are, you know, again, again, just cvs, you know, this has come up a lot of times in the past.

Our idea of the last 30 years is if I can find all the vulnerabilities and I can see what people are doing against them, maybe before they do things to me, I can put those together. That's coming to an end anyways. But this is a good exercise in the, in the end era of that.

So do you want to spend human time looking for vulnerabilities? No. Automate the living crap out of them and do that in reasonable environments.

I would call 'em a tested environment so that you're not just wandering off into a AI hallucination. Um, put that in a, into a human process where the humans are focusing on the human, on the other side of the class and not on the AI vulnerability tools they're using. Mitch, people have been talking about this issue forever, and we have plenty of other tools out there that scan code and sometimes at least suggest fixes.

Maybe open AI is going a step further, but how does this stack up from what we already have out there? Well, The current, current generation of code fixing or finding vulnerabilities may be recommending fixes if they do that is vast and wide. I mean, everybody from Microsoft, AWS, Google Atlassian, name the company to people you've never heard of, you know, people that I haven't talked to yet, um, that are in this, in this area.

Art of Arc is another attempt at that. But pretty much, at least all that I've seen are essentially doing what we do today. They're just automating, using AI to go find vulnerabilities i's a little interesting 'cause it looks at the repositories.

Others happen in the IDE, the scan code when others happen at the point of check-in. So there's different points in time, but they all rely on a human to look at the results and say, yeah, uh, that's a problem. I'll go fix it.

Or Yeah, that's a problem, but your fix is not good enough. We have to go do it. Or I'll accept your fix.

And as we generate more and more code, um, especially as we generate code with AI that has more security vulnerabilities, that's, that's a problem. Um, there was an interesting study I eee study that, uh, did some analysis on, when you're using AI tools, tools to development, you just don't generate the code and then go do something else. You sit there and iterate it.

I iterate it with the, with the ai, with the model or agent that you're using that about the fourth iteration of it working on the same code is when you would start to see incremental vulnerabilities being introduced. So it's kind of build builds on itself and then starts making mistakes. So these are all interesting, I think.

Um, yes, I'm glad we're doing it. And yes, we'll have agents that will start to do some auto remediation, but I think until it, it fully agentic isn't gonna matter until we fix it at the point of origin, what I call the, the code secure code og, that's where we need to solve. It's before it ever hits the screen or the repository or anything else getting spit out of the code generation process.

Well, and that speaks to exactly the kind of problems we're having right now that I don't expect to persist a lot longer because, you know, drift, right? And I, and, and again, on in venues like this, we have folks like us talking about things that I think are very fu uh, future cells will assume they are, are obvious, right? If I have an LLM and it's generating code and putting it, its into its own context window and rerunning on that, you know, they're reading their context window as they go forward.

And I'm three or four iterations down a down a path, you know, drift is a thing. Why? Because you're, you're generating on your, your, your own code.

And this is not a necessarily a bad thing, you know, it's kind of how we do it. You know, I'm generating code and you guys are consuming it and so on and so forth. But that's, that's this brittle untested environment we talk about all the time, right?

You can't expect a person or an LLM or anyone to write things and then, and build on their own writings without, you know, cross matching, verifying, coming back to reality. And we literally, there's five of us on this screen, we're literally bouncing ideas back and forth to each other and then return to this LLM, this little model of a, of a human predicting words and say, sit in this little box by yourself and tell yourself stories until you can't tell where, where you are anymore. And we're we're surprised.

It, it's, it, it makes up fairytales. It's like, how about, you know, making sure it's tied to other things and not just in this little box. Anyways, rant.

So here's my think on it. I wanted to welcome all and AI into the quest for the holy grail. Yep.

No, you gotta get the, the coconut things that go like this. You remember? Yeah, yeah.

Sorry. Hey, is she a witch? Does she float?

No, but, but seriously, we've all been, you know, looking for this brass ring off the carousel to pull, uh, to, to combat co vulnerabilities and to do it as far left as possible. And Mitch, you want to do it at the, you know, at the, at the instant of the big bang. Maybe we'll get, we'll go that far back.

There you go. It's not gonna happen. Let's, let's look, you know what, there really is no Santa Claus either.

Hey, I'm not gonna touch the God. So you, Hey, we've been through this before. There is a Santa Claus, Virginia Cover your cover, your children's ears.

Absolutely Canham. Since we are talking about medieval stuff though, I just want to throw a a nod to Kate there and say, you know, I love that idea about the, you know, the code. I think if we're gonna be medieval, we should have, you know, like in the bad days when they had, you know, a plague and somebody would walk through town ringing the bell saying, bring out your dead.

I think there should be a bring out your dead code day. That's a brilliant idea. I love it.

Actually, Not dead dead yet, your mouth with the medieval and Kate thing, but you get nicer with it. Wait, I think that was a compliment. I'm not sure.

Well, I like that much, but, but Right. You know, I'm wanna go there for a second because a, Alan, you're right. You know, there was no, uh, um, um, fairy godmother, um, uh, you know, holy grail, however, right?

You know, Mitch, who, you know, what you bring up with, you know, more, uh, you know, creating more, more vulnerability with the AI code. The difference is that, you know, while there may not be a holy grail, there's a difference between nine and one, right? You know, there's, everything's on a scale, and we're used to the world right now where vulnerabilities are this level, this percentage of whatever.

We have metrics for them as humans generated. And we put in process around to try to make that a little bit better. And then we throw AI at it again, you know, random, you know, isolated AI that makes the problem worse.

We can't forget that there are, you know, I think we're sort of at an eight in this one and always have been. We've like eight, nine, sometimes seven in, in the, in vulnerability pervasiveness. I think we can get it down to one two, right?

Yeah. The, the volume of vulnerabilities. And cha you, again, I don't think chasing vulnerabilities has much legs left at all.

But I think the volume we can drive way down and you can't make it perfect. You absolutely cannot no matter what you do. But we're used to this level of slop and that that's, you know, we can machine those bearings much, much, much closer.

Totally agree, Chris. Yes. I, I don't disagree with that either.

Here, here's what I think's gonna happen. You know, it's an arms race. As we learn to generate more code with ai, we'll also learn to generate more secure code with ai.

And, you know, for those of us who have been in security a while, this, you know, this has been a holy grail. This has been, um, you know, uh, uh, a blood kind of thing that we've been pursuing for, for all these years. My question, let me bring it back to the beginning or the press.

You know, the premise of this is, is OpenAI the company that's gonna do that? Anybody believe that? Well, I don't believe that you should be using the LLM that you use to write the code to verify whether the code is secure.

Absolutely. And I don't believe we should be using the LLM company as my security company. I think, I think people who have a bit more security expertise and experience, though, for all I know Open AI went out, Hey, look, they probably lost about $11 billion last quarter.

Maybe they spent a billion of that or something on some security folk aqua hire or something. But I would imagine that there's security companies out there who are using ai, who have real deep security expertise that are working on this problem as well. That sounds like a great company Now, Excuse me, That sounds like a great company Yeah.

To start. Yeah. And I don't think Open AI wants to grow up and be a security company.

I think it's more of a, uh, do something in the near term to just address vulnerability so that they are doing something while everyone else is also doing that. I, I don't see them surging ahead and taking the lead on fixing insecure code. But I have a feeling everybody Who makes these tools is suddenly running over to philanthropic going, Hey buddy, how you doing?

Yeah, yeah. They're all gonna need one. Chris, I'll give you the last word then we gotta hang up here on this.

Sure. So, uh, you know, some hope for, for the future, shout out to two folks out there in the AI world. You got Mark Weatherford, uh, chief Policy Officer at Nvidia, key player, just, uh, just recently.

They're bringing a huge legacy of understanding into that hardware layer. And Miles Brundage, you know, not with OpenAI, but he was the policy guy out of Oxford into OpenAI and out in other ventures. And you can see the structure behind it.

Like, and again, I don't wanna pick on open AI as too, be easy, but should we trust any one L-L-M-L-M company? Anyone, anything? No, but they have roles to play.

So Opening Eye should be moving in this direction. If they get it right or wrong, I'm sure we will hold them accountable. Absolutely.

Let's take a break. We're gonna come back here and we got Mike Mike's down in the National Harbor this week at an Ingram event, and he's gonna give us, uh, an update. You're watching Textron Gang, Discover Textron Group, the epicenter of tech innovation.

We are your go-to for reaching, 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 Textron Group.

Hey folks, this is Alice said, I am in a hotel yet again, and this time we're in National Harbor for an event called Ingram Micro One. It's their annual conference with the channel. The channel is made up of all those folks who are resellers and system integrators and consultants that, you know, are the, are the fiber that kind of holds everything together out there.

And especially for SMBs, they rely heavily on these folks. And, um, Ingram most people don't realize is, you know, generating, you know, mo you know, close to billions and billions of dollars in revenue every year. And this is because they sell a lot of stuff and they're one of the big two in the IT distributor space this week.

They're talking about how they're using AI agents to start to automate some of that process. And the first thing they did was came up with a sales agent that enables their salespeople to look at, say, macro economic news and correlate that back to some of the historical sales data they have to recommend to some of their partners about maybe you should go talk to these companies in this particular vertical industry about buying this, that, and the other. Because, well, it's more relevant now than it was yesterday.

This is part of a whole process that's starting to occur where AI agents are gonna be both buyers and sellers. 'cause the people who buy stuff will be putting AI agents into their purchasing apps. And somewhere in the ether these things will somehow magically negotiate with each other and, um, we'll see how that comes.

And then as a side note, Daniel Newman from the Tuum group was here with Tiffany Bova. And, um, you know, Daniel was up on stage and said, you know, one of the things that I found interesting was that he said, Hey, you know, one of the most overhyped things for 2026 is this whole notion that AI is gonna replace the need for humans in workflows. Well, you know, as soon as I get off my backup on my chair, I kind of like said, Hey, um, that was pretty cool and pretty fair.

And then Tiffany had some great insights about how, you know, if you don't have your data house in order, you're not really gonna be able to kind of leverage all this stuff. So, you know, it was like, you know, rational thought was kind of occurring on both ends here. And I thought that was, uh, well just outstanding.

And as you kind of think about all this though, it says, you know, they also did a straw poll and just said, how many of the partners were ready for ai? And I think, you know, maybe 20 to 25% had their hands up as they're doing something and the rest, were still kind studying their options and what they think is gonna be happening. But Alan, um, how automated do you think that this whole purchasing cycle is gonna get?

You're asking Me? Mm-hmm. Well, you know, I, let me say this.

It's a, we're in an evolutionary, but a rapid evolutionary phase right now. I, I think we're gonna see tremendous automation, especially when you're going down channel sales like this, right? Like Ingram's main, you know, bread and butter business, it's gonna be automated to no end with multiple sales teams being able to, you know, point people into that, into the sale, into the buy process.

However, I think we are barely scratching the surface of an agentic AI automated sales future where I'm, you know, let's think about the, the effect it's gonna have on the channel, like in about Ingram's real business. If, if I'm going to buy through Ingram, there's, what's the advantage? Is it easier, cheaper, faster, better somehow, right?

And when we're talking in an nature of agent ai, if Ingram's agen AI isn't the chisel, I'm going direct because it's all automated anyway, Right? And to your point, on the purchasing side, we'll have AI agents that will be instantly comparing prices and availability in split seconds and making certain decisions. And the other part of it's that I don't quite get though, is like, so let's just say, and maybe Chris, you can help me with this.

So I'm a seller and I have an AI agent that I have programmed to optimally sell things at the highest margin or best margin possible. Now, you as the buyer or a purchaser, have an AI agent that's designed to buy things most efficiently and that the least cost possible. So won't these two agents just kind of meet somewhere in ether and cancel each other out and call us for help?

How's that gonna work? I, I think it's a perfect example of brittle systems, right? You can see the race condition right there, right?

And everybody, every buyer and every seller, you know, locks horns immediately and collapses. And this is, you know, so, so, uh, as we gave a demo for, uh, uh, built a demo for Gati ever and his GTFO, uh, pro series recently on how two ais can develop trust over time. And you can take that and put it in a lot, a lot of different scenarios, but, you know, hadn't discussed that through in a sales scenario, but it's the same thing, right?

You know, if I am the human who's setting up an AI system, I'm the buyer. I'm on my end to go out and do these things, and I encounter this brittle situation, you know, it just immediately crashes against, you know, as you're sort of setting up, Mike, what do I do next? Right?

I still need to buy things. What's the next step? Well, the next step is to replicate, you know, Ingram channel.

I love channels. My entire career has been about channel. I was like, Hey, would you like to be part of the, you wanna be the company that's selling my product and getting in, in early and making a lot of money and get that human to understand, see theirselves in the, in the story where they're rich and famous, and a lot of 'em were right?

And it's a relational thing. So I need AI systems that will develop and maintain a relationship that will track say, okay, I paid less, um, um, that one time and, but they're always a little bit late, or sometimes they're like 10% of the time, but however, depending on the condition, I've been asked to go out and buy this thing. And if I have enough understanding, like a human would more or less, you know, of the, of why I am buying these things, I can look at my sources and say, here's the tracker and here's the history, and n navigate that forward over time.

So I think it's really, it, it's because it is such a, we're not avoiding this one, right? This is happening. I'm telling you, I've started a company, we're not hiring a lot of salespeople to do that by hand.

And the relationship map of the people who might be customers is much more like the inside of my head, but now it's in computers. You know, we need companies to understand who they're buying and selling from, and not just because you're VP of sales or your three salespeople personally know these relationships. Hmm.

And and to your point, Chris, I think my cousin just came last weekend to visit me. And, and one is about to graduate with, um, a degree in advertising. And she was speaking to me and I, I was listening and I was fascinated at what she's learning.

And I thought about how it actually makes sense that we are overloaded with all these advertisements of, you know, and it's almost creepy to some extent of, of what you know, of what's being suggested. We say something and next thing you know, we have a bunch of ads, but in the class that she was taking, they started talking about brands and how, um, we are becoming as people, we, we want to be, um, look at brands and, and say, okay, this is a good brand and this is what I want. And I think the same is gonna work, um, where we're gonna be.

We are becoming smart consumers, and we want, we are so tired of being told, like, just by this, uh, for whatever reason, we wanna know, now I have, you know, this is what I need. Give me suggestions on how to either fix what I need, if I'm a do-it-yourselfer and I'm trying to fix something in, uh, you know, I'm in my house or, you know, from, from everything to, you know, the type of technology I want. I think, um, this is changing the game, and I think it's important, And I think it works out okay in the end, right?

You know, I've, I've been, you know, a, a Pollyanna for, you know, since before computers since the eighties, you know, and so forth that it's not all bad. And look at what, what I've seen in advertising and sales in my life that people try to adver literally spending money to advertise things to me that I'm physically incapable of buying, right? And much less so now as you're saying Kate, now we're in the world is, it's a little creepy how well, you know, what I would buy, right?

So our targeting has gotten that good to the point that it's bad, right? You know, you can't just creep up on your customers all the time. They'll stop buying from you.

But we're getting a resistant pressure to that. I think The future could hold more than just automation and, and buying and, you know, buying and buying at the lowest price. If, if it really, if it really was effective, we could be headed to a future that looks more like arbitrage, like the way stocks are sold.

I mean, you see this manually being done today with open AI getting, you know, this much bandwidth and compute reserving it today, they're capturing that because they need it. They don't wanna lose it to, to their competitors. I mean, I could imagine, you know, products, which, which does happen to a degree today, products being bought and sold, because right now there's a window where I can buy this as a channel person, as a reseller, and, you know, buy it for a dollar 50 less and in quantity make up, you know, make some good money on, on doing that, or dump something that's, that's not performing well.

So you maybe we're moving closer to an arbitrage kind of scenario. Let me tell you how that's gonna play out, because at the end of every fiscal quarter, every buyer's gonna wait till two seconds before midnight for the best deal possible. 'cause they know that that vendor's desperate to drive mark that revenue into that quarter.

That's what we do as humans. So why, why wouldn't the computer, You know, what I sort of wonder when I read this article is what does the sales kickoff with scope look like with all these ai a AI agents? Well, they're all in the President's Club.

Um, all right. Hey guys, I've gotta pull the plug on this one. We've gotta move on.

It's been a great gang today. Chris Gate, Mitch, Mike, thank you. Thank you for watching.

Stay tuned for the rest of Techstrong, uh, tv. Immediately following, um, I believe do, do we have a tech field day today? Yeah, Tech Field Day, which I am at, so I'm happy to, to be joining everybody there.

So watch the tech field. Very cool. It's actually Network Field day, but yes, network Field day.

Very cool. We will be back tomorrow though, with more gang and more fun and more good discussion. Until then, though, this is Alan Shemel for Textron.

We're out.