Techstrong TV June 25, 2025

Hey, everyone. Are you ready for the A to a era you're watching? Textron Gang.

Hey, everyone. Happy Wednesday. Welcome to our Wednesday edition of Techstrong Gang.

Um, I'm Alan Shimo. Hope you're ready for a, a, a quick 45 minutes of discussing three hot topics in tech today. As I'm doing going forward, I'm not gonna waste a lot of time introducing people you've been meeting every day for the last year and a half or two.

So let me just say that our gang today is Mitch Ashley, Chris Bla, and the dean, Mike Azar, coming at you from Denver as well for the open source Summit. Let's jump right into it. Mike, big announcement out in Denver at this open source summit from the, uh, LF Linux Foundation, Google, uh, donated their a to a protocol to a new, uh, project that LF is fostering or hosting or running.

That includes Google, Microsoft, and, uh, a bunch of other big, big names in the area in the space. And with the, with the idea of, you know, how do, how do our agents gonna talk to agents? How are they gonna interact?

Like APIs weren't enough, but what you are out there, you wrote an article on Textron AI on it. What, what, what do you got? Well, it was kind of interesting to watch the keynote because a lot of the people in the audience were still at the, what's a two A again, fans, uh, like, yeah, all right, well, this sounds interesting, but not many of 'em actually had built in an agent must much less try to interoperate with anything.

So they were kind of, you know, it was met with what I had like to call golf applause, which was kind of like, you know, this is good thing, but we don't know what it means, kind of thing. Um, I think it's gonna be a significant deal, and it has a lot of implications for MCP as well because, um, what you did see is a hundred vendors are now rallying behind this a to a protocol, and that's a good thing, and that creates a level of standard and interoperability we're gonna need. There's this other thing called the model context protocol that Anthropic created that is still kinda led by them, and they haven't quite gotten around to donating that to a consortium.

And the question then becomes is they're gonna be pressure on them to do the same. Uh, I think maybe, but will it go to the Linux Foundation or somebody else? I don't know.

But Mitch, this whole area now has, uh, an alphabet soup of acronyms of things, and I know you put a post together kind of outlining the different pieces of that. So maybe you could explain the relationship between A to A MCP and everything else that we're playing with here. Yeah, this is, uh, a report I did on, they'll update it regularly.

Matter of fact, right after I published it, this, this happened about the Linux Foundation talking about, you know, where are the top agentic AI open protocols? Meaning they're not going through a standards body in a formal sense, like IEEE or IETF, something like that. So the two people, if you know about him, the two you'll know about, certainly his MCP, that got the most kind of buzz because it's a different kind of project and I'll talk about about that.

But A two A was introduced by Google, largely launched at their Google Cloud next, and it had some support at the time. Um, and they addressed two, two different things, but they're very similar protocols. Uh, first of all, they're based on kind of a restful type of interface, something that we're very familiar working with in software development.

But MCP model context protocol is about getting access to things outside of an l lm, how do I get access to a database or to a service, or to an application, or to something that will enhance the reasoning of the m to IT agent? Agent to agent. And by the way, agent to agent doesn't mean we're, we're splitting alcohol a anonymous into half.

It's, that's a different thing. A two, a H, two A is really about agents talking to other agents and discovery of what, what the services or what capabilities of the agents are. And this, and keep in mind this is a, this is still evolving.

It's not a mature thing yet. Uh, the authentication protocol for that, there's still some work to do to beef that up. What is the modality of it?

Is this a text, a image, a voice video? What, what are we talking between each other? And then an ability to interact between agents, but still protect the state.

In other words, what's happening inside my agent? You don't need to know about, I'm a black box. You just tell me what you do and I'll talk to you and I'll do the same for you, blah, blah, blah, blah.

There's still more around orchestration layers and things like that for this all to evolve in. Personally, I'm looking at all of this landscape of, of open protocols and saying, okay, where's, where's the pony in there, where's the Kubernetes of agents gonna come from? Is that gonna be an evolution of a two A probably more likely that than MCP?

But, you know, that's sort of making, making the sausage kind of detail. But the, the big thing is externally, it probably a much bigger impact than it did necessarily at the, uh, open source summit. It really sent some shockwaves to say, great, glad that Google did this.

And there's one reason why this was important, and one reason why MCP doesn't have to do it yet Google created it and people don't trust. If Google's got control over this, they're not going to, they'll sign up for it, but begrudgingly. And, but we better not put all their eggs in that basket.

It's an open source project with multiple contributors. And kind of like Kubernetes, Kubernetes being handed off by Google a while back, you know, there, there's much better hope with this being an open standard and being widely adopted. So that's why I think it's a big deal.

I I couldn't agree more. Right. And and thankful to the background on that, Mitch.

And this, you know, I was thinking about this before we got into this episode is a couple months ago I was on this show saying, you know, AI is neither artificial or intelligence, terrible word, you know, phrase, uh, I was kind of wrong. And, uh, there are better words, you know, in civic AI terms, we talk about semantic and solid, right? These are the team members we work with.

And this move, you know, as Mitch as you said, this is huge. This is seismic. You know, we didn't happen three months ago because we weren't ready, and three months from now is too late.

Things move really fast in this space. And a as you folks know, I mean, I'm literally doing all sorts of a agent work right now by copy and pasting from one AI's output to the other. A is interface using rock and open AI and claw and all these systems.

And we're doing amazing things. And I'm doing that in the context of the civic AI ethics framework, you know, that we've developed. And it's all working just fine.

And I don't see any path forward that lets us take advantage of all the stunning capabilities that all of this evolution, this technical evolution gives us, other than getting to a point where we can trust these, these semantic team members, you know, to behave as reliably as humans, which isn't perfectly, you know, the many, one of the four of us could lose our minds on air right now, but we tend not to. And we deal with that. And we're at this point now where we have enough latent structures laying around that this summer.

I, I expect it to bolt together. And by the fall, we'll have different conversations internally. So I have a few thoughts on this.

First of all, to the, to the issue of trusting Google as it relates to the open source world and CNCF and lf, you know, Google didn't do bad with Kubernetes, right? It was Borg or whatever it was before it became Kubernetes, but they donated it to the LF relatively early on, right? Way before, uh, official release.

It was still in Bader and stuff. And, and it's been in a cn, you know, the cornerstone of CNCF from, from get go, where Google gave themselves a black eye in this space was the Istio saga, right? Remember, they, they were going to donate it.

They, they did it, then they didn't want to donate the ip, just the copyright or vice versa. And there was this long sort of drawn out battle around Istio, which eventually Google did seed over to the CNCF, but that created an opening for other service mesh, linker d and others to really, so today, I don't know if there's a, Istio is not the preeminent service mesh right there, there are several. Um, the way I look at this though, from a bigger point of view is we, we haven't determined best practices for, for ai, for agent, for agent to agent interoperability, et cetera.

We're in a, we're in an era of emerging practices, emerging protocols. Some are will, some will win, some are loose. But in today's world, where you do need that found, we, you know, we have this foundational model that Harry Selden put in.

No, only kidding, Harry Selden didn't put in it. But, um, we have this foundational model. If you don't donate your standard or wannabe standard to a foundation, it is looked at sort of sideways as potential lock-in potential big brother vendor.

And so I fully expect MCP to be part, to be donated at some point. It would be nice. So this new project that LF started, Mike, and I don't remember the name of it.

It's, is it eight? Is it the A to a A two A? Yeah.

Eight, two A Roger, Maybe, maybe it's a home for MCP too much like CDF, right? Has spinnaker and, and, and Jenkins and a bunch of the other CICD tools. It, uh, that it manages or fosters.

Um, it would be nice if we could tie a nice little bow around it that way. Yeah. But they might also give it to, you know, there's, the Apache Software Foundation is just as legitimate at Home, and there's the Eclipse Foundation.

There's many of 'em. Yeah, I think there's, there's a couple of distinctions. I agree with what you're saying, Alan, but, and the, and the, but is with Kubernetes, you know, the, the race for owning that space was really early and nobody else was in it very much.

I mean, you had what doc, uh, Docker had scout and a few things that were around No, no, there was, well, doc had Swarm, Swarm, sorry. I said, but You also had rancher. You had rancher.

You had Mesos. Okay. But now, and that's my point is, is who outside of our circle knew of those companies.

If you weren't into cloud native co containers, blah, blah, blah, blah, you don't know about that space. In this situation. You're talking about the titans are all fighting over ai, and who's gonna win?

Who's going to, I want to control it, but I can't control all of it because my customers won't let me. It won't work with me if I, if I lock them in. So I have to support open things and do it in a way that lets people talk to other services in addition to mine.

Um, but I also don't wanna give away the farm too. So it's a much different competitive situation and who, and who was in it. And, uh, I think the feverish pace of it is also what's pushing that.

And I think why Google controlling it too much is an issue right now for the big guys. All the big guys to really play and play well together in the sandbox. Well, you know, so last Friday I had a conversation with, with somebody, I can't even really describe it too well, because I don't wanna say who it is, but one of the folks technically behind, you know, the big AI world we're in.

And it confirmed a lot of my thoughts. It was a very nice person, knew what they were doing, uh, but didn't understand where it's going. And, you know, for purposes of clarity, I gotta say that, you know, I am, I am now one of these folks.

ai is the company we're starting up that has a website now, will be a corporation by the end of the week probably, and we'll get into this space, tiny little startup. But to your point, Mitch, you know that this market is undefined, moving fast. And to your point, uh, Alan, about open source, uh, we've decided virtually everything we do is open source.

The whole civic AI process that led to this starting the company is all open sourced. Almost all of our corporate processes are in GitHub. Why not?

You know, it doesn't save any time. Things, things are moving too fast, as I've said, as a security person for 35 or so years now, it's like, you need to decide what you actually need to protect, because actually protecting things is hard. And if you've gotta protect everything, I'm sorry, say this nicely down here.

That's not a good choice. I might call you a name. Don't do that.

Right? You know, there's very little inside a company that you really, really need to secure. Put all your efforts in that, and, and maybe as wire wire is doing, post everything online, everything, you just get rid of the questions in the first place.

And, and, and to a, you're pointing, you know, you know, I'm an open source capitalist, like, companies have to be profitable. I respect Google and everybody else, you know, making their choices. But I think the choice to keep things in-house and control them is untenable.

So there are a couple of caveats in this thing. So first of all, the thing about eight oh a is now it's customizable and extensible according to Google, which could be a good thing, which also means that there's gonna be multiple flavors of this thing. And how they actually interoperate with each other remains to be seen.

And MCP for all the noise and press it gets, you know, when I talk to people, it too is kind of perceived as very lightweight. And to Mitch's point doesn't solve the orchestration capabilities and how you pull data. And so there's just a lot of stuff that needs to be done in terms of strengthening both these platforms and figuring out how they're gonna work with things like GraphQL or whatever else it may be out there.

Um, so it's very early days, and none of this stuff is enterprise ready by any stretch of the imagination, I think it is. And, and, and I'm gonna, you know, and again, you know, since I'm, since I'm one of these players now, a tiny little startup, we could be wrong, but we're taking the approach starting now. But yes, it is done, done right, and it is enterprise ready, and it's, you know, the, the approach being taken.

While I respect everybody in the, in the industry, you know, doing different things, I think we're taking the wrong approach. You know, this, you know, this is not about having an ai. There is no Skynet.

This is about, you know, getting past, you know, they are not humans. They're, they are, they are, you know, creations of whatever, but they act like humans. And we need to function with them as team members, like we do with people and not like systems.

So we're gonna try that approach and we'll see what happens. Do They have rights too? Yes.

Okay. There we go. Yep.

I, I think we can Find in Cannon, Think do on this, but Chris, we, we'll come back to that point in a minute. But to your earlier point about publishing everything, everything's an open book, if you will. Look, that's GitLab, right?

Mitch knows this. Mike knows this, right? Sid, Sid, uh, Brianon, Jay, uh, and the GitLab team, I mean, they literally published the book on GitLab.

There's a book on GitLab, their product roadmap. Everybody's salaries, who reports to who everything going on is, is, is an open book. And, and it served them well, right?

GitLab went public, successful company still. Um, but there's a difference between a company following an open source mantra and, and an open book transparency versus donating your open source project to a foundation that opens it up to coopetition. And that's the real beauty of this a to a thing.

And we have it, you know, we said, oh, the hundreds of other companies, not just hundreds of other companies, the biggest names in the market are part of this consortium, part of this group, right? Including companies that normally compete against each other. And that's the beauty, that's the beauty of the foundational system, is that cooperation model where you have companies that wouldn't normally com, you know, cooperate.

'cause they're competitors all saying, Hey, let's throw in for the common good, and then each of us will build from that common ground bedrock and go our way with it and see where it goes. It's, it's worked in cloud native, certainly. Um, it's worked in other areas, right?

Look at the, the different flavors of Linux, the different TROs of Linux, right? Um, it remains to be seen if it'll work here, but I think it, to Mitch's point, it gives people a sense of, man, I'm not getting locked in. Look who's involved in this.

It's gonna be okay. It may not be everything I I want it to be right now, but it's gonna be okay. And, and you, and you, you, you and I have been around a long time around open source.

I mean, you used a blog about it for network world and folks, um, you, there's so many different models that have happened. But in this scenario, you know, governance is always key. When, when, when company, when one company isn't controlling an open source project.

And so it's about those companies agreeing on what needs to be added and then contributing to it. Um, I think the, now I'll be on, on the optimist side of it, the optimistic side of it says, um, they know they need to work together, because if not this, what else? None of them is gonna create something that is gonna compete with it, at least not right now.

And in order to demonstrate what agents can do, you know, agents aren't an island they can't work on by themselves or work just within them. One vendor's agent ecosystem. They've got, they've got to be able to work across those boundaries and do it successfully without, you know, unnatural act by agents.

Um, so there's a real strong, um, incentive not just from customers, but just to get in, be in the market and be competitive, and to accelerate their own adoption and position around AI and agents. You know, the downside of it is, you know, there's always foibles and game playing and things that can happen to and arguments about. We don't like your implementation of that and my implementation of this.

Those things get sorted out on projects and kind of depends on how the governance structure is set up there. So that really will determine a lot of the success of this, I think. Absolutely.

Hey, we're over time on this one. We gotta take a break. Come back 16 billion with a B records leaked.

Yeah, it's just, just another number. You're watching Textron Game. Hey folks, we're back.

And there's yes, 16 billion records, a KE credentials that have been discovered on the dark web somewhere. But there's a debate as to whether or not these are actually new things, or if it's just somebody kind of aggregated all these things and put out a report about it. 'cause maybe, well, we have lost track of 16 billion records over the years, and I guess the bad guys are trying to make it easier to, for the other bad guys to find this stuff.

But Chris, what's your take here? Is this news in your mind? Or is this kinda like, you know, a little bit of hype?

It's not news. No. Yeah, in, I'm trying to think.

It was somewhere around the turn of the century if it was when the Veterans Administration's servers got hacked, and a veteran friend of mine and one of the security mail lists, you know, was complaining about that. I was making the point that, you know, I have put my social security number on so many mini mi graph forms in my life by that point, and have gone out into the world, and God knows where, that if my identity and my trust and my, you know, my, my electronic artifacts are attached to that, then I have other problems. And we got in the back and forth.

So I posted my social security number on the list, you know, of like a hundred of the top security people in the world. So there was this great silence, like, look, and this is a perfect example. Say 16 billion records, a hundred trillion.

Yeah, it's all out there. So how do we build these systems and why do they trust anything? And I, you know, I feel like a broken record these days, but this has been laid down for a long, long time.

You know, folks like Fred Cohen wrote this down years ago, John Kinder Bag, and the, and the Zero Trust thing is another example of how we're going through this. We need to have systems that base trust on Morgan, just these, you know, ephemeral, uh, uh, connections. And they're coming together.

You know, our last segment talked about this. The, the application of systems that work in the human level, you know, semantic narrative bases, and understand human stories, not just digital stuff, gives us the capability of the speed to get the kind of security stuff. That's because if not, wait, we think next week there's not gonna be 25 billion or 73 billion or 150.

Yes. That's the world we live in, and the systems don't function if that's how we identify with cells in developed trusts. I was gonna say, if it was 15 billion, I wouldn't be concerned, but 16.

That really, that puts me over the edge. I just, that's Changes. Yes.

Like I said, are you, no, did you just show up? Right? Do we have about 7 billion people in the world total, right?

Yeah. Oh, that's two passwords per person. That's assuming every single person was online and had passwords.

I mean, so here's the thing. Bad guys like publicity too, right? They, you know, there, there's something to be said there.

They want the, they weren't the, the, uh, when they take a scalp, right? The marks on their belt, on their wampum belt or whatever, they have OKRs too, don't they? They've got goals to meet you.

No, they have KPIs and bonuses, KPIs, but, um, but here's the thing. This is not new. It's not all from one hack.

It's, it's an amalgamation. And as someone who runs a pretty sizable mailing list here at Text Strong, I will tell you that the shelf life of people's email addresses, as well as their passwords and credentials are ephemeral at best. And every day that is one day further away from when that credential was stolen, is another day where that credential is more than likely useless.

And so, go ahead, Chris. And I was, you reminded me on a, on one of these shows a couple months ago, you'd asked me about, you know, quantum, uh, quantum computing and, uh, passwords and photography and so forth. At the time, I said something similar, what I'm saying now, you know, just expect things, you know, security's about time.

But I found out down the whole semantic pathway, there's a different answer to that. You know, uh, you know, cryptographic semantic cryptography, basing keys in cryptography based on narrative strings and the weights and the, the, the, the fundamentals that humans use to think with and AI models use to compute with, or whatever they do, that turns out to be really, really good. If you just talk to one of these, one of these ais at length like I do, they will come to recognize the way you speak amazingly today off the shelf, right?

So my advice, un unlike most of the times I'm asked about these sort of things, normally I say, don't change anything. This doesn't change anything. And this doesn't change anything.

However, start looking towards systems that don't fall over when it this happens because it's already happened. It happens all the time. There are better ways to do that, you know, they're merging.

Now, there are complex systems that we're aware of, and if you're in a critical environment, do those, but start leaning towards something that's not gonna fall over because somebody hacked up database Fair. And, and I a dumb question. Go Ahead.

No, they're on the dumb questions, Mike, you remember that? I am. Why the hell are we so independent on using social security numbers everywhere for everything?

Because, you know, Like it was never intended for that. Never intended, Never intended for that. I mean, and, and that's a uniquely us thing too, by the way, right?

I, I don't think you see that in the rest of the world. Um, but, you know, but it begs the beg bigger question, and then Chris already hit on it, is we gotta get outta this trap of authenticating ourselves based upon passwords or, or date of birth, social security number, that kind of data. There.

There's, you know, whether it's biometrics or something else, there's gotta be a better mouse trap we could build here and just break this cycle of, you know, it's like a hamster wheel. Well, we're already so numb to attacks, we don't even rubberneck at it anymore. You know, nobody says, oh, 16 billion.

Oh my God. Yeah, they keep driving. They don't even slow down.

You know, that's how numb we are to this. So, you know, it's time to get rid of passwords, go to passcodes, go to biometric, go to please use the password Manager for all that stuff. It's, it's like back in the crime reporting days, right?

If, if someone's house was broken into, but you know, the amount of goods stolen was less than a hundred thousand dollars, it wouldn't get covered. Well, cybersecurity is the same way. Now, if the breach is like less than 10 million, everybody shrugs Drugs, small breach, small, small potatoes, unless, and this has always been true in, in the cyber world, unless you are the zebra whose day it is for the lion to choose you.

And that's the fundamental truth of cybersecurity. 16 billion, that's a really big herd. What are the chances that the lion picks me?

But when the lion picks you, what you gonna do? Um, anyway, let, let's call a break on that one. We'll come back and do C block here.

Tesla's doing robo taxis. You're watching Textron Gang. 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 Textron Group.

Hey, guys, we're back. And yeah, a laud Musk is out really with a, an actual robax, I think he's been talking about it now for a better part of a year or maybe more. And it too got, you know, a lot of initial kind of hype, but now it's like a test pilot program running around Austin, and yet we see Uber and, uh, whammo are already putting cabs out there.

So, Alan, I gotta look at all this and say, you know, did, did Elon miss the boat here? I mean, who cares if I can get a Tesla? I already got a million other cars I can hail.

Well, first of all, I think a lot of people in LA thought those Waymo's were Tesla, and that's why they, you know, were attacking 'em and burning 'em and everything. But it, it's interesting. Let me just stay on that La Waymo burning theme for a second.

For whatever reason, self-driving autonomous vehicles have come to represent the technological elite. The one percents the halves to a certain subsection of the Audi of these population. And I think that's why you saw them attacking these Waymo vehicles.

It was a way of damaging property without maybe hurting someone. That being said was, is Tesla late to this? No.

You know, a lesson I've learned is that big successful companies, Dell, apple, don't pride themselves on creating a market. They don't, they don't pride themselves on being first to market. They don't wanna be the missionary.

They want a market to become established and have a certain girth and then come in with a plan to dominate it. Now, maybe I'm giving Elon too much credit, and he's been too busy on his ketomine or whatever the heck it is, but I think they were waiting to see if this autonomous vehicle market is real enough yet that people will trust it and everything else, and then start in Austin, which is, you know, obviously a big tech town and go in and try to become the dominant player, right? I mean, Uber's been trying, look, Uber, truth be told, Uber was originally founded with the idea of autonomous vehicles, right?

Not, not having humans in that mix, but, um, certainly, uh, Waymo's done, done. I mean, you saw all over the RSA area in San Francisco. You know, they, they've proven the model.

I think, I think Tesla's still a little early in it. Now, the flip side of it is, is do they have the goods? You know, they're going to get a certain amount of attention and people will give them a lookie because it's Tesla and Elon, especially if you're 30, one of the 30% of the people, well, even those people don't necessarily like Elon anymore.

But, um, but there's gonna be a segment of the population that say, oh, it's a Tesla. I'll give it a shot. But if as we're seeing there's been, uh, incidents caught on camera, and there's been rumors of, you know, not, not behaving well and accidents and injuries, 'cause we're in a very kind of precarious state in this autonomous driving, right?

A lot of people think it's still cool, magical, whimsical. I don't know if they really trust it and I late, early, or in between. I think that's the biggest issue in front of Tesla and Elon and the whole industry building people's trust.

As long as he doesn't scrap a SpaceX rocket to it, I think we're okay. He did that with a Tesla, didn't he? Yeah, that was early on, but when they weren't blowing up or they stopped blowing up, Chris.

So two, two comments. One, one, you know, not company specific. You know, as we're talking about, you know, right now in this very episode, talking about whether we can trust an AI agent instead of a human agent to decide whether our database talks to something else, and the complication with that, you know, here we're talking about, you know, a, a AI agent driving a vehicle with humans around.

And now in the first example, that database may actually control traffic lights and human life to be at risk and so forth. So there's really no great difference except that we physically see cars driving around when we're standing there, and it makes us think about it. So essentially, it is exactly the same issues unless we can trust, you know, these, these semantic AI agents, you know, to act like human agents in similar roles.

We can't have any of these things. And the second part of that, uh, interestingly, I've been spending a lot of time working with Crock, uh, recently, you know, the, the Elon Musk ai, a great deal of time, and we're working on right now, uh, a project to look at 400 millisecond response time to get an as bomb across multiple supply chain hop hops inside SpaceX in, uh, infrastructure. So I'm getting a good feel for, so bro is an interesting example of an AI here.

com, um, and it, it's a single entity that understands the whole thing. So it's, you know, I don't, I think everything you guys said is just true, is really, really early. If we can't trust AI to get two databases talking to each other, can we trust 'em to drive cars?

No, obviously not. Do we need to work out that both those right now, both because companies are spending billions of dollars making robo taxis and actually driving 'em around the streets, and so we don't have five years, five months we gotta work, we move this one 'em forward. And because on the IT side, yeah, you cannot run big data houses and big companies anymore today without some semantic age and help, and we don't trust them.

So all that's gotta work out real soon. Now, RSN as we used to say, right? It's also, it's about trust and it's also about use case.

This is the best use case you didn't notice. We don't have any cities that get a lot of snow that are piloting, you know, self-driving cars yet just 'cause that we're not there yet. So these are, you know, what is Austin hot?

Lots of sun and occasionally get some rain. So it's a great environment to, you know, pilot other than a lot of people around you wanna kind avoid eating. It's a good environment to, to pilot that.

Wait, Wait, wait. Waymo just applied to, uh, for approval to New York City for their service. So they, maybe they have some figured out that snow issue.

Oh, cabs take out people all at the time, so they don't care. I'm a huge, I've always been a huge self-driving, uh, car fan since childhood. You know, I expected to be a, a tell my grandkids like, oh yeah, me and grandma are used to pilot metal vehicles and so I really wanna ha see this happen.

But for the same reasons, I believe with the whole civic ai, not just our project that go, there's a lot of pieces moving around, figure this out. We have to figure this one out. No, Waymo doesn't have it figured out.

No. You know, Tesla doesn't have it freight out. No, these things are not Stitch we'd driving in the streets because the four of us can't decide if they're safe to have do two databases talking together.

But we have the parts and we can make it. So, and the fact that these companies are taking the risk, hold on a second, the equal risk that may drive it to an answer. The fact that the four of us can't decide doesn't mean that there's not an underlying truth there.

It's ju you know, let's se it's just that maybe the four of us are just tri lights or what, whatever, right? That doesn't mean that the damn thing doesn't work. I mean, rainbow is, oh, oh, it, I'm sorry.

It, I get excited. It does work. You, we actually can do, you know, put the two things I said together.

You know, the car shouldn't be out there right now 'cause we don't have the ethical model to trust them. But I'm also deploying AI at enterprise scale today. 'cause I, I think I, we have a way to do that without the risks that's not only adopted.

There's the same thing. You can't, you can't have it both ways. Yeah.

But, but listen, there's a reason why Waymo started in San Francisco and there's a reason why Tesla is going to Austin because these are heavy tech talents. And I think tech people, my people, you watching this out there, we're more not susceptible, but we are more likely to give it a try than some dude, you know, in Times Square on Broadway is, and I don't blame him. Times Square gets pretty messy with traffic, right?

So, so I have a question though. Like, so where are the big car makers in all of this? So won't Detroit, Japan and Europeans all go build their own fleets of cars that are autonomous?

I'm sure they're working on it. I, I'm, I think they all are working. Aren't they all have plans?

But again, back to what I said before, they're not, they're not looking to be missionaries. They're not looking to make a mark market. They're looking to dominate a market.

And once the Waymo's and the Teslas and the, and the Ubers and what have you had their way, you will see GM come into it and Mercedes and BMW and Toyota and the rest of the world's great car makers. Um, that, that's just, you know, it's too early for them yet because they make no mistake. What's the difference between a robo taxi and just your robo family car, right?

There is none other than maybe the robo taxi has some built in payment activity with a meter or something, right? But if we're going, if we're going robo, we're going robo. It's not just a taxi thing or maybe it's, if it's Software is the difference, right?

Yeah. Right. Conversion soccer.

And to your point, I was having this conversation with somebody yesterday and it was like, well, what's the difference between a robot and an autonomous vehicle? Because isn't a robot essentially driving the vehicle? So won't Always, it's a specialized, it's, it's, it's not driving the ve it's a specialized robot.

If, if we're gonna take a, you know, an Elijah Bailey kind of view of robots, that's two Asimov in one day, guys, gimme fi gimme something on that now. Two different series, Two different series to vote. Well, well, Emily, you know, I'm trying, trying to think of anything useful to add.

And you know, you and I and David Bryn have been discussing, and you know, his book film people, you know, we're having these conversations about the ethics of things and it's, it's, it, I I feel like a bit of a missionary. And you're right, the big car companies aren't, shouldn't, wouldn't, don't. When I was Cisco, I wouldn't assist.

You know, we just watch out this, right? But it's such a simple thing. And there's some speculative, fic fiction.

David did some good stuff, you know, when we stopped. And it's, it's interesting, you know, 'cause these machines we've built, we built out of words and meanings. That's why I say semantic because that's how they work.

They're not gears and cogs. They're not ones and zeroes. They function by using the things we use to function for the same purpose and our, our inclination, our hesitancy to refer to them using words we use for people.

You know what I mean? That's how we get ants. That's why we can't have nice things.

That's why we keep getting caught in all these things for fascinating reasons. Because as we think, you know, when I think about a coffee cup and I wanna do something with a coffee cup, I use different structures in my head than when I'm talking to you. So if I talk to you like I'm talking to a coffee cup, my brain doesn't work and the things come outta my mouth.

Don't make your brain work. And so we've made these machines, they are, they're not human, but the only way to make them work is to talk to them and think about them like they are people. And in teams, you know, when you're having teams, I had a meeting last week with two humans, gr and Lumina, and it wasn't about ai, it was a one hour working session on the space supply chain system, and it was just four team members working on something.

If you, if we think about it that way, then we can all live together. And if one of 'em goes rogue and kill us, kills us all. So can I.

Right? That's, that's any human is huge existential risk to all other humans, but we somehow work it out. I got one question for, I'm brushing over the Chris, I got one question for I'm just you, that's the way we need to go.

Sorry. Did you ask them to open the pod bay doors? I have one question.

What are you talk about at your meetings? And normal. Normal, This is funny, right?

This might go rogue and kill people just Like this. There's four people having a conversation at the end. What's the next steps?

Move on. I Need, I think I need to stand your ground loft for ai. That's what I got outta That.

You're gonna start shooting them. Alright. Hey, you on that note, we gonna, I'm gonna have to pull the plug, Mike, when you're talking about getting violent.

Oh wow. It's enough. Uh, great conversation today guys.

Hey, heads up Thursday, tech Strong TV is live at Platform Con in person in New York City. Stay tuned. We'll be, uh, streaming live there, I think around 11 through the rest of the day.

Eastern Time. 11:00 AM through maybe 4, 4 30. Um, looking forward to that.

Platform com by the way, is all week on the virtual aspect. Mitch, I know you have a big post up on LinkedIn, um, so you could go check that out. It's free to do that.

Mike and Mitch are in Denver actually today, and they were there yesterday and Monday. I don't know if they'll be there. We are you there today, Mike?

Wednesday? I'll be there today. Last day is today.

Yep. So three days there at the OSS event. Chris is in meetings with some of his AI friends.

There was a time where they'd lock you up for that stuff. By the way, I'm doing an analyst panel myself and someone from IDC and Forrester about platform engineering and how it's impacting developer productivity. That's, I might have to sit in on that, Mitch.

Maybe, maybe we'll do an article on platform engineering com that come around It. Great. I'll give you some questions to throw at the other analysts.

I'll take the softballs. You're not supposed to say that out loud, Mitch. Oops.

We're recording this. Are We alright? Until tomorrow.

This is Alan Shimmel on behalf of Mike, Chris, and Mitch, have a great day everyone. We're outta here. Hey everyone.

Welcome back here to Techstrong tv. You know, I haven't had my friend do on here in months, may, maybe even a year. Let me introduce you to, uh, do Leor DOR is the CEO and co-founder of Steeler DB joins us today from, from Israel.

And welcome, do and thank you. And, uh, we appreciate you coming on. How's everything with Steeler db Alan?

Good to see you. Good to see everybody. Thanks for hosting me, uh, is doing great.

Uh, we're working our behind offs in order to, uh, improve our releases. Uh, so all, all good. We re recently released, uh, X Cloud, which we'll probably talk about and, and variety of other things.

Very cool. Dob, just for people who maybe are not familiar with you and Cila, give them a little bit of your history and a little bit of the history of Sierra db. Sure.

Um, so I'm an engineering. My roots always been, even now that I, I don't write code. I involved in the product technical person.

And, uh, in, in the past, uh, I built a terabyte router company and, uh, later on I parted with my co-founder Avi, and we created the KDM Hyper Hypervisor. Uh, that was 20 years ago. Uh mm-hmm.

We, long time ago, we were involved with, uh, like we brought K DM from the beginning and we highly were involved in Linux after Red Hat acquired that startup. And, uh, more than 10 years ago, we created CA db. Uh, CDA is a highly scalable database company.

It's a no SQL distributed database. Absolutely. com and everything is there.

You can, people can download and, uh, kick the tires. Absolutely. And for tho from you people at home watching it, it's spelled just the way it is under door's name on the video.

com. Alright, we got that all out of the way door. You guys recently announced, as you mentioned, uh, X Cloud, Correct?

What, what's X Cloud? So, um, uh, we have, uh, our order release, which you consume, uh, eh, with a self-managed version, the, the enterprise version or source available version, and also a fully managed database as a service in the cloud. It's called C Cloud.

Uh, recently we released a new one, a new release called X Cloud. What's special about it, people love the X these days, so we added X to the product, but, but really what they X does is a, allow us to be the most elastic database in the market. People need elasticity.

Uh, why would you need elasticity? Because, and, and number one, you first need the database that can perform and can scale with the high availability to be a mission critical, uh, database. This is the, the, the number one need afterwards when you, uh, start to use it, then, um, uh, if you scale to a very high extent, it costs money and we like to make money, but sometimes your, your peak level varies and the usage varies.

Uh, there are cases where a Black Friday or similar event come around and, and then you need to scale your deployments. Uh, this is more predictable scale. There are unpredictable scales that, so sometimes you, you have a really good success, and then many users come to the web, to the website, and you need to scale a database.

In many cases, the, um, the elasticity requirement is daily people come, uh, wake up, go to, uh, our customer's websites and, and our customer's mobile app and whatever they do generate loads. Uh, during noon they eat. So load decreases a bit.

Then there's the second pick of the day, and in the evening they go home and rest. So there's, if you pro provision 100% for the peak, you are wasting on resources. Uh, on the other hand, uh, traditional databases and prior to this X cloud release, uh, if it's very hard to, uh, always meddle with the, uh, number of, uh, servers and infrastructure related to, uh, uh, a deployment, we, we have deployments, uh, at the side of the PE petabytes.

So moving around petabytes, uh, within, um, within several hours is extremely externally difficult task. Uh, X cloud is a release that allow us to do it faster and allow us to double or quadruple, uh, the throughput for particular customer within minutes. Um, so the, this is what's the, the value of X cloud.

I can talk about what is there under the hood as well. Excellent. Um, you know, in many ways this whole idea of this elasticity, I feel deja vu.

I was told this about the cloud in general, right? That was one of the things about in the cloud, you burstable, but it, but what it proved to be in the cloud door was like, if you blow up a balloon, you know, and it stretches out, and then you let the air out of the balloon, the balloon is still stretched out and mm-hmm. And for so many people, elasticity on the cloud is sort of a one way ray, a one way, you know, direction.

Yeah, you can always make it bigger, but how many people really do make it smaller? Is this something also true in sealer or, or can? Is it truly that elastic?

True. A, absolutely true. B, b, everything you, you, you, uh, you, you've described, um, a a and it's also true that even regardless of the infrastructure and how cool X cloud is, many times people just keep on adding more and more data.

So if, if you just add and you don't delete, then it is what it is. But, uh, that we, we have different capabilities. Uh, there are normally two reasons why you'd want to scale out and, and why you'd like to scale in.

Number one is storage. So if your storage grows, you, uh, uh, will, will scale out and add resources. If you delete data, we'll automatically, uh, scale in back again.

And, uh, our current release have two unique things. Number one, it has, uh, autoscale with the storage. Uh, the, the storage automatically scales and, and, uh, our compute is bundled together with the storage.

At the end of the day. I'll tell you a secret. We run servers, the thing that's called servers, and the server has, uh, memory and disk and networking and compute altogether that, that's some of the reason we have really good performance.

We bundle compute and storage. Um, so, so, uh, if the amount of storage people save decreases, we automatically decrease the amount of servers. And, uh, we do it with 90% utilization.

Uh, so the server, the, the disk capacity can go up to 90% of, uh, of the shared, uh, uh, cluster infrastructure. Uh, it, it's very high percent. Now, previously we went, uh, from 50% to 70%.

Now, because we're very elastic and we can move data very fast, uh, we can go up to the very end, uh, up to 90% and still allow people to use it as, uh, and, and it's all automated. So when you go, uh, to 90%, we automatically provisioned more servers. If you go below 90% there, there is some threshold, like 85, then we automatically decrease servers.

And we also do it in a way where, um, let, let's say if you have, if you choose to have very big servers, let, let, let's say you have three gigantic servers. We, we support in all servers up to 256 CPUs. Uh, if you add, uh, if you'll run out of space and you'll add another gigantic server, then suddenly your utilization will be low.

So it's, it's bad for you to add a big server if, if you just need like another one or 2% of utilization. So what X cloud is, is doing, it automatically selects the server size for you. And we mix server sizes.

So if you only need extra 5% along with the, these very big servers, we are going to add a small, tiny servers with, uh, two vcps next to those other big servers. And we will keep, replace it automatically for you without you knowing as long and it translates to value. So you need to pay the, the total cost of ownership will be low because we were targeting 90% utilization of this capacity.

Uh, the, the other, uh, reason why to scale is sometimes, uh, throughput and CPU consumption. And this is even easier because we, we need to move less storage. So it's, uh, provision new servers with less storage.

That's easier for us. And we do that and we allow our customers to, to scale in and out multiple times an hour. Got it.

That, that's a great, very in depth. Thank you very much for that. You know, dore, all the news today is with ai, ai, agentic ai, generative ai, LLMs underlying all of this though, is, is data, is is the, the, the, you know, that they're training on, that they're storing, that they use.

How has this whole kind, you know, it's a whole world unto itself. How has this affected this field of business? Um, it, it's definitely drives our business.

Not only ai, but absolutely, uh, uh, it helps to drive our business. We, we have three main, uh, pillars with, uh, related to ai. And number one, we have traditional machine learning.

We have feature store, so everybody who needs to customize automatically their, um, their infrastructure and, and fit it into customer use cases automatically. And segment that. There is a industry standard called, uh, machine learning feature store, and CLI is used there.

For example, TripAdvisor is a customer of ours, and they use feature store in order to find the best recommendation and the, the best deals and the best, uh, advices to their customers. Th this is number one. Number two is, uh, database usages, uh, AI usages that needs, uh, large scalable storage underneath.

W we have a very large, uh, um, vehicle vendor, which, uh, I cannot say their name, but it's one of the largest. And, uh, and they use, uh, CLOC to train their AI model for, uh, for, uh, for, for, um, automatic AI sales driving and cars. So that's another, it's not just, uh, vehicles, but it can be anything else that drives a lot of data.

They need to access a lot of, a lot of data and a lot of objects, standard access pattern in, in databases, but it happened to drive ai. And the third one is, uh, vector search. Uh, so vector search is, is a component that supposed, uh, a rag, um, automated retrieval to together with the agent, agent agent ai.

So you can run your query, but, but your query, uh, your query, you, you are going to query a customer of ours. And, uh, the, the query should automatically go to private repositories that, uh, chat GPT cannot access and augment the data retrieve, like, uh, um, when is my flight, uh, will going to land? Uh, it, it needs to fig to retrieve that, uh, your ID to figure out where is your particular flight is supposed to, what is the number?

And then get, get to the predicted, uh, eh landing, for example. And, and this is what, uh, vector search, uh, allows databases to do. And, uh, with GA or Vector search, uh, product, by the end of the, the year, it's not yet.

Well, it, it's only, uh, closed be You heard it here. Okay. Thanks for sharing that with us.

So, I, I just wanna make sure we hit the, the main points of this X Cloud offering. So with, with all of these things that you'd mentioned already, really what this results in right, is you, you, you've improved compression, improved streaming. So not only are you helping to reduce the storage, you know, cloud costs, but it it's the network as well, right?

Your bandwidth and, and everything else that, that comes in that, and that's an important piece of the, of the equation as well. Um, the other thing I want to make sure our audience understands if they're not familiar yet with SA is all this is offered as a basically database as a service, correct. DB as a service.

So you don't have to worry about your infrastructure. When we talk about storage, it's, it's all art of the seal a offering, right? You don't need to do all that.

Absolutely. We, we, we have high amount of automation that everything is done on behalf of the end user. All, everything.

We, we don't do anything. Our, uh, a manual ourself, everything. It's, it's, uh, the, the elasticity level is supposed to be that, uh, uh, the cluster is breathing, can, can breathe all the time.

You just need to set, this is my utilization, this is my ex expected set of, uh, SAP power, and that's it. The, the cluster will continue from that point onward. Um, as, as a customer, you just need to be connected and everything runs under the hood, including a b backup.

For, for, for example, uh, if a customer, for example, runs at, uh, 89% utilization, uh, and, uh, we run a, a, a daily backup. Uh, so, so could be that the backup will, uh, generate a snapshot and the snapshot, uh, we'll generate some data and, uh, temporarily, uh, will go beyond 90% capacity. Uh, this, the, the second we trigger that automatically will provision more servers, uh, to the cluster.

Uh, because it's just an event of crossing 90% utilization. Uh, once the backup, uh, will be complete and, uh, the image, uh, backup, uh, snapshot will be loaded to S3, uh, then the image gets deleted, uh, the, the cluster will have less than 90% utilization and we can automatically decrease the amount of servers. Everything is automatic.

It's, uh, it's really, uh, fascinating to see all of those use cases run under the hood and to see the metrics and to see it happening. Absolutely. Just wanna also make sure Dora X Cloud, this X new release.

This next generation available now. It's available now, correct. Just go to the website if you sign.

And what about existing sealer customers? Do they automatically upgrade to it or no? Um, it's, it's not like, 'cause the, the previous generation you used to, uh, um, the divide the data in, in using some older algorithm and do we have, um, migration process in order to move between the old generation to the new one?

Uh, it, it's a, an automatic process, but, uh, it, it's need to be, uh, it is a procedure we, we do together with the customer. Excellent. Alright, Dora, I think we're about outta time.

Thank you for coming on. Congratulations on this next generation of Sealed db. You guys always stay one step ahead, it seems of, of what's going on there at the leading edge.

So keep doing what you're doing. com. Check it out.

We're gonna take a break here on Tech Drunk TV. Will be back in a second. Hey guys, thanks for the throw.

We're here with Boris Kuer, who's COO and CFO for Smooth Stack. And they do a lot of work in training and we're talking about whether it folks are ready to become AI managers. Boris, welcome the show.

Thank you so much, Michael. Happy to be here. What do we mean by that?

'cause a lot of it folks would say, well, what do you mean? I gotta think about becoming an AI manager. I am as prepared as anybody else.

And maybe more so they would, at least some might argue. What's your sense of what's going on here? What is different about AI as a new emerging technology that it, people need to maybe rethink some of their assumptions?

Yeah, so, um, the landscape of software development has changed, uh, drastically, um, over the last year, I would say. Um, and I would say the biggest difference is that, um, software developers now are really generating about the last 20% of, um, of code, right? The first 80% is created quickly, um, by ai, um, for developers that are adopting ai.

Um, and it's that last 20% where the architectural thinking, um, and the real engineering happens. And so there's a big reliance on AI to, uh, generate code quicker, faster, um, and, uh, and jumpstart, uh, velocity. It's not hard to imagine, and especially in the land of software development, how the future might be, uh, a bunch of AI agents that it, people are now managing to accomplish a task.

And many of those tasks they would previously have done manually. What is it about managing and orchestrating AI agents that's different from the way it normally does their job or the tasks they perform? Yeah, so it's really interesting.

There are a lot of differences between an AI agent and a human right. AI agents don't get sick, they don't need to sleep, they can work round the clock. Um, but what they're missing is that business context, um, and overall kind of context of what needs to happen within a project.

And so at the end of the day, um, you can't hold an AI accountable for bad code. You can only hold humans accountable for bad code. And so who's overseeing what the AI is doing?

Who's prompting the AI to do what it needs to do? Um, and at the end of the day, uh, making sure that what you end up with from a product perspective meets the requirements of the business and the stakeholders involved. We tend to think of software developers as a type of knowledge worker.

Um, but are we all evolving into something that's not just, say, a knowledge worker as much as maybe we're knowledge managers? And that requires a slightly different mindset, Some somewhat, um, I think you have to have inherent, you can't just, folks are calling it vibe coding these days. And, and I think that means different things to different people, but you can't just become a vibe coder where you don't understand, um, the context of what the code is actually doing or how it's written and rely a hundred percent on AI to, to generate that code.

Um, AI isn't writing perfect code. It's writing plausible code. Um, and plausible code can be very dangerous if, if it's not reviewed.

And so it's really incumbent upon individuals, humans, um, to understand what that context is and actually make sure that the code is not just plausible. Um, it's, uh, secure. We're not blindly trusting the ai, um, we're validating the AI generated code before it merges into production, et cetera.

Mm-hmm. So I would, I would argue that the knowledge, uh, knowledge work is still required and heavily dependent on, on humans. Mm-hmm.

One of the things you will hear from software developers working with these tools is that they find that they are reading and reviewing more code than they're actually writing. And is that the way this job is gonna evolve? Um, potentially.

Right. So really the idea is, and we can equate this to folks that have used chat GPT to write an email, um, or, you know, any, any other kind of thing that they use it to, to prompt, um, some sort of verbal communication with. You can prompt the ai, but the AI is not going to give you a perfect result, right?

You still have to review what the AI spits out, and you have to alter it to fit your specific needs. And that change in what the actual end product is, could be dependent on how well you're prompting the ai. Um, but in, even if you prompt it perfectly, uh, you're still probably going to want to, uh, review it and edit it to, to make sure that it meets your needs.

I think coding, um, you can draw a big, uh, comparison to that from, from the coding perspective in that you can, uh, eliminate a lot of the menial kind of upfront, uh, coding structure and, and tasks that are required, um, and really add the finishing touches that require the most amount of brain power from, from the individual. So, um, I would say that, um, yes, in, in some sense that is true, but you're essentially what you're, uh, able to do by utilizing AI is do a lot more because you can eliminate some of those upfront tedious, uh, kind of tasks and really focus on, on the meat and potatoes of the the problem. What do we need to do in terms of training for folks?

Because a lot of times when I talk to people, the general expectation is that somehow, or that they will magically go on their own and learn how to use these tools and figure it out for themselves and, you know, and then the boss gets a little annoyed when he finds out that that's not happening. So is there some smarter way to go about kind of exposing people to these technologies and training them? Absolutely.

Um, if you don't have a rigid training process and a change management process, then um, folks will just adopt and use AI at different levels and they'll do it, um, based on their own, uh, judgment, um, or they won't adopt at all. Um, and as, so I think, you know, as leaders kind of think about this problem, uh, there's two things that, that they should do right away. Uh, the first thing that they should do is set up a, an AI usage policy immediately, right?

So, um, things like establishing guidelines for prompt design, um, tool access, data sharing, um, educate developers not to blindly trust ai, that's, that's a big rabbit hole that that folks can, um, go down, um, monitor AI tool usage and track adoption, uh, validate AI generated code. Um, the second thing that that leaders need to do is proactively, uh, incentivize adoption of ai because your more senior developers are not incentivized to adopt ai. And, and the reason why is because their productivity actually declines initially because they're really good developers.

They've been doing this for many years, um, now they're introducing a new tool, um, and they're having to learn that tool. And as a result, um, productivity goes down, which is not typically desired in a senior developer. Um, so I think the change management component is, uh, is very significant when thinking about, um, you know, what, what leaders actually need to proactively do in, in terms of, of training and ultimately getting adoption because, you know, leaders are looking to AI for, uh, increased productivity.

Who should be in charge of ai? 'cause I think there's an assumption in a lot of organizations that this will just get managed by the CIO and the rest of the IT team, but I also see organizations where there is now a chief AI officer, or sometimes it's the data officer that's assumed those responsibilities. Is there, is there a right answer here?

I don't think there's a right answer, and I think the answer can change depending on the size of the organization. Um, some companies can afford to hire a chief AI officer, um, and, and have a specific role carved out. Um, and some, some companies, quite frankly, cannot.

Right? And, and I don't think that AI should be, um, just geared towards larger companies. I think that there's a place for AI in, in small and medium, medium-sized companies as well.

Um, I think it really depends, uh, but I think the, the most important thing is to have somebody that is charged with determining, um, how to utilize ai, what are the use cases within each organization, how to effectively adopt it, um, without introducing security risk and, and all of that other stuff. And, um, ultimately how to increase production to, um, to give your company an edge. Um, so I don't think that there's a specific role, but I think that, um, somebody at the company leadership needs to give that, that lantern to somebody.

Um, and, uh, it should be somebody's, at least a portion of their time spent on, um, on determining these things and really setting up this, this process. And, and again, a part of it is, uh, the change management in our case within our company. I'm involved from an operations perspective, but, um, I partner very closely with our CTO, um, with our, with our internal training program as an example to produce, uh, AI native developers.

Mm-hmm. Um, how much in this going forward, do you think is a technical challenge versus really what amounts to being a cultural challenge inside organizations and it's an exercise in business process re-engineering? I think it's, I think it's really both.

Um, I think that's a really good question and, and something for leaders to think about. Um, you cannot, without considering both of these things, you cannot successfully adopt AI within your organization. Um, some of it is as simple as, you know, companies are not utilizing AI because they don't have a use case for ai, or they're not aware of what the use cases are for ai.

Um, and so they, you know, they're, they're slow adopters or, or late adopters. Some companies really want to utilize ai, um, and they understand the value. They have clear outlined use cases, um, but they can't get the adoption because they don't have, they haven't laid down the framework for their employees, uh, to easily adopt.

Mm-hmm. So among the companies that you have seen making this transition, well, what are they doing and others are not, They're putting a specific focus. Um, I read an article over the weekend about Intuit and some other companies that are actually in their job postings, uh, posting, uh, looking for folks that have adopted AI within their coding practice.

Um, you know, sometimes companies will look at, uh, AI negatively, oh, you're just, you're cheating, or you're just using AI to generate the code. Um, and some companies find real value in folks that are what we call native AI developers, meaning, you know, how to develop on your own without the use of ai. Um, and you're able to adopt AI and use it in such a way where, um, you're able to essentially pour lighter fluid on your work and make it burn a lot quicker, right?

And hotter, you're able to do a lot more with, with, uh, limited resources. So, um, I would say that, uh, um, it, it's really, really important to focus on that. Alright, well, folks, you heard it here.

The thing about AI is it changes the way that we interact with machines, but it maybe just, maybe it will also change the way we need to think. Hey, Boris, thanks for being on the show. Thanks so much.

Appreciate it, Mike. All Right, and back to you guys in the studio. Hey everyone, it's me, shimmy.

Welcome to another Shimmy. Says, I wanted to talk to you this week about a lot of what I'm seeing in the news around ai, whether it's ag agentic, ai, a generator of ai, or whatever flavor of AI you're, you're into today. Um, but it's around AI and jobs, right?

We're starting to hear this again, that AI is going to take your job away. 42% of people are nervous about AI taking their job. Uh, the state of New York is now starting, uh, to track a jobs lost to AI and AI job loss tracker.

I think it's something premature, the state of California trying to again, pass legislation that will limit or somehow regulate use of ai, seeing it in the EU or seeing it all over. But on top of that, we're hearing Microsoft is preparing yet another round of layoffs, and they're going to, you know, they're claiming it's a result of jobs being eliminated due to ai. We've seen it with Salesforce, we've seen it with a lot of the big tech vendors, and I get it right.

I, I understand the angst out there with people because this is what the press and the media are just, you know, go into town with. But, you know, I think a, a little reality check is in order here. Number one, I, as it exists today, I don't see AI necessarily doing people's jobs.

I think it helps people do their jobs better. It helps people do their jobs faster. It, it, you know, it helps people do their jobs.

It doesn't replace people in doing their jobs. That's not to say that at some point in the future it will, and probably sooner than later it will replace some jobs. But from everything that I see at this point, most of the output from AI still needs to have a human involved in the pipeline there, in the chain supply chain there.

And at best, it's an enhancement. It's not a replacement, right? So digital workers, as, as the term is being used for AI, that will replace people.

I, I don't see that real today, but here's what I do see, I see a lot of these companies that are doing layoffs, claiming that AI is taking these people's jobs, using that as a, as an excuse. A lot of these people, a lot of these companies rather, especially in the tech industry, you know, during CO maybe even before COVID, they, they were hiring like drunken sailors. It seems like we couldn't get enough DevOps engineers, we couldn't get enough developers, we couldn't get enough security people, we couldn't get enough tech talent and, you know, price salaries were going sky high, good developers making three, $400,000 or more a year.

And, You know, we were hiring them as fast as we can get 'em. And they would, and as soon as we'd hire 'em, some, you know, head hunter would call them up and tell 'em, Hey, come over to the next job and you'll make another 40 or 50, and they'd hop and, you know, so the average tech worker was staying at their job for less than 18 months before moving on. So the industry as a whole went on a tremendous hiring bench.

And I, I think what we're seeing now is still the remnants of that binge working its way through the system. But I think it's somewhat dishonest to say that these positions are being eliminated today because of ai, AI will do people's jobs at some point in the future. But as I said, I, I really do think that it is a, uh, a helper more than a, it's a co-pilot.

That's a good way of putting it. I heard someone else say this, so I'm not gonna take credit for it, but AI today is a copilot, not a pilot. And I think that's an important distinction to be made.

Um, that being said, though, what could you do? What could you do to protect yourself and make yourself valuable in this new AI economy? Well, number one, I think you gotta treat AI as a tool, as your friend, not your enemy.

You've gotta invest the time in learning how to use ai. And I don't care whether you're in tech marketing or something else, you've gotta learn to use ai. 'cause if you don't, if you resist that, then you really do run the risk of being a dinosaur.

So learn to use ai, embrace ai, learn to make yourself more valuable in this coming economy, because I, you know, I, I've seen, I've seen these things come and go and the people who embrace it and don't stick their head in the sand are the people who benefit the most from it. So go do that. Get, go take an AI class.

There's plenty of online courses. It just play with it yourself. I mean, you can talk to it, it talks back to you.

Now it's, it's pretty easy. But if you are intimidated by it, and then I do think you're gonna be a casualty here. So my advice is go take care of that.

Um, embrace ai. Don't stick your head in the sand and wait for your, wait for someone to take your job. But as I said, I think that's more in the future.

I think a lot of companies are using, using that as an excuse right now to trim their payroll, even though they're doing very, very pretty well profit wise and Wall Street wise. But, you know, this is the way of the world. Until next week, this is Shimmy.

Thanks for joining me today.