SPEAKER_03 0:47

Wow.

SPEAKER_02 0:48

There you go.

SPEAKER_03 0:49

Wow. This is like warm-up before he goes on stage, right?

SPEAKER_02 0:53

This is it, man.

SPEAKER_03 0:54

For the show. And I used to make fun of how easy grunge was. Smells like Teen Spirit. Is that what you were trying to play there?

SPEAKER_04 1:05

Yeah, but that smelt like shit, you know.

SPEAKER_03 1:08

That was like it felt like that was kind of the song for a while, and then you were just going Pacific Northwest.

SPEAKER_02 1:14

I appreciate that.

SPEAKER_03 1:15

Yes. You know, a little tri a little extra trivia about that song. That was my wife and I, our wedding song, our first. Yes. We got married in Seattle. And you know, right in the middle of the grunge deal. I know it seems weird, and like it is weird. You guys are freaks. It's supposed to be like Frank Sinatra or something like that. Yeah, exactly. And uh yeah, you know, it was the times. We were, I don't know, you know.

SPEAKER_04 1:45

You were young. We were young.

SPEAKER_03 1:47

We're we were crazy kids, and Smells Like Teen Spirit was our first song. We were at a club, we were at uh P, you might remember celebrities uh down in Pioneer Square. I mean, it doesn't exist obviously anymore, but it was around back then. And um that was right when all the stuff was there in the DJ. We yeah, we it I don't know. It's a weird story, but it happened, and that's who we were. We're raw, young, crazy kids. Yeah. And um, there you go.

SPEAKER_04 2:15

That's awesome. Yeah, a little bit of Tiffany history, little Tiffany trivia there, yeah.

SPEAKER_03 2:21

You know, and you know, we were talking about Van Halen and and um David Lee Roth earlier about writing the songs and everything. You know how there's like so much fake news on all kinds of media now, AI generated news, stories. Like, so basically, everybody already knows this, but we'll just say it. So on Facebook, the Facebook, it's no longer really mostly about seeing what your friends are doing or anything.

SPEAKER_01 2:46

Oh, yeah.

SPEAKER_03 2:46

Not only is it not only is it the whole self-reinforcing bubbles that got them in trouble, but there's just all kinds of just news stories about and they find out oh you're into music or you're into movies or you're whatever, and you just start seeing in these long-form things and they suck you in. So, related to Eddie Van Halen, and obviously there's a lot of talk in the last two weeks about Michael Jackson because his movie's out, and even though the critics are panning it, it's headed towards a billion dollars because the critics are panning it because they don't include all the bad stuff about Michael Jackson, you know, that happened. It's just the concerts and all the love, which is fine. And so, and it's in fact, I think Thriller is now back uh in the top 10 in the Hot 100 billboard list. Um, you know, it's cool to see young people, you know, because I mean that was a phenomenon. But there was like this story, it was talking about how Eddie Van Halen getting together with Michael to do Beat It, the guitar solo. And so the whole story, you're reading it, you know, and Eddie Van Halen gets a call from Quincy Jones, yeah, you know, and he's like, Well, he's like, Yeah, you know, uh Michael to show he's for all kinds of music, he's gonna do a rock song, and it's called Beat It, and he needs some help for you and all that kind of stuff. And and you go, and he's like, Oh, yeah, I got time to come down to the studio next Tuesday, and then he comes down there, whatever, and he's like, Yeah, you know, because uh, you know, we need you to do this song, you know, Thriller's ah, and it's a real success, and so we need your help, you know, doing a guitar solo for Beat It. And I'm like, wait. If the album Thriller is already out, then it's too late for you to write the guitar solo for a song that's on that album. A lot of people obviously didn't get that. Like the fatal flaw in the AI was he's you guys aren't getting it either, are you? No, no, yeah.

SPEAKER_02 4:45

It's an impossible story.

SPEAKER_03 4:47

It's an impossible story. Yeah, it's an impossible story, but we're just waiting for you to get to the end of it, that's all. Yeah, I guess.

SPEAKER_02 4:57

It's only a one-hour show, Rob.

SPEAKER_04 5:00

We only have a TikTok attention span. Come on. Oh, sorry, sorry about that. No, no, no, people need it.

SPEAKER_03 5:05

No, no, no, I've touched clip, but there's like so much.

SPEAKER_02 5:09

No, I know, but people don't, you know, it goes over like you said, that the uh attention span is light, and so people are just scrolling through, like, oh yeah, it must be true.

SPEAKER_03 5:17

Yeah, and they're like, you know, the because the real story is Eddie Van Halen was never with Michael, he was like on some island somewhere and recorded it separately, the whole thing.

SPEAKER_04 5:26

They were never together with Steve Lukater.

SPEAKER_03 5:29

Yeah, they were never together.

SPEAKER_04 5:31

Steve Lukather threw it threw down most of the guitar parts actually, and then Eddie Van Halen rearranged that bit, you know. Yeah, right, you know. He did that whole breakdown and then so that he can yeah, because he didn't like it just being like, you know, uh a single part or uh uh motif. So he changed, yeah. He does what Eddie Van Halen does, does all kinds of crazy stuff that sounds incredibly cool. And of course, he didn't get paid for any of it. It was like one of the greatest, most popular songs and biggest money-making songs in the history, you know. Right, yeah, but he didn't care. But you know what? That was that that's who he was back then. You know what? That's the thing, authenticity. But even a ton of credit to David Lee Roth, he was incredibly authentic himself, you know. Well, sort of he kind of riffed off that one other dude. I forgot the name of this individual, but you know, he did all this like you know how many people could do this dancing type of stuff.

SPEAKER_03 6:38

Yeah, how people could jump into the splits in mid-air on stage like Diamond Dave. That was impressive. It was very impressive.

SPEAKER_04 6:45

Yeah, and doing kung fu with the you know the little staff and yeah. Yeah, amazing entertainment. And you know, the artists there aren't that a lot many artists that do that who don't rehearse those parts. It's just energy, right? They're just energy energy as they're performing, and even Eddie like running around and jumping and doing all this stuff, uh, you know, all these weird you know.

SPEAKER_03 7:12

Oh, remember when he was like like suspended, like had the wires on him, he's flying. Yeah, yeah, yeah. Yeah, that was good stuff, man. That's good stuff. It was as Leonard Nemoy would say, it was pure energy.

SPEAKER_04 7:28

What song was that? They they sampled.

SPEAKER_03 7:31

I wanna know what you're thinking. And the song was they kept playing the clip from Spox and Pure Energy.

SPEAKER_01 7:39

Yeah.

SPEAKER_03 7:41

Uh this is a music show, and sometimes we talk about IoT and tech and things like physical AI.

SPEAKER_04 7:48

Oh my god. Yeah, we got um, you know, uh our fill of well, not really a fill of it, but we brought the physical AI message and grounded it in sensors at sensors converge, right?

SPEAKER_03 8:01

And uh yeah, we hung out. Tell me, tell me all about it. I want to know all about sensor converge, man. That was a good show. Um it was good convention.

SPEAKER_04 8:13

Uh I can't believe how much um positive feedback there was on the floor. I mean, the traffic seemed like light lighter than last year, but they had moved the date of the event a month. Uh it's usually in June, but because of like um, I don't know, the World Cup, they moved it to May this year. And uh, you know, I guess what they didn't want happening more was uh you know, the international well, you know, a lot of the attendees are from around the globe, right? Yeah, right. Maybe people don't know that, but people like to converge in Silicon Valley for the semiconductor stuff, or I mean the sensor stuff, uh, because a lot of the customers are there like Apple, uh, like Meta. And so it's like it's an interesting conference because this is one of the few conferences where Apple is very visible. Apple never shows up to anything, yeah. And and uh and Qualcomm really makes a huge presence there as well because uh you know uh they're all they're competing with uh Qualcomm. I mean uh Apple when it comes to technologies bringing a lot of that mobile stuff, you know, that to the mobile world. But yeah, it it's it's Ruby. Yeah, Ribbian, you know, the the LiDAR, you know, all the sensors for the autonomous vehicles, which basically crashed because I don't know if Luminar company Luminar, the LIDAR company that some people were tatting was gonna be the next big thing because autonomous vehicles were going to be everywhere.

SPEAKER_03 9:55

They are everywhere. No, not really.

SPEAKER_04 9:58

Uh they are there.

SPEAKER_03 10:01

I was in Scottsdale and I did two different rides. I know, yeah. Soft. It was awesome. I didn't die. My wife didn't die. Oh it was great, it was great. That's the future. And you know what? Lots of people, when they think about physical AI, which is like a wrapper or whatever, it's a lot of a lot of reusing, a lot of the same IoT stuff. I get it. Uh, and but uh gosh, everybody's talking about physical AI and robots all the time and embodied intelligence, and and I get that, but I really feel like what's going on with Waymo self-driving cars, nothing compares to that. That is the epitome of physical AI and edge AI on wheels.

SPEAKER_02 10:45

That is sort of a perfect storm of tech.

SPEAKER_03 10:48

It is, it is, yeah, and we lost him. Yeah, yeah, I think I think the censors heard him and said, well, we don't want to hear about this. You know, we gotta kick this guy out. We gotta kick this guy out. Yeah, no, uh, you know, the uh it wasn't that long ago, and I think Tesla was the same way with their full self-driving, you know, it used to be rules-based stuff. And the problem with rules-based, it makes sense early on because that's kind of like humans, and we've been doing rules forever. But after a while, trying to figure out do rules for every edge case and every everything, it it you're chasing new rules that you need to apply to how your car operates in space. And so, you know, it wasn't that long ago that Google had Waymo switch, you know, to really using AI and you know, doing machine learning and really learning, and the fact that they trained Waymo cars on digital twins, yeah, using a world model digital twins of the planet Earth and trained it because that's the best way to train something is with digital twins. Yeah, like in the matrix, learning. Yeah, like in the yeah, yeah. And so, you know, the when when they when Google switched from a rules engine-based thing to digital twins and true AI for all the nuances of the real world, and it's like, yeah, these Waymo vehicles have been trained, they've done one billion trips already, you know, in this digital world model, uh, uh, which is way more than they'll ever do in the real world. Um, and then the AI, but it but you know how like I'll talk about how like I'll do some IoT stuff where it's always a combination of deterministic and probabilistic, yeah. Because because there's things happen in the most no, no, I need to check it out. Um, but things happen in the real world that you can't account for, you know? And um, and so it's like like for instance, when you think of the Waymo car, and it's got a zillion sensors on it, and I know Elon makes fun of it because the Tesla doesn't have anything coming off of it, and Elon believes that cameras alone will be fine. Um, I feel like Waymo is like over-engineered for safety and overthinking, but it works, and they're actually in the lead by a wide margin, even if it looks silly with all the things spinning all over the place. But the part that can't just be total AI is weird shit happens. A pedestrian walks right out in front of your car. You know, cars get into your lane, all kinds of crazy stuff happens, and um, and you need to be able to react to that. And so it's always got to be a combination of, yeah, I've been trained a billion times on a digital twin world model, but I also need to take care of real life as well, because it's messy and you gotta deal with it.

SPEAKER_02 13:46

And so uh the newer by the way, the newer Waymo's look a lot cooler. I mean, the yeah, the current Waymos are just Jaguars that are retrograde. But the newer Waymos that are designed that I've seen a few of them, they're much sleeker.

SPEAKER_03 14:01

Because they had the they had a they had a big booth at CES.

SPEAKER_02 14:04

I mean CES, they showed some new Waymo's CES, yeah. So they'll they'll get them out there. But I agree, I think uh you know Waymo is way ahead and uh um you know, accommodating kind of and doing kind of the ultra low latency intelligent uh AJI. Uh it's really important, right? Because that's they they seem to have nailed it. And uh yeah, once you take the Waymo's though, like the third or fourth time, it becomes sort of uh blasé, you know, sort of just getting in and like okay, get me where I want to go.

SPEAKER_03 14:33

So my my wife Kathy is like, and they're playing like that music I hear when I'm at a spa. You can change the music.

SPEAKER_02 14:40

That's the the killer feature is you go in and you play your own music on uh Waymo. Absolutely.

SPEAKER_03 14:49

That's a good point. Um it's it is it's way more expensive though than doing an Uber, and maybe that's just because I saw in San Francisco it's cheaper.

SPEAKER_02 14:59

It was cheaper than an Uber.

SPEAKER_03 15:01

I'm going from I'm in Scottsdale, South Scottsdale Road, going up to north part of Scottsdale Road. So let's say maybe seven or eight miles on a straight road to go to dinner up in the Kieran thing in North Scottsdale, just a straight road, and it was like $43.

SPEAKER_02 15:19

Yay, and I'm and I was like, and I was like, maybe it's prototype pricing or something, but I don't know.

SPEAKER_03 15:24

That's what I was wondering. You know how like every tech, the early days, it's the most expensive version that it'll ever be, is the early versions of it, and they'll get cheaper over time. So I didn't know. But you're saying it was cheaper for you when you wrote it.

SPEAKER_02 15:36

Yeah, and for if you use the Waymo app in San Francisco, um, I've looked at it and it was it was you know a few bucks cheaper than Uber, a few bucks cheaper.

SPEAKER_03 15:44

That's great.

SPEAKER_02 15:45

That's great. But it was fine. The cost per mile, obviously, is a lot more because the cars cost five hundred thousand dollars.

SPEAKER_03 15:51

So Martin's got so much gear all over it, man.

SPEAKER_02 15:55

So but they have to figure out the cost per cost from cost per mile.

SPEAKER_04 15:59

Yeah, yeah. I guess like at the show, there's been a shift because automotive is not as big of a deal as it was maybe three, four years ago. Um, yeah, definitely a shift toward whatever humanoid stuff. But I think the star of the show in terms of like uh spotlight device was the smart glasses.

unknown 16:21

Right.

SPEAKER_04 16:21

A lot of glass, a lot of airbags. Yeah, micro really leaned in on on that stuff, TDK, TDK's uh um CTO cassette tapes. Yeah, exactly. Exactly. I I used to love TDK uh cassette tapes. Those are the ones on the case.

SPEAKER_03 16:39

TDK versus Maxel.

unknown 16:41

Exactly.

SPEAKER_02 16:42

Yeah, you know, we actually talked to them about that. They had that uh this these are some of the smart classes, by the way. Sentient, sentient glasses. But um, yeah, TDK's got I talked to their marketing director because they're one of our uh strategic partners for Edge AI, and it was like, you know, you guys are doing some really cool stuff in Vencents and all that, but uh you still got the the the cassette tape thing going on. In fact, they did a deal with Stranger Things, she said, where there was a t-shirt, a TDK t-shirt. Oh, some sort of like brand deal with Stranger Things. Like the Stranger Things came to them and said, Hey, can we you know show this t-shirt? Where then they started reselling the t-shirt as like merch.

SPEAKER_03 17:19

I love it. I love it. It's a little bit there. But they're doing it.

SPEAKER_02 17:25

And they've also acquired a number of companies. Even sense, they acquired a a wearables um kind of glasses frame company, and so they're really going into the sensor space heavily. But Sintient was there in TDK. Doug Leonard was saying a lot of wearables, a lot of med tech, too. I saw a lot of med tech, a lot of biosensing, muscle sensing, uh, you know, things like that. So uh I only saw one um one booth that had a couple of goofy, like kind of dog robots and stuff, but you know yeah, they were just trying to sell dog robots.

SPEAKER_04 17:58

But uh, but they're all they're all yeah, they're all remote control, and then they have program moves. So when you see these these robots, there there's very little a AI in orchestrating or uh street fighters their behavior, yeah, uh or even perception. It's basically like Tekken.

SPEAKER_02 18:21

Tekken, yeah, that's it.

SPEAKER_04 18:23

Yeah. Or Street Fighter.

SPEAKER_01 18:25

Oh, you a Green Fighter. Yeah.

SPEAKER_04 18:29

You know, it's but you know, the the joke is people think it's autonomous or it's exhibiting some sort of human behavior. It's all programmed. It's just smoke and mirrors, you know, even the stuff that uh industrial, a lot of the industrial stuff, um anything that's humanoid, uh, most of it is just remote control.

SPEAKER_03 18:52

Which is kind of cool, you know? Does anybody remember freakasoid robots?

SPEAKER_04 19:00

I isn't that like a song.

SPEAKER_03 19:03

Freakazoid robots like enter the dance floor, kind of never I mean, I am I the only one that knows anything about music history. Midnight Star. Yeah, it's it's midnight star. Midnight stars. Yeah, midnight stars. Absolutely. That's good stuff, man. I love it. Yeah, I I would do the music beatbox style right now, but it would be embarrassing.

SPEAKER_04 19:25

I I I had to learn that song because I was in a and then you know it's funny because sometimes you don't have a keyboard or you have extra parts, and I'd have to figure out the guitar parts, and I'd make it make it kind of like heavy metal.

SPEAKER_03 19:40

Yeah, yeah. And we had to program into the Waymo cars, no parking on the dance floor. Yeah, there you go. I'm gonna give you a ticket. Oh my god, we're so bad. Yes. Oh, so another news. Um you know how I've uh you know, I I I've used our friend Claude a lot and written a lot of code, and it's helping me with some use cases and what gear to get when I talk about that sustainable development goal book, you know, it's like handbook like piecing together the bomb or how to do every use case. Uh anyway, but I thought I would go see how good Gemini is with writing code. And so I've spent some time this week kind of creating IoT, you know, lightweight IoT digital twin thing with uh with Postgres and stuff. And uh yeah, it's pretty good. It's not the same and it thinks differently, you know? Interesting. And so it's it's it's interesting to just see what it does differently from what I what Claude does, you know. You you're still achieving goals, but yeah, differently. Yeah, it's been kind of it's kind of fun. It it doesn't do things as completely and crank out here's the file, and here's the whole project for Visual Studio or whatever. Um, you're still it's still just showing you on the screen, you're copying, pasting, and putting code into stuff like that.

SPEAKER_04 21:09

But uh but it's it might actually be better, right? Um it's pretty good. It's pretty good actually. You can you can you can deal with each step of the code code you know decoding versus having some something just like spit something out and then you have to you know pack everything and validate. Yes, right, yeah, yes.

SPEAKER_03 21:31

So I was doing some complicated thing, and I've done this in the past with a digital twin system I built where you know our telemetry data for a particular device, I I precreate what the digital twins schema looks like, right? You know, the properties, data types, all that shit. Um, and then the device is sending in telemetry, and then it goes into a message queue, and then you're pulling stuff out, and you have to do some kind of transformation that says, oh, device one, two, three maps to this type, you know. This uh digital twin model, and then I would loop through a background process and write a bunch of code to convert what this JSON that's coming from the device and put it into a strongly typed uh table in the database, you know, so it's easy to query and stuff like that, because I'm a fan of easy. Um and so uh and so I've done that a lot, and it's a lot harder to do that and everything. And and sometimes mistakes can pop along when things aren't perfect and data's flowing in, because right, data's not always perfect, right? Um and so and so, and then so with you know, in Gemini, it's it's in love with the uh I'm this is so technical. I'm sorry, audience. Um like in Postgres, you know, different databases, like so you know, you have all the NoSQL databases that got really hot, and then all the relational databases had to like go, oh crap, we better have JSON support too, and we better support all this other stuff. And so uh yeah, you have JSON support in everybody's deal. But one of the things that uh that Postgres uses is something called JSON B. It's a binary representation of your JSON data. Um yeah, yeah. So it's so it's smaller and it's way faster, and you and you query it. And so instead of me dynamically creating a new table with the columns and data types and everything for the to represent the digital twin from the data, um it's not even trying to do that. It's just like, yep, here it is, and it's just this JSON B column. And yeah, and it's not like we've never done that where we put a whole world of something in one cell, one column, you know, one row. Um, but it's it yeah, yeah, it's like a blob. Yeah, it is, it's a blob. And it but it's interesting, and it and it and it actually works pretty well. And I'm sitting there going, well, will this scale? Can we, you know, and you're having I don't I don't know about you guys. I I think I know about you guys. I mean, I think we're all different. But I mean, I I I use AI heavily every day. We're all we're all individuals. No, you know, I just mean that we're I know that we live in a bubble that the average person doesn't live in. And so, you know, I find I use AI heavily every day. And I know that maybe we we might be in the 1% of people who do that. I don't know. But it's interesting. Anyway, there you go. So try out Gemini for coding. It might be interesting.

SPEAKER_04 24:20

There you go.

SPEAKER_00 24:21

Great endorsement. I mean, I can tell you, I can tell you this. Um, first and foremost, I think the whole prompt engineering thing and everything like that. I think it's cool if you really want to learn how these things work, but I think at least from our side, I mean, there's a distinct point of view, right? I mean, so like when we're talking twins and everything like that, like we would this whole week we've been having this huge back and forth internally about the state of twins and 3D versus operational and all of these other things. And the code that represents it because of the stack, right? Uh so you can take a look at those like, you know, I mean, I I won't say names, but there are the those that do the 3D rendering, okay? Sure. Um and and for that, they're you know, a lot a lot of times they're using WebGL. And for those that don't know, just Google WebGL. I mean, really, it's just a uh a way of of browser rendering your environment. Um but there's a lot of scaffolding that's that's underneath all of that. And and so when you're when you when you do a comparison of a person writing that code in in context to what they're looking for as an output versus having that conversation with Claude or any AI, um the way that I approach it, so we like I said, we had this conversation. The way I approach it is in a project, right? Co-work in a project, and I lay out the frame. Here's here's everything, and here's but everything is governed by a mark tag. A what? Mark Dank? Yeah, an MD file. Oh, okay. All right, um because I can edit those, but it becomes the contextual single point of truth. So now I have a referenceable point to go back to whenever any hallucinations take place or anything else starts to drift or anything like that. I can I can catch that. Yeah. Um, so so it I've kind of left the the the point of prompt engineering and landed on a series of markdowns that create the foundation by which the output is driven.

SPEAKER_04 26:50

So they they call that, yeah. There's people that are calling that context engineering.

SPEAKER_00 26:56

I mean, they're gonna they're gonna give it a name, you know. Yeah, I know. I you know, just I mean, I'm like give a point. Give it a name. I don't care.

SPEAKER_04 27:05

I mean, yeah, I'm just saying I'm just saying I'm I'm helping you associate some of the new buzzwords and uh new job roles that uh the prompt engineering folks are assigning themselves.

SPEAKER_00 27:21

But uh and see that's that's the weird part. That's probably why I don't get called to do keynotes and everything like that, because I'm gonna say I could give less than two fucks about you know what name you're calling it. Yeah, because it's a distinct point of view, and your markdown doesn't look like my markdown.

SPEAKER_04 27:36

So yeah, yeah, yeah, exactly. And they're all all just bit becomes like goofy. It's like these SEO guys trying to all of a sudden become like context specialists, you know. And the problem for them though is that you have to become a domain expert. You actually need to know what kind of context to feed and where the gaps are. And you know, the funny thing is you see all these context engineer guys talking about, well, you know, you're missing context, you need to have comprehensive context. The thing is, is you never have the full context.

SPEAKER_00 28:11

No, especially in the world of twins, because that stuff is changing.

SPEAKER_04 28:17

Yeah, and and you know, one of the things that really stood out to me this year at Sensors Converge was uh the reminder that sensors drift just like AI models, and that it drifts from reality over time, they can degrade, they can they have all actually very similar problems to AI, but then you layer AI on top of it, right? And then you compound the problem because your AI can drift and it can dec uh degrade over time and then become disconnected from whatever abstracted reality that it's been trained on. So you uh the you know, as I was talking to engineers and and some of the guys from uh you know Edge Found uh Edge AI Foundation was there, you start to realize how difficult this stuff is. And the guys who are up on stage pontificating about physical AI, you start to realize how little they know about what it takes to make this stuff usable, and then long term, what the cost is to actually maintain good. Because I and and so just tremendous props to the the engineers and the technical folks who actually deal with this stuff that probably have to painfully watch these bozos get up on stage and talk about exponential bullshit and sit there.

SPEAKER_02 29:51

14.8 trillion to go 14.8 trillion.

SPEAKER_04 29:54

Wow, what the fuck are you talking about? Right? What 14? Oh, yeah. That was yeah, um, so one of the keynote speakers won't call his name, but he put out like 14, what is it, 14.8 billion as a number? Do you think it's a trillion audience or trillion? What is this number? And he's asking the audience to guess, and it went cyber criminal economy. The problem is it it was recorded, right? But guess what it was? The 14.8 trillion dollars. That was the combined net worth. No, no. Actually, apparently not worth as much as everyone thinks, but which is kind of cool because that's a you know a humble statistic is associated to somebody who's quite large and exaggerated. But um it was the combined, what was it? Uh what was it?

SPEAKER_02 30:58

It was the valuation of the top seven, the next seven.

SPEAKER_04 31:02

And then, and then you know, associating that to potential economic value. No, you don't do that. Valuation, stock market valuation is not the economy. It's like the number one uh misconception. Anyone does that up on stage in my book, you've lost you. I'm not listening to you anymore. Uh, but yeah, it was pretty funny.

SPEAKER_02 31:27

Um that wasn't my keynote, though. That was someone else suggested.

SPEAKER_04 31:30

Yeah, and but but just to clarify, it's not the only thing that does it. You know what? It's a poison, and um it and it's a fallacy that's being picked out by uh Silicon Valley. There's so many it's a Silicon Valley talk track that valuation somehow equals economic growth. But then you when you look at some of these charts, GDP growing at what a Kaggar of like we're talking about global GDP, like 4%. And then, you know, technology is growing at what the valuation is way past that. So how are you gonna get a return off of an optimistic 4% global growth? But a lot of that's gonna be driven by the uh developing economies, not the mature economies, right? Mature economies are growing at like you know one, two percent. And the developing economies don't need AI to drive growth, right? Because they're going to take on all of the shit stuff that the developed countries or the advanced economies don't want to do, you know, the stuff that everyone complained about globalism. So none of this stuff makes any sense. You know, and you know, when you do uh the forecasting, Rob, you probably know you have to do uh you know what's called boundary testing. And if your the boundaries that you're assuming are just way exaggerated and wacky, or you know, you you have yeah, you have nothing. You're just at that point, you know.

SPEAKER_03 33:14

It's like iterating through an array or you know, whatever, and you have to do bounds checking, otherwise you can have an overflow.

SPEAKER_00 33:20

Yeah, but I mean at the end of the day, there's no one that's able to predict anything. I mean, when you get to three years out, you're already off. But going back to your point, Leonard, about let's just talk about sensor drift and hardware, you know, that's why when you look at tech specs today and in the past, they always said plus or minus, equivalent to or less than. I mean, they always had a spec because the performance is going to fall off. And if you start to monitor those things and show them over a period of time, you will see that degradation take place. But but all of this, how is any of this any different than technology advancements that we've had in the past? Sure, this is probably the most that we've seen where people without a tech background can actually jump in and do something technical or appearance that of something technical. Um, exactly. Um I mean, but but but those that are operating, okay. So, like I don't even like the term art of the possible because it's pop anything, I mean, hell, anything's possible. Yeah, if you put the pieces together. So let's let's talk about the art of the outcome deliverable. Yeah, all right. Let's talk about that. Let's talk about that. That is an art because if you can actually build something that drives an outcome and deliver on that at scale, now we got something to talk about. And oh, by the way, when you do that, that that valuation piece that you're talking about will expose itself.

SPEAKER_04 34:53

Yeah, that is kind of weird that you say that you you don't get you should be invited up on stage all the time. You know what I'm saying? It's like it's just people are scared to bring you up because you just tell it like it is. It's like you know, it's like basically, hey people, um, you see this guy, you see this guy, you see this company, this company. Well, they're wasting your time. I mean with a bunch of marketing nonsense, you know.

SPEAKER_00 35:21

It's it's it's but but you know, I get it and to some degree because people are like, you know, come on, you have to be somewhat professional. It's like I'm totally professional. Oh, yeah. The fact that you don't like the language that I use, I I I can walk out anywhere on the streets right there and talk to somebody, and we're we're exchanging ideas using that language. Yeah, I mean, if you don't like that end of it, I get it. I mean, you know, that part of that doesn't change the fact that everything that I've done, I'm not gonna say something that is as suspect, except for they didn't land on the moon. But I mean, we're not gonna say Rob. I'm not gonna say that's an opinion, right? That's an opinion.

SPEAKER_02 36:02

You know, the point is we're all with you right up until that point, Bill. We were right.

SPEAKER_00 36:09

Now we're bearing on. I have to remind people that I will do that. I love the debate myself, but but when we're talking about literally, you know, the businesses and how I've built the businesses that that I have now and what we're exactly. What happened?

SPEAKER_02 36:30

You got a little glitch.

SPEAKER_04 36:31

Yeah, a little glitch in the matrix still now. You had a glitch in the matrix. Oh, it was trying to shut me down.

SPEAKER_00 36:41

Yeah, it was trying to try to it was trying to censor me. You were being censored. I mean the the the things that the things that I've built, the companies and everything. I mean, all the people that work with me. Um, I mean, it everything's on the table. It's not my way or anyway. I learn from them on a daily basis. Yeah, I'll tell them, here's what I think, and this is why, and they'll be like, No, you're wrong. Here, here's what it is, and here's the data that supports it. Okay, cool. Change my perspective, right? Then on the on the on the fly, and we'll credit them for correcting me.

SPEAKER_04 37:16

Oh, yeah, oh yeah.

SPEAKER_00 37:17

I mean, I I've yeah, everything, all of this is still a conversation, right? I mean, all of the you know, when the thing that I don't like is the is the BS behind it when people are like, oh yeah, we're we're edge-based, and then we start talking about their architecture, and it's like, okay, so how do your messages get from here to there? Oh, well, back over here, oh wait a minute, you're sending them back to the center cloud from the center cloud back down, so you're still doing a round trip. Yeah, your edge-to-edge communication is not really edge to edge, it's edge cloud to edge.

SPEAKER_04 37:51

Yeah, yeah. And you know, that was one of the discussions that was um at least um uh quite frequent at sensors converge where we're talking about how edge AI, you know, how there's an AI AIOT camp, right? It's like IoT with AI sprinkled on top, right? Then you have edge AI. Edge AI is actually a different beast. It's almost like you know what we talked about, Pete. It's a counterpoint to IoT. It makes you think differently. Like, so everything that you were uh talking about earlier, Rob, how you're dealing with streaming data and then you're you know aggregating it somewhere and then using a digital twin or some kind of ontology to normalize it and uh give it meaning, that flits. That whole model is uh you know put on its head because edge AI actually is about it is really about how do you force everything as close to the sensor as possible and then have that translation you're talking about, Rob, uh done very locally. And the only thing that you're really trying to transmit is context and meaning, whether it's through metadata or some other kind of format. But you know, it's not to say that you're going to have things shipped one way or the other, it's a force. Do you know what I'm saying? It's got a lot of power. Yeah, exactly. And and and it is with you. And the force is strong in this. Yeah, exactly. Force is strong freaking Yoda.

SPEAKER_03 39:32

You know what? You know what? I keep hearing things from folks here that reminds me, and I know, Leonard, you do posts about this where you're like, people still don't know what IoT is or whatever. Um, and and Bill's like, you don't know what the edge is. I mean, you were just in a in it, it was like the idea, like, we're gonna have to have a conference that says that's called like you don't know what you're talking about, and we're gonna learn you something because I'm surrounded by freaking idiots. Oh my god. Oh my god, let's just talk about digital twins and people not knowing what they are, because I see it on the daily, okay, at least weekly. Now some of them will write they're writing about digital twins, and I catch them, I catch them in that Dassault systems digital modeling thing and not the real digital twin thing. And they go, Yeah, these are great, man. If only there was a way to, if a digital twin, if you could like kind of make it alive and have real data, then you would actually have something. I'm like, I'm like, I got the record play. I'm like, I'm like, I'm like, bro, a digital twin is alive. It's alive. And it's not a digital model that you created in Katiya. It's alive. Physical twin, digital twin, data's flowing, and data's going, and it's alive, and you can look at it. And there's latest state, there's historical state, and there's AI models being created from that. It's freaking alive. And edge, and you said, and I love the idea of peer edges. I remember writing shit about doing that too. You know, people are like, oh, I'm just doing this whole round trip. I'm like, you know, it's it's bonkers. A IoT, yes. I remember when people started talking about that. We know exactly who you people are. It was because AIT has failed, IoT has failed, so we're gonna put an A in front of it. We're saved, we're saved, hallelujah. Um, it's like I may not get to the AIOT promised land with you, but I know you're gonna make it, you know.

SPEAKER_00 41:45

I mean, but but here's the thing, you know, it's like it again, a lot of the stuff that we do under the hood is not pretty.

SPEAKER_01 41:54

Yeah, yeah.

SPEAKER_00 41:55

It really isn't. It's it's truly understanding the numbers, and it's not it's not following any prescribed rules. If you don't develop your own point of view about this stuff and then move to that point of delivery and scale, why are we even having this conversation about valuation?

SPEAKER_04 42:18

Yeah, exactly. And you know what, like, even with all this edge AI stuff, all the you know, like uh Pete and I uh we we cut a little video and we mentioned this. This stuff is like difficult. It it's really fucking hard. I and again, hats off to the guys who are deep into tech, right? And so I'm not gonna go around and pontificate how edge AI is the greatest thing in the world. It's gonna it's gonna be a force. But the engineering, the technology, all this stuff, it's gonna be difficult to translate this stuff into actual reliable, you know, um, valuable applications, just like IoT, you know. You and Rob, you're absolutely right, man. Um people still don't get digital twin. It's amazing.

SPEAKER_03 43:16

Now they're calling it synthetic data and all this other shit, and it's like or the people, the people who think that digital twins is only about simulation.

SPEAKER_00 43:25

Well, so here's the thing. Yeah, yeah, yeah, yeah. Let's let's let's look at let's look at let's look at you know some of the movements in the market. Just period, right? Bentley systems acquires sessium, right? They acquire Cessium, and then they come out with a new product because they're merging the two, right? So you got you got the BIM model stuff with the Cessium Web GL front end, and we're gonna, you know, and now we're gonna start grooming in operational data. You know what? My word to any own. Anyone that is doing those type of things, good on you. Go for it. And they're like, well, aren't you afraid? You know, that's a threat. I'm like, do you understand how big this space is? I mean, it's huge. And and you, and you've got, you know, you got all of these other frameworks and omniverse and this and that. You got all these kinds of things. If that's if you want to go that route and go, we're we've selected, we've landed on this framework. This is what we're going to do. These are the things we're going to be doing. We're going to groom in operational data. You're going to see these things. Good. Go do that and scale. Because at the end of the day, the technology is the enabler, the business is the driver of the economics that you're going to use to try to get to some level of valuation where either you go public and the people buy in even more, or you get acquired and try to iterate on that and grow even bigger. And so the conversation is more about the application of the tech, your point of view, and your delivery mechanism.

SPEAKER_04 45:02

Yeah. And you know what? Um, you're right. And the reason why I think the IoT hype collapsed the first time is everyone thought that that was like a hyperscale opportunity. And you're going, you know, they saw these massive TAMs, and people who threw out these gigantic numbers had no friggin' idea what they're talking about. And the bottom line is this. I had a chat with Andy Doe. Uh he's um with Embedded Works. He's in the Gap of the Ugly just with you, except more on the IoT front. Um, you know, and he did some stuff for SoFi Stadium. They uh put some sensors on 400 manhole covers, right? Okay, and I'm like thinking, he's telling me the story, I'm going, man, it it reminded me of uh our episode, um Rob, when you and I talked about IoT plumbers, right? Yeah, this business of quote unquote IoT is it's a trade business. It's not some guy sitting up in a uh, you know, it's like what Ian. You hang out with Ian, you're a RT, right? Yeah, does he sound like he's like some coding dude that just w writes software, maintains software, you know, in some ivory tour? We're out there with the welders, man. Yeah, it's a fucking trade business, right? You have to buy a truck and you gotta roll that truck out to get on site, be in the friggin' hops field like you were, right?

SPEAKER_03 46:33

Watch the videos or the pictures every week from our friends at Caliper in Australia, and they'll show you what real IoT is. Every week they've got something, they're in some manhole or in some creek somewhere or next to a railroad, and they make it real. And it's and it it is. You're out there, and uh it's a trade business.

SPEAKER_02 46:53

It's there's another uh company called Caliper, by the way. If you want to talk about deploying things in the real world, they do like sewer sensors and all kinds of stuff in Australia, and they're on the you know, telemetry on rail rail temperature and stuff. That's amazing.

SPEAKER_00 47:08

Well, you know what they have to do.

SPEAKER_03 47:10

Is that that's that's the company you were just talking about? That's what the company I was just saying. You're right. Well, you said caliper I did say caliper, yeah.

SPEAKER_04 47:20

Yeah, for listening. That's what happens when you don't pay attention to it. Well, there was a glitch in the matrix. It's all right.

SPEAKER_02 47:27

Yeah, by the way, there's another company called Caliper. I don't know if you've heard of that.

SPEAKER_00 47:33

This guy. Yeah, but you know what? I I do think you were right. Uh, you know, I mean, all all all well, a few jokes aside, I do think you were right. We should do a conference. Uh, and and it should be, I mean, I not like these other ones. Look, come in and listen to us, yeah, yeah. Come in and engage. Let's have those conversations. Yes.

SPEAKER_03 47:55

What would the conference be called? Like, you don't know shit, they not like us. We're gonna learn you something. I'm sorry.

SPEAKER_04 48:05

Like, we're gonna learn you something.

SPEAKER_00 48:07

I just just call it bring your tool, Bill.

SPEAKER_03 48:10

Yeah, there you go. Yeah, yeah. Well, you know, well, it's like, you know, like Walker Reynolds has done those prove it conferences the last couple of years in Dallas, you know, where the industrial IT guys have to show up and show it really working and solving a real problem at the conference. And everybody get everybody gets put on the spot in a real big way because he was all like pissed off at conferences because you got all these booths and it's all marketing. And he's like, We need to have a conference where you prove it. Yeah, prove me, prove to me that your product does, and he assigns them a task, and you have to show up or you can't be there. Yeah, it's crazy.

SPEAKER_04 48:47

Yeah, well, it'd be great if he came on.

SPEAKER_03 48:49

I mean, you know, we had a couple of conversations about um him showing up, and he has he, you know, he has like a yeah, but it'd be cool to have on just uh yeah, I just think it's it was a great idea for conference, and I think that's kind of what we need is we need to, you're obviously wrong, everyone's wrong, as it turns out, and we're the only ones that are right.

SPEAKER_00 49:14

Yeah, yeah, that's a great way to get people to come to your yeah, no, I mean look, my whole thing is is having creating that community of conversations. I mean, realistically, you know, I mean I really like talking to people if they well, we're we're edge-based, and anyways, well, I can't divulge that to you right now because that's our secret sauce. You don't have no secret sauce, you don't have no sauce.

SPEAKER_03 49:40

You know, the only sauce you got is the sauce you got from Kane's at lunch today for your chicken. That's all the sauce you got.

SPEAKER_00 49:47

We're just making sauce. I mean, we we will I will open up the I will open up the code and show you how we do what we do. I have no problem with that.

SPEAKER_04 49:58

Well, yeah, the only sauce that you have it ends up being experience is you solve the problem and you have a pattern for accelerating it for any other scenario, right? Recognizing that every situation is different. That's like uh you know, that's systems integration 101. That's engineering in the field 101, right? And you know, that's the only secret sauce that you can have, but it's usually like I don't think it's it's technology.

SPEAKER_00 50:25

No, our secret, our secret sauce is the people. We've been doing this for over 10 years, and the experience, you know. Yeah, it's it's it's the people. I mean, and and and you've got to be crazy.

SPEAKER_04 50:38

It's a friggin' trade, man. And there's nothing wrong with it because you know, these h commercial HVAC guys make a shit ton of money, man. They make so much money, it's like crazy.

SPEAKER_02 50:48

Especially in the summertime.

SPEAKER_04 50:50

Oh, yeah. Exactly. Summertime. Locksmiths.

SPEAKER_03 50:57

Locksmiths make a shit ton of money, man, because you got and think about all the use cases you're talking about here when we talk about these things, they can all be IoT eyed.

SPEAKER_00 51:11

That don't change the fact that you're gonna call them when that IoT part of it kind of fails you.

SPEAKER_04 51:16

Yeah, like what's happening in my house right now? I don't think we have like IoT connected lights and they're all like friggin', you know, yeah, it's easy to operate.

SPEAKER_03 51:24

Going back in time, like whenever someone says, What is this IoT all about? You know, and I'm like, Do you ever watch the weather on the news at night? And they go, our weather station at the CTAC airport says it was 75 today. I go, Do you know what that was? It was a sensor for temperature at a location, and it was wirelessly transmitted back to him place where the weatherman can put something on a screen and say, it was 75 today in Seattle at the CTAC airport. That's it. Yeah, that's it. It's sensors, and we're sending the data to somewhere else and maybe converging a whole bunch of stuff together, and that's what it is, you know?

SPEAKER_00 52:05

Yeah, Leonard, you mentioned your your lights. You you know what you might want to consider home assistant. Home assistant. Yeah, you you like to you like to tinker around, use home assistant, and building a open source. That is like, you know, I mean you can get a Proxmox box and yeah, yeah, yeah.

SPEAKER_04 52:28

You know what? I I I bumped into that. Yeah, okay.

SPEAKER_00 52:32

Can you put that on a Raspberry Pi? Oh, I mean you could, you could. You could, I mean, I I I run I run all my stuff through Prox Proxmox. Proxmox.

SPEAKER_04 52:45

Well, you can run all kinds of crazy stuff on uh on a Raspberry Pi these days, right? So yeah, but you still gotta harden it, man. That's the see that's the real thing. Can you press up and down the stack? Yeah, and then you know, think about it. That's gonna be a really valuable thing if you can you can deliver hardened secured um stacks, right? From hardware to the platform to all the endpoints that connect to it. Um because you know, man, I I I I just don't see how people are gonna be able to manage securely all these agents, right? It's it's insane.

SPEAKER_00 53:27

Um when you say when you say all these agents, what are these agents doing?

SPEAKER_04 53:32

You know, doing all the hypothetical shit that all these AI companies are talking about. You know, going off and doing research or doing, you know, buying shit for you and doing all kinds of stuff, whatever. You know what I'm saying?

SPEAKER_00 53:46

It it's it the the the bigger part about it is putting the rigor around it to actually understand that the workflows that you have in there, number one, aren't drifting. Number two, have guardrails around them inside of the of the of the containers that have guardrails around them.

SPEAKER_04 54:02

I mean, you you've got a super No, no, and you know, like I said, you know, Paul Cunningham, CTO of Cadence, he got up on stage when he's talking about Cadence, uh, their um, you know, uh their whole super agent framework, right? He goes, We'll uh um agentic AI will require millions of lines of code. I don't know if they uh it just he got he slipped that through, but I took a picture of it and then later on I gave him a high fives like, yeah, you get it, man. Yeah, you're beautiful. Crap you have to stick on top of agents in order to make them use like usable. The the I mean, and that's the that's why I struggle with um these TAMs, the investment that's going in is you see guy companies like Cadence with amazing engineers trying to use this for you know semiconductor design automation, blah blah blah, right? You know, auto verification, and you talk to the the people who are actually doing the work, and they will tell you a very different story. They say that this stuff is not usable, or we have to use it in a very measured way, and then it just looks like scripting. So, what we're getting excited about at the end of the day is like friggin' RPA. Yes, millions of lines of code. What does that tell you? You are stripping, like we've said before, you are stripping agency away from the LLM, and at that point, what do you have? Because you want it to be an LM, no? Yeah, you just want it to be deterministic, you don't want it to go off and do whatever the hell.

SPEAKER_00 55:51

Yeah, I mean it's like hilarious, it's a joke. It it like I said, I think it's gonna take years for this stuff to get to a maturity level where it should be trusted. I mean, it's why it's uh it's it's why what I mean again, our point of view and how we're employing this is very specific and and gated by existing or in works about best practices and standards, which is gonna limit the art of the possible. Yeah, um, that is that I mean at that point we're like, nope, we we're I mean, we we could entertain that, but it's gonna take us a while to even get there. And why would we want to do that? I mean these things that just pop up. It's like we're here to solve this problem and we know what it is, and we've articulated it and we've documented it and we've said this is what the outcome wants to be. That's what we're working on, you know.

SPEAKER_04 56:47

And by the way, it's there there's probably some people who are watching this going, man, these guys are horrible for go-to-market support. And we're actually we're not. If you have a we're not full of shit marketing and go-to-market um, you know, strategy, this content is perfect for that kind of stuff. If you're full of shit, you're gonna find everything that we've just said super uncomfortable.

SPEAKER_03 57:15

That's right. Because we don't stop till you get enough. Yeah. Yeah. That's who we are. Okay, because of course it's got a lot of power.

SPEAKER_01 57:24

Yeah.

SPEAKER_03 57:25

You know, just to just to just just to go off on a tangent here, Bill said something. He's talking about his Proxmox, you know. I'm curious. I I don't want to call out any names. I do spend a lot of time with a lot of people who are in the virtualization game, and they've all think their virtualization is really special. And most of their virtualization technology is just a wrapper around the KVM that's baked into the Linux kernel. And a lot of them, they're all of them, I'm gonna say almost all of them have the same go-to-market strategy. Oh, ever since Broadcom bought VMware, and now they're charging more and all this kind of stuff, and so we're gonna capture all that value. Why isn't something like an open source thing like Proxmox just running away with that? Or if it can't run away with it, it needs a company or someone to wrap around it to offer a service.

SPEAKER_00 58:17

So what do you think, though? I mean, my assessment of that when I when I started digging into it and saying, you know, hey, I I kind of like the way they structured everything, it's open source, we can do this. Um, it's it's it's a small group of like 12 people in Germany. Okay that that there is, I mean, they would have to really form themselves in such a way that they were able to scale and execute in that space. And and literally at that point, I would say ignore VMware, pay no attention to them at all. And I mean, that would be my advice to them, would be ignore all these big players, play a different game. Stop, don't play their game, play a different game.

SPEAKER_03 59:02

Exactly. You just gotta face that man in the mirror, and you just gotta play a different game, you know.

SPEAKER_04 59:09

What is up with all those Michael Jackson and all these Michael Jackson?

SPEAKER_03 59:13

Because it's the hot movie that's out right now about it could be a real thriller. But it doesn't even exactly, I don't even think it needs to be them. It's open source, they put it out there. Remember, the whole planet survived on open SSL with one guy maintaining that thing.

SPEAKER_00 59:34

Oops, yeah, but but Proxmox is not open SSL. I mean, when you're talking about when you're talking about virtualization, there's various planes that you have to consider, which OpenSSL just kind of said, hey, this is a socket. We're gonna secure it.

SPEAKER_03 59:55

Oh my god. But yeah, Proxmox that works with clusters. I mean, it's pretty impressive for what it does. It's not trying to be the virtualization. KVM is the virtualization, but it needs wrappers to do clusters and lots of VMs and stuff, and it does the LXC containers like you see on Ubuntu. I mean, all of this stuff. That sounds like that sounds like a whole new startup right there that we're just gonna deliver Proxmox open source virtualization and containers and take over the world. And the new GTM is we're gonna grab all that stuff, and we're only gonna charge you like five cents a day or something like that.

SPEAKER_00 1:00:30

Yeah, just that that that that right there. Uh contact Rob if you if you would like to be part of that company. There you go.

SPEAKER_03 1:00:38

But we're gonna get a trillion customers. Yeah, so trillion. That's how it's gonna be up in volume, like the car dealership.

SPEAKER_04 1:00:53

There you go.

SPEAKER_03 1:00:54

Yeah, you go. I don't know, could be something there. All big companies I know of, every one of them uses the city.

SPEAKER_00 1:01:00

Yeah, but it doesn't have AI in it, so they'll they'll never get funding. Yeah, we could do that. If you want to do that bootstrapped and never take a dime from investors, give me a call.

SPEAKER_04 1:01:11

All right, all right. There you go. All right. I think we got something. This will officially be the longest. Wait a minute.

SPEAKER_00 1:01:21

Before we go, can I say one thing that just hurt the hell out of me this week? Okay. Uh oh. Come here. Listen. All you people out there that keep sending these messages talking about how you're gonna help improve sales because of your AI assisted elements. Stop it. Stop it. Stop sending me DMs in my LinkedIn telling me that you've got something and stop cold calling me. Stop it, stop it, stop it. Just stop it. You can't even the minute that I ask you, they oh, I get these emails. Oh, we've looked at your company in true north, blah, blah, blah, blah, blah. And we could help you increase sales by and I'm like, really? What do we do? Yeah, uh, well, uh, Dissy, goodbye. Click. You know, I'm like, I stop it. Just stop.

SPEAKER_03 1:02:15

I know we told you to don't stop till you get enough, but in this case, we want you to stop.

SPEAKER_02 1:02:22

You need to beat it, right? You need to beat it. You need to beat it.

SPEAKER_03 1:02:24

You need to beat it, man. Yeah, exactly. Exactly. You need to beat it. You're right.

SPEAKER_00 1:02:29

That is my PSA for today.

SPEAKER_02 1:02:32

All right.

SPEAKER_00 1:02:32

That's a good PSA. I'm in.

SPEAKER_02 1:02:33

I'm in. All right.

SPEAKER_04 1:02:35

Woohoo! You're gonna take us off.

SPEAKER_03 1:02:39

Bill should take us out.

SPEAKER_04 1:02:40

Yeah, you yeah, you're so angry and uh on fire.

SPEAKER_00 1:02:44

Look, I mean, first off, I was late, so you know, sorry about that. But there's been a whole host of things uh going on, and we want to thank you for hanging out with us this long. Um, I mean, ideally, we always have good stuff to talk about, but we're looking for that engagement. Why don't you guys hit us up? Let us know what you want to hear about, let us know what you want to talk about and come on the show. That would be awesome! Yeah, we want to we want to thank everyone for hanging out with us. We want to bring light to uh to elevate uh what is the new and now elevate communities, elevate communities, yes. I mean it can't yeah, I haven't seen it in a minute. Uh, because it those things are important and we've we've migrated away from just uh just providing laptops and things like that, and we're looking for more engagement there for 100 100. Uh good the edge. I mean, everyone knows who Pete is, and I mean, and and I mean he's a giant when it comes to edge and edge AI and everything like that. I I mean look, we're we're going to get engaged. I'ma tell you that. And and uh we're gonna see what happens there. But you guys have a wonderful weekend. Stay up with us, smash that like button, come on back and see us again next week.

SPEAKER_04 1:04:05

Take it seriously at your own risk.

SPEAKER_00 1:04:10

That's for sure. And ring that bell.

SPEAKER_03 1:04:17

And keep your eye on the man in the mirror. Exactly, exactly, because that man in the mirror is gonna say the IoT coffee talk bros and girls are the number one thing on YouTube. I mean, we're we're aren't we?

SPEAKER_04 1:04:31

Oh yeah. Well, we're probably like the last remaining IoT uh titled show in the YouTube or the hold out.

SPEAKER_02 1:04:39

There you go, you got it now, good. Or the market. Exactly.