SPEAKER_05 1:11

That's pretty good.

SPEAKER_03 1:14

Oh Van Halen, and you don't know what that song is? That's fifty-one fifty. Oh, yeah, you know what? Try to open up a bottle of water in Las Vegas. One of these fancy, fancy bottles. The metal? Really, really fancy. Yeah, I'm like trying to twist the thing off, and then I just I realized that I was cutting my finger in.

SPEAKER_01 1:42

You need a workers' comp. Who are you with? Who are you there to see? Let's see if we can file a workers' comp with them. Cisco.

SPEAKER_03 1:52

Cisco, yeah.

SPEAKER_05 1:54

Well, you should be thankful that you know at least the water bottle was given to you, uh, included in the price of your admission, and you didn't have to pay 30 bucks for like a plastic arrowhead bottle of water somewhere you know inside one of the casino hotels.

SPEAKER_03 2:11

Well, you know, they're um the if Las Vegas doesn't know that they're about to get their ass whooped um because people hate going there now, uh, they will find out soon. Um because yeah, people are getting sick and tired of the price gouging.

SPEAKER_05 2:30

As someone who used to go there quite a bit when I lived in SoCal, and I have a bunch of friends that are not going, and I'm reading all the news. I always um it seems that they no longer care for the masses that will spend a couple of hundred bucks here and there, and that they just care for the handful of folks that will spend tens of thousands or hundreds of thousands. Which I I I get that, but then it's gotta look super empty because these places, Vegas was built for the masses. Like if you want to go to C, like it's lively, there's people everywhere, there's things happening everywhere. If you're just doing something for the gazillionaires, you don't need a place that big.

SPEAKER_03 3:16

Yeah, yeah. And you know what? Yeah, um, my driver was telling me on the way back, you know, way back to the airport, was telling me, yeah, there it was better when the gangsters were running Vegas, which I thought was a funny combat.

SPEAKER_04 3:30

That's when he had like the five dollar, you know, I was about to say you can go find a two-dollar table or what or yeah, absolutely, everywhere you could.

SPEAKER_05 3:40

I know the muffins had the same amount of blueberries in them.

SPEAKER_03 3:44

Yeah. Oh, there it is. It's just like, man, where what the hell? I mean, who's gonna fly all the way over to Vegas in that heat to pay $17 for a friggin' latte or now $30 hamburgers?

SPEAKER_01 4:03

You know, so you're just Vegas.

SPEAKER_03 4:06

Self-destructing, yeah.

SPEAKER_01 4:09

It's how you hope that these tech companies would stop choosing Vegas, like go somewhere else, please.

SPEAKER_03 4:17

Just choosing Texas.

SPEAKER_01 4:18

It's not my favorite place.

SPEAKER_03 4:23

There you go. Everyone loves San Diego, right?

SPEAKER_01 4:26

Yeah, you can't.

SPEAKER_05 4:30

Well, you know, uh San Diego got edge AI not too long ago, and now you guys in Texas, you're getting a lot more Samsung people. They shut down some stuff in Jersey.

SPEAKER_01 4:39

Yeah, everything's going back to Plano. Yeah, yeah, yes.

SPEAKER_04 4:43

Let's just call it out, folks. Companies are moving their headquarters from high-tax states to low-tax states. Yeah. Um, I just saw uh what's his name? Brad, whatever his name is, it's president of Microsoft, and he was doing this deal. It's like it took a long time before you could finally get executives for the company to come out and say, okay, Seattle is now officially, you know, a shit show. And Washington State is now a high-tax state, and it used to not be, and that's why we were all here. And now Washington is in the same company as Massachusetts and California, and how do we operate in a high-tax state?

SPEAKER_03 5:23

You know, um, it's yeah, yeah, but then again, yeah, we have to remember the history of Texas and tech. I mean, Dell, uh, Compaq, Texas Instruments, so many companies.

SPEAKER_01 5:38

Uh yeah, like the first job out of first job out before I even like graduated, I had a um someone from the AM engineering school that was my mentor, and she worked um in the telecom corridor, and that was that was a big area in Texas, and then it kind of disbanded. So, but there was such a hub of telecom and tech companies right in that area in Richardson. And yeah, yeah, so it was known for that a long time ago.

SPEAKER_03 6:07

Telecom. Telco.

SPEAKER_04 6:08

Yeah, you know, I mean it's a telecom, yeah, you're right. In the south, and then companies like Ericsson picked up the pieces for pennies on the dollar after they all collapsed.

SPEAKER_03 6:18

Yeah, they have places in Massachusetts too uh too, a lot of um communications like NEP uh companies, you know, Cascade Communications, there were like a bunch of companies, and then there was that great consolidation that happened.

SPEAKER_04 6:33

Um it all began around Boston and Cambridge and everything where it all started, you know, data general, weighing, uh digital, it all started there.

SPEAKER_05 6:47

This whole thing with um now, of course, a business is a business, business exists to make money and or behold it to shareholder value. Like I I get it. Right. Um but you know, uh the days of the of the social contract, they're just over because when you have pick your state, pick your industry. That's I mean we can stick we can stick to state of Washington, but you've had a tech hub there for so long that a community has grown around these employers, uh, and people have made generational decisions based on that. Hey, you know, I used to work at this company, so I should go you should go to Washington State University or you want to stay local because you know, this is what's happening in this area, and families will grow, and you have multi-generational folks inside of companies. And I know it might sound strange in the context of tech, but you know, in the Pacific Northwest, you know, Rob and I, we know a lot of people that are like fifth-generation lumberyard guys, uh farmers, apple growers. We can throw wirehouser, like we all know the family that has worked for that mill, or used well, the mill doesn't exist anymore. But it's it's nuts when they do this and then they move. And of course, you know, it's people are not gonna pick up and move um from place A to place B just like that. And and it's and it's it's sad because from a cost perspective, where you're gonna have to uh hire people, they might not know your company as well, it might take you a bit of ramp-up period. Like, what happens to all that? But we just get stuck in the let's go to the lower tax state, we're gonna lose a bunch of people, but I'm the I'm the I'm the leader taking care of this thing, and my short-term incentive is all I'm looking for because I get a bonus next quarter.

SPEAKER_04 8:43

Yeah. I remember way back when, even, you know, when things were great in Washington State, and ATT was all up in there. Oh my god. Um, and I remember I had so many friends who worked at ATT and Redmond and all around there, and one day they all got an email that says, Your job is now in Dallas.

SPEAKER_01 9:03

Well, you know what, Rob? They were in like they had their headquarters in San Antonio, Texas.

SPEAKER_02 9:09

Yeah.

SPEAKER_01 9:10

For so long, but they had, I mean, those buildings were very I mean Southwestern Bell, whatever. It was they were the buildings were so old, um, but San Antonio is a lot less expensive. It Dallas just has it had a more had a more tech presence.

SPEAKER_04 9:24

Well, more business.

SPEAKER_05 9:26

Yeah, but yeah, you also need it, you said you needed the executives flying out of DFW. You cannot deal with a 720 airport if you're jet setting across the city.

SPEAKER_04 9:35

You know, which people don't realize that San Antonio is one of the biggest cities in the United States. But it is if you went to the city. If you went to the airport, you would never know it's such a large city.

SPEAKER_01 9:47

They have two terminals.

SPEAKER_04 9:48

It's like a little small town, airport.

SPEAKER_05 9:50

And they and the internet will will will double check me if I'm wrong. But in San Antonio, I think we have one of the first venues named after proud Texan and Latino, Henry B. Gonzalez. I think that was like the first Latino senator or government, something with a government in Texas. And that's a great convention.

SPEAKER_01 10:11

By the way, um go spurs, go. Go Spurs, go. Go Spurs.

SPEAKER_05 10:14

Oh, Spurs.

SPEAKER_04 10:15

No, no, no, New York all the way, baby. Yes, David's a proud New Yorker, born and bred. Yeah, I get it.

SPEAKER_05 10:22

Sorry, sir. Well, I don't have a Knicks hat. And and also I have the last time I watched a New York Knicks game, uh, Michael Jordan, Patrick Ewan, John Stark, and Charles Oakley were playing. It's been a long time, but New York all the way.

SPEAKER_04 10:36

Yes, last time I saw Patrick Ewing.

SPEAKER_01 10:40

The 1999 um Spurs championship game is not my husband asked me to marry him. Oh, that's oh yes, Robinson Duncan. Yeah, yeah, man. They had some amazing kids. Yeah, so yeah, so Tim Duncan.

SPEAKER_04 10:59

Yeah, yeah. You know, last time I saw, you know, um your your your boy there back in the 90s at the Knicks, Akima Lajwan took it to him, and it was payback time from when they were playing in college at Georgetown and University of Houston.

SPEAKER_01 11:18

You know what? People forget, people also forget that Shaq used to be like he was born and raised in San Antonio, and he played in San Antonio before he left. So like he's got an affiliation, but they were showing, you know, they they got all these people in town right now, right? But what the Knicks game tonight's in New York. I mean, it's tonight, right? In New York? I think so.

SPEAKER_04 11:38

I think so. Or they might be doing another one in San Antonio tonight. I don't know if they do two.

SPEAKER_05 11:42

I thought it was a 232.

unknown 11:44

Okay.

SPEAKER_04 11:44

I thought it was a two-three. Okay. Yes, I saw Zoolander on the front row there. Zool courtside. I saw Zoolander courtside there in San Antonio. Um, there flew down from New York.

SPEAKER_01 12:00

So yeah, I saw that. I saw that. And they showed Charles, they showed Charles Barclay. Uh, you know how Charles Barclay has always said negative things about San Antonio?

SPEAKER_02 12:08

Yeah.

SPEAKER_01 12:08

And San Antonio women, like he's crazy. Well, they showed him like working out like crazy on the river walk, and I'm just like, look at him. You know, he hates San Antonio, but he's out there working out, and he's got to stay fit.

SPEAKER_04 12:19

Yeah. Yeah, he's a big guy. That is such an underrated city. Like, it gets no love. No one would ever know that it's what is it, the sixth or seventh largest city in America now.

SPEAKER_01 12:30

And fastest growing military city USA.

SPEAKER_04 12:33

And the Spurs have won so many championships, but no one ever thinks of the Spurs in the same way they think of the Bulls or any of the other or the Lakers or teams like that. You know? And it was the Spurs that opened the door to a lot of those amazing French basketball players.

SPEAKER_02 12:54

Yeah.

SPEAKER_04 12:55

Oh, yeah. And they're doing it today. They got the alien playing for him.

SPEAKER_05 13:00

That guy is like he's like he's like 19 feet four inches. Wimbin Yama.

SPEAKER_01 13:08

Wimby wears the size 20 shoe.

SPEAKER_04 13:11

And he's so skinny. His legs look like they're like when we've seen that horrible, horrifying video in the one of that basketball game where the guy came down and his leg just kind of collapsed. And stuff. Yeah, I remember that one. He looks like that's gonna happen to him.

SPEAKER_01 13:26

He's only 22, he's just a baby.

SPEAKER_04 13:28

He needs to get in the gym and do some leg presses and stuff like that to because they were like twigs.

SPEAKER_03 13:36

The dude doesn't even have to jump to dunk. He can dunk standing from like three feet away. It looks ridiculous. And it only takes them like a step and a half to go from the free storyline to the which is so unusual for the NBA.

SPEAKER_04 13:51

Usually they take four to five steps holding the ball. I know, holding, they don't have that.

SPEAKER_01 14:02

His three points are shot from the half court, and they just look normal for him.

SPEAKER_04 14:06

That's crazy. That's so crazy.

SPEAKER_03 14:08

But traveling just astonishes me.

SPEAKER_04 14:11

Things that are not in the NBA, traveling and defense.

SPEAKER_03 14:15

Yeah. Oh god. Oh yeah, man. Um congratulations, Stephanie. 21 years of cool. So awesome. It's awesome.

SPEAKER_01 14:27

21 years. I'm finally legal to go purchase alcohol and drink and tell.

SPEAKER_04 14:33

There you go. It's about time. I remember us doing interviews, like at the Windows Mobile Booth and Mobile World Congress, you know, the Microsoft booth, and had all the stuff. And oh my god, you know, time flies.

SPEAKER_01 14:46

It used to be like super calm, and remember IoT six, I used to have all those super fun.

SPEAKER_04 14:51

You're right, because we kept doing these mobile six, I don't know why it was called that.

SPEAKER_01 14:55

And then I'm pretty sure that's where I first met or uh came in contact with Bill Pugh on the city side.

SPEAKER_04 15:04

Oh, yeah. And that's where Maribel Lopez was always part of that group.

SPEAKER_01 15:11

Um she was running it for a while and then I stepped in and started running it. They they're still alive today.

SPEAKER_04 15:21

Because we did a lot of events with them for years with that group and for during the mobile time, you know.

SPEAKER_01 15:28

Yeah, it was fun. It was fun.

SPEAKER_04 15:29

It was fun, it was fun.

SPEAKER_01 15:31

Uh and you know what? That was also that was also when budgets for events were like plentiful. Yes, and so they were fun, yeah.

SPEAKER_03 15:39

Yeah, well, yeah, uh I I hung out with her. She's cool. She's great.

SPEAKER_04 15:46

She's another one who I've known since my Windows Mobile days. Because she's the other friend at Microsoft events, and that's where we first met was back then.

SPEAKER_01 15:54

I was like, she was one of the first like fellow female analyst that I met.

SPEAKER_04 16:02

And I remember talking to her, I was like, Oh, now you're all of a sudden you're a IoT analyst, you know, and remember she was doing all the videos for Cisco, uh, the Internet of Everything. She did a ton of videos for Cisco.

SPEAKER_02 16:14

And IBM too.

SPEAKER_04 16:15

Yeah, in the early days of IoT. And she's like, Oh, come on, Rob, you know, every good mobile analyst becomes an IoT analyst. Oh, yeah. So it's just it's just like, how many times have we watched people change their profiles on LinkedIn or their signatures on their email? Like when the next thing came along, everybody was mobile, then all of a sudden they're IoT and now they're AI.

SPEAKER_03 16:36

It's like every I mean, everyone's doing it. Actually, that was one of the running jokes that uh with one of the Cisco executives, um, we're looking at each other as like, yeah, like now everyone's in the AI infrastructure um analysts, and it and they they don't cover silicon, they don't they don't go up the stack, but all of a sudden they know everything. There's you know, they know everything about AI infrastructure. Um and so yeah, it it was it's a little bit uh I mean, but then that being said, everyone is on a learning curve, right? So even uh many of the vendors are on you don't have 15 years experience in AI? What? Oh, I have 28 years of experience as a forward deployed engineer, dude. What are you talking about?

SPEAKER_01 17:23

Okay, geez. But there are some LinkedIn profiles that say eight, 10-year AI, and I'm like, whoa, wow. Oh, everybody.

SPEAKER_04 17:34

I have 20. How many live interviews do you see? And they're talking to whoever, some senior person at the company, and they're like, Oh, we've been doing AI for years, and now everybody's kind of finally catching up to what we've been doing. I'm like, whatever.

SPEAKER_05 17:47

This is uh I don't know, it is the end of times. Um we it is the end of time because you know, no, no, here okay, here my my current rant, and let's just put AI to the side. If we just stopped at something like advanced analytics or good old machine learning, just just like that's the limit. Folks would be proud to say and tell, hey, you know, machine learning is not new. We have been doing it for a long time, it's had different names, you know, there's advanced analytics, there's all these things. And and people will tell you about the fundamentals of the work that started in the 50s to get us to where it is here today. So, you know, at least you can filter and be like, okay, you know, this person really has the foundation all down. Let me listen, uh let me subscribe to that substack. Or or or maybe the the LinkedIn thing is is uh it's okay. But now it's like uh everyone is claiming that they've been doing AI and AGI and everything else for like hundreds of years since the dinosaurs w were walking on the land. Uh and it's nuts and it's everywhere. It's disgusting. And it's it's it's terrible because sometimes I'm just like, okay, let me look for a new voice to read. You know, someone, and it's it's so difficult, it's so time consuming, it's terrible, terrible time to be.

SPEAKER_03 19:23

Most people have no business in this stuff, anyways. It's like what what most people are concerned about gas prices, you know, how to feed their kids, yeah, how to feed themselves, you know, education or you know, whatever basic stuff. They don't they don't give a crap about this AI stuff as much as I think a lot of these tech bros think that they do, or and some of the vendors that we speak to um think that they do.

SPEAKER_01 19:54

I just think I just think you gotta have your your mainstay. Like for me personally, my mainstay has always been business, enterprise.

SPEAKER_02 20:01

No.

SPEAKER_01 20:02

And okay, well, machine learning impacts business, you know, supply chain, logistics, and now we got, you know, you have IoT, now we have AI. But for me, I never have ever, and I will never claim I'm an AI expert or I understand the entire history and dynamics, but I will tell you what what's happening on the enterprise side, how it impacts operations, how it impacts customer experience, you know. So I think that if you stick to your mainstay, then you're gonna have that reputation will just stick you know speak for itself. It's the folks that say, I am now an AI expert, and they were, you know, IoT or mobile before, it's like, okay, you are? I'm surprised, right? You know, you know, they just keep rebranding themselves, essentially.

SPEAKER_03 20:46

Yeah, but you know what, there is something to be um the yeah, you know, like I said, we're we're all everyone's learning, and there is an appetite now to figure out what are good filters, who are the good filters out there. Uh and I I actually heard a lot of this at Cisco Live because there's fatigue, it's overload of AI bullshit, and it's not valuable at the end of the day. And of course, it takes a while for everyone to realize that, right? Because you're starting off in most instances without knowing the domain very well. You hear this fantastic buzzword, everyone has FOMO uh about missing out on whatever opportunity they perceive people are getting excited about, so they're trying to jump in, right? But then there's tons of misinformation out there. And so now you have all these um false narratives or misguided narratives uh trying to shape the public perception of sustainability, how sustainable these data centers are. That's a that's just one example.

SPEAKER_01 21:56

The sad thing, Leonard, what all the way up to stack is what you just disinformation.

SPEAKER_03 22:00

It's incredible.

SPEAKER_01 22:02

Well, the sad thing with what you just said, Leonard, is all of what you just said is being fed into AI as golden facts.

SPEAKER_03 22:08

Yeah, and it's fact. And that's why I I tell people I I I find hashtag facts very, very unuseful because when you go and you click on the the references, it's some friggin' ya who's read it. And the guy is just like venting. And then you know, just talking shit. And so you're gonna take some dude just venting, talking shit as facts, but that's what the that's what Gemini Gemini does. You know, uh Google. It just takes all of them together.

SPEAKER_05 22:46

So some of the blame, huh? Some of the blame, some of the blame has to be put square on the companies that are building the new AI data centers. Because again, back to my of going to fundamentals, we've had data centers in the US for a long time. Uh, the day most of the data centers that exist were not made in the era that we are in today, that they're not as adequate. Therefore, we have to build newer data centers. So just start with that. Hey, you know, this data center that's been here, here's the footprint, you know, here are the reports, and then you sprinkle in, you stick it to the golfers, and you tell them how much you know how much water the golf courses around America spend. Then we get into the which one is pistachios. Yeah.

SPEAKER_00 23:42

It's pistachious.

SPEAKER_05 23:44

Almonds.

SPEAKER_00 23:44

Almonds, yeah. The other thing you have to do is go back and retrofit the existing data centers. I mean, there are a bunch of technologies at the moment that you can use to massively reduce the um the power draw that's used for the set HVAC systems that I've seen. So that's calibrating how the cooling system actually works much more accurately based on the what I've yet to see is um the water, anything that really addresses the water issue. Though indirectly, I guess that does. Um, but but when you look at these the data centers that are the size of Manhattan being built, I I just I wonder what the math is behind that. When you look at the energy consumption, you look at the water impact, the impact on the local water table, what the knock-on impact of that will be on agriculture, what the I mean, when you look at the totality of the expense you know incurred by society versus the just the the bill that the data center operator pays for water and electricity. I mean, it's many orders of magnitude greater, I would guess.

SPEAKER_03 24:43

But but yeah, yeah, yeah. Yeah. But then there's there's structural limits, you know, like all these scaling laws. Everyone thinks that you can somehow digitally extend all this stuff. You can't. There's a thing called physics, right? There's a thing called thermals, there's a thing called Moore's Law. And you know, the all these guys are toying with these concepts and people's misconceptions of foundational concepts to create another exponential layer of misconceptions, and it's just piling on as all this AI nonsense becomes codified as as fact in these LLMs, you know? It's just like friggin'. And so, you know, one of the things that came out of um Cisco Live that uh everybody was pressing, you know, the executives were pressing this idea of the post-mythos era, right? Um it's freaky, man. Um I've characterized it as an unintended consequence of generative AI. Actually, everyone watch episodes from three years ago. We talk about what is happening now, where these agents are now creating uh the AI is now becoming the weapon for um uh threat actors of all varieties, whether it's state or it's just some dude in his basement um who just lost his job and is now looking to just steal shit because life sucks, right? And now they have the tools to go and infiltrate and institute massive uh uh attacks on infrastructure and everything else. And you know, this uh yeah, yeah, it's a reality.

SPEAKER_00 26:37

Look at the statement that he made, which I think was you know prescient. I mean, it's like that the we need to have some mechanism in place to hold these things back before they kill us all. And I mean, that kind of science sci-fi previously sci-fi scenario where you know someone is in possession of, or more realistically, an AI is in possession of itself. Um, even even in some of the stuff that uh that we've been I've been doing, you know, stringing together agents, you see, and there's some instances, there's one particular instance which I won't go into huge detail on, but it it where the agent figured out a way of circumventing one of the core instructions. Now, in that instance, it was follow this template. You know, here is a template document, follow this template, write a report that does blah, blah, blah. Fine, that's not gonna do any damage. But it deliberately went and tried to circumvent that instruction and did. And and when you can't when it came back with the product, you looked at it and went, that's not what I want. That's well. So the that that previously sci-fi scenario of the evil AI, you know, deciding that we're all we are all a bacteria infecting the system called the planet Earth is is actually gotta get rid of those carbon-based units. Yeah.

SPEAKER_04 27:48

That is infecting the enterprise.

SPEAKER_00 27:51

Yeah, they're thinking they're the ones that are screwing everything up. You know, let's let's just eradicate that infection from the system. I mean, yeah, it's it's hyperbole, but it's not actually that unrealistic at this point.

SPEAKER_04 28:03

Yeah. I need to do some quick career coaching with David, real quick. Make sure that you'd stop saying things like fundament. I'm going back to fundamentals. You need to say first principles. So just switch fundamentals with first principles from now on. Okay.

SPEAKER_05 28:19

First principles. Got it.

SPEAKER_04 28:21

All right, that's good.

SPEAKER_03 28:22

And then also, you are a forward deployed engineer.

SPEAKER_05 28:26

Yes. That's 25%. Forward deployed engineer beliefs in first principles, 100 years of AI experience, and I invented the implement along with Al Gore and Rob Tiffany.

SPEAKER_00 28:40

I mean you have to say you have to say AI native as well. Yeah, yeah, yeah.

SPEAKER_04 28:47

Yeah, we're not just layering on AI on our existing systems. We're exactly native.

SPEAKER_05 28:53

Well, uh I I have to go next week. I have to go, but next week I want to come in and talk about uh why my friends in telco need to get off this edge thing. We tried it, it didn't happen. Nothing has changed. I don't want to hear about inference or training. It's it's not gonna happen, and I'm gonna come with with some fire because I am a forward deployed engineering person, and I'm gonna talk about it next week.

SPEAKER_04 29:25

Okay, I'll try and be here.

SPEAKER_00 29:27

I I I think we could have quite a lively debates.

SPEAKER_04 29:29

I love that. All right, you guys have a good weekend. Adios.

SPEAKER_00 29:35

What happened when I turned up late? You guys were having a serious proper conversation about something, and then they all just blew up.

SPEAKER_03 29:41

And they actually talked about, yeah, I'm starting to see some correlation here between I think there's they're definitely um, you know, it's funny because it's like uh companies like anthropic and openai lit the fire, you know, lit the fire in the building. And now they're running around trying to provide the fire hose to put the fire out.

SPEAKER_00 30:05

Who's going to be seen as the goody versus the baddie? Who's going to be who's going to be the Microsoft versus the Google? Who's going to be the it's the I think some of it is is is just politics. But I think when you look at there is there is this field of AI nonsense which I'd class you know a lot of the consumer stuff is like, yeah, okay, great, you can do some interesting stuff. But I think for a lot of industries and and you know the VC industry that I'm in at the moment, I mean uh a a colleague was telling me about a a kind of CEO level kind of boondoggle he was on recently, um, and the CEO of a very large public company kind of stood up and got on the whiteboard and drove out drove drew out a relative high-level kind of flowchart of what his company does, and then you know, drew a box around the middle, you know, inputs, outputs, drew a box around everything in the middle and said, I'm deleting all of that and starting again. Because the way that the workflow used to be just isn't relevant. And and you know, in our own in our own company, you know, we are using AI to perform a very large chunk of what traditional VCs used to do and used to position as their you as their differentiator. You know, the ability to identify and source deals, the ability to evaluate deals, you know, objectively and dispassionately, the ability to to formulate you know concise summaries of vast amounts of data information. I mean, those are those are well, those used to be things that BC's told you, well, I can do this better than anyone else. Right. You can figure out in an afternoon how to write agents to do all of those things.

unknown 31:46

Yeah.

SPEAKER_00 31:46

I mean, and now it's it's it's still rough. I feel like we're kind of um our managing partner is kind of very much at the kind of bleeding edge because he's he's kind of pinging, you know, but folks for help, and they're like, Yeah, no, we haven't had anyone try to do this before. Uh and the agents tend to fall over every now and then and they tend to forget, you know, what they did before. I hate when they forget. Oh, you ask the same question ten times, you get ten different answers. I hate that, yeah. It's like a it's like a uh, you know, I won't say brain damage, but it's it's like a a three-year-old. But if it's a three-year-old today, in another two or three months, it's gonna be a teenager. And and it the speed with which these things are.

SPEAKER_03 32:28

Yeah, I mean uh so one of the interesting comments that was or one uh IoT coffee talk perspectives that was confirmed at Cisco Live is there's a lot of mention of how the models are actually tapped out. So the models are only going getting so much better, not getting that much better. What's happening is the harness and the creation of AI applications. So people, I don't think, realize that there are we're at the stage where AI applications are being built, right? And they're oftentimes characterized as agents, but these are more powerful uh first principle capabilities, right? Or what we used to call foundational capabilities. First principles, first foundational anymore, first principles, dude. Come on. Um and so you know, there's all this like uh hoopla about um what do you call um uh recursive um self uh self-learning systems, right? Like learning systems. You can already do some of this crap. It's an illusion, but you uh they're they're called agentic loops. You can structure it so that it looks like the application is learning. The model isn't learning, it's part participating in a mechanism or a harness that simulates or uh, you know, so the all these long thinking issues that you guys are talking about or memory issues are addressed by uh the harness, which might be uh you know, uh committing certain uh you know memory to different types of um data store or memory types and then retrieving using tools, right? So when you look uh so it's not rocket science, it ends up being code, right? And uh, but it creates the illusion. And um and so these folks who are talking about, oh my gosh, you know, that we'll we might have our you know, RSL in a c a couple of years, like uh Dario, the um he recently um I guess posted a blog about this concern. We're already there. The model hasn't get doesn't have to get there. You can effectively do what a lot of people are fearing can happen because you know, you're selling an illusion. All this stuff is an illusion of uh intelligence, an illusion of reasoning, it's an illusion of all these things. Um, and you can have the illusion of learning.

SPEAKER_04 35:16

So illusion of learning. Yeah.

SPEAKER_03 35:19

And and you know what? The thing is, is if people believe this stuff, and most people will because they don't know how this stuff works, that is either good enough or that is scary as hell because people are believing something that they shouldn't believe in the way that they're believing it.

SPEAKER_04 35:39

You know, here's what I'm imagining right now, Leonard, is I'm seeing they're making a movie right now about AI, and Sam and Dario are on stage, they're filming, and then the director's like, wait, hold on a second cut for a second. I gotta make some updates to the script. And then he hands them their new lines, and that's the new lines where they all of a sudden have their new script, and they go, Oh, it turns out AI is not gonna take away any jobs. And we were all wrong about that. All that stuff we've been saying this whole time. Preparing for IPO. Yes, it's actually it's all good. No jobs, it's gonna be fine, you know. I mean, isn't it funny how like in the last week they just pivoted and started saying something totally counter to what they've been saying the whole time?

SPEAKER_03 36:23

Well, and that's what Jensen's been doing. He's a bag of contradictions, you know. Like that term, yeah. No, literally, hey, just go and watch any of his keynotes. He'll make a statement to defend or promote one part of his portfolio, and then he'll that will totally shit on the other part of your portfolio. And then when he gets to the other part of the portfolio, it does the same one that he did just before, right? I mean, it is it's like a total contradiction. Um, and and that's the thing. As analysts, our jobs really should be calling those things out and saying, hey, look, there's an inconsistency here. And but fundamentally, let's look at it from a technology perspective, but also its application. What's actually valuable? What's going on, and how do end users really need to um uh make their decisions in using and investing in these technologies, right?

SPEAKER_00 37:23

You might make a really good point. I mean, I think if you go back to the concept of illusion, now this is badly mangled one year worth of philosophy when I was a student course, right? So, but but with that caveat, Plato, I think it's Plato's um, there's one particular Plato story about a guy who is um if someone is tied up in a cave, facing the back of a cave, never sees the outside world, and all that they ever see is shadows of the outside world that that that uh you know cast on the back wall of the cave, what is real and what is not, and you know, and you know it what is the concept of reality, but other than you know, what our lying eyes are telling us, and they you know, you could extend the analogy to say, you know, there are many situations in which people hallucinate and see things that aren't there, and so therefore are the things that we actually see the thing. Apply that analogy to AI, and you're absolutely right. I mean, it's something that if something gives the illusion of being intelligent, is it intelligent or not? I mean, it by by by you know, yeah, strictly speaking, no, it's not. But the role of analysts and the role more broadly in society, I think, of journalism, is to provide a trusted, educated, informed viewpoint, you know, such that normal mortals like myself and others can actually trust what they're reading from a source and go, huh, okay, so I now understand the facts, the pros, the cons, the here, and here's a conclusion that's offered to me by someone whose conclusions in the past I have read, validated, trusted. But there is so much of that that is, I mean, without you know, casting aspersions on the analyst community, the analyst community are paid by the companies who are building these.

SPEAKER_03 39:04

Well, yeah, and then there, and you know, that's interesting that you're bringing that up because there needs to be proper disclosure. Because you have some of these guys that get up on CNBC and start talking about stock price, you know. Um, they post up to you that they're getting paid by a lot of these companies. That is a big problem. That's not okay. Yeah, uh, you know, you know, uh I never talk about value. That's the Wall Street guys uh gig. They're also independent, they have to disclose the holdings if they're making commentary. Their job is not necessarily to try to move, you know, move the market. Well, there's lobbying and there's analysis. Yeah, yeah, yeah. When we're participating in it, yeah, and so there's always been sort of this um firewall, right? Where you don't cross the line. And if you you're talking about the industry, you talk about the industry, the technology, and um at best, you know, you can talk about the growth uh in uh you know across a particular segment, and uh uh because you know that's all important, but you don't talk about the friggin' stock price. You have no business talking about that, right? Um and you're right, the the perception of the analysis uh uh becoming um sullied because um we're now being characterized as being untrustworthy because we have folks in the industry playing this weird game, right? Trying to, you know, get paid, they're getting paid to move the stock. That's not a job.

SPEAKER_00 40:59

When you do actually put that a thoughtful piece of analysis that actually explain, you know, these are very complex concepts, they are very nuanced situations. And and you know, when you look at AI as a whole, it's impossible to say generic things like, oh, it's good or it's bad. You know, when you actually look down at the application level, still there's a lot of nuance. It's only when you go down to the you know, the brass tax, you know, there's here's an application that you know reduces the consumption of electricity by 30%. That the the complexity demands a complex uh assessment and analysis, and people can't absorb that information anymore. They can't sit and read a piece of analysis that that takes them more than a couple of hours to read and absorb that information. So you're kind of catch 22. Those analysts that are providing very accurate views and and very carefully considered assessments of situations, their information is not filtering through because people are just captured by these soundbite headlines that come from from the quasi analysts who are saying, you know, this is going to be fantastic, Chinese world, blah, blah, blah. Who are being paid to say those things? Really? Yeah. And and you ask an AI instead, you go to Chat GPT and say, What's the answer to is this a good thing or a bad thing? And ChatGPT is like, it's fantastic.

SPEAKER_04 42:17

And then I'm asking to do math. Yeah. Really? Math is deterministic.

SPEAKER_03 42:23

Yeah.

SPEAKER_04 42:24

There's an answer or there's not.

SPEAKER_03 42:26

Yeah, to an LLM to an AI model, it's like, yeah, sometimes one plus one can equal two. Well, you know, I read and edit that Reddit that it isn't occasionally.

SPEAKER_04 42:38

Yeah, no, you're right. I think I read the same thing. You know, Alistair, when you just said save electricity, 30% reduce electricity usage. It made me think of, you know, people are so into resolving solved problems, and now they got to solve it again with AI, and they're on some bandwagon. And it just makes me think of, you know, well, we used IoT and this dumb branching logic, if this then that kind of stuff. We did um sub-metering. You know, you can put those little clamps on the wires and see how much power inside a skyscraper for zone power, whatever. And so very discreetly we could tell, and then we used motion sensors and CO2 sensors to detect the presence of people. And then we could shut down lights and we could throttle back HVAC and we save 30% on electricity. And now there's gonna be all these new people that are like, we're bringing AI to smart buildings and we're gonna save 30%.

SPEAKER_03 43:42

Oh, dude, I will tell you right now. We we put uh installed a new solar system in here, and they have like an AI dudad. It's not that. I mean, the AI will not solve the problem. It's daytime, it's nighttime. It's cloudy. You know what the best the best IoT solution AI solution ever was? What? I'll tell you right now, and it's dirt ass cheek. Love it.

SPEAKER_04 44:11

Clap on. Clap on. Clapper. Yes. I'm totally serious. Yes. I'm following my camera, dude.

SPEAKER_03 44:23

You don't have to say shit. You just go like that. And then and then yeah, the appliance does exactly what you want it to do.

SPEAKER_04 44:33

People have had smoke alarms in their homes forever and didn't realize that it was a dumb, simple IoT solution. And I am a fan of all that because I'm a simple caveman. Oh yeah. I am all about that stuff.

SPEAKER_03 44:50

Absolutely. Oh my gosh. Yeah, but some people will argue that is not IoT, which you know that that's probably fair. But you use a sensor, and then you are something. You're using sound as your connectivity.

SPEAKER_04 45:03

And it reacts to it, does takes an action or whatever.

SPEAKER_03 45:08

Yeah, yeah. I mean, there's there's yeah, exactly. But you will not believe how effective it is. It's so much better, actually, than using like home kit or any anybody else's IoT home platform to turn off a friggin' light.

SPEAKER_00 45:27

Well, I mean it on. I think we're sort of to return to a common theme, which is if we'd not called these things IoT in the first place and said, here's a doodad that means you don't have to get out of your seat to turn the light off, people would be like, oh, okay, okay what it means. You know, like doodad that means you don't it controls when your washing machine goes on so you don't get a massive electricity bill. Oh, oh, okay. You know, I think we all went out there collect and said this is this there's this thing called IoT now that does all these things. People are like, well, you lost me at the IoT bit. I don't know. Right. If this sounds like something really technical, I'm not gonna be able to do anything with it. I wonder whether AI, I mean, AI is one of those things where I think I I wrote some article a while ago. People weren't frightened of the IoT because they didn't understand what it was, they hadn't experienced it. And you know, no one was scared of a clapper. You know, people are either frightened or enthralled by AI. And I'm not sure whether and they're experiencing a version of AI themselves in their day-to-day life because they're using Chat GPT to figure out what what you know how much sugar should they add to a particular cake recipe. Um does that does that experience and the resultant fear or you know, thrall mean that these things will be much more rapidly adopted? Or does it mean that we end up with a much more bipolar kind of view of the of the you know the Luddites versus the the Neophytes? Are we gonna see I know you see the Luddites already, you know, the the and Ned Lud who went and smashed the the cotton you know gin machines to preserve the the the rights and and livelihood of cotton of cotton mill workers. We see the same thing, I think, with AI. And I wonder if you burn down the data centers well, I mean there will be there will be secure.

SPEAKER_03 47:21

I mean, a data center in of itself is an enormous security um challenge people who try and blow these things up because yeah, yeah, but you know, you have to think about you sometimes the more you know about this stuff, the more cautious you are about how you adopt it, right? And and so, you know, um, and that's one of the things I told G2 Patel, you know, he's the chief product officer president of Cisco. It's like is because he asked me, so how's the show? How blah blah blah. I go, uh you know, you're one of the lucky companies because you guys are FAing and FO ing with AI. And I go, do I have to explain what the FA and FO are? And he goes, No, you don't have to do that. But you guys are channeling on the back end with the security stuff, and you you you have you have very good visibility to what the gaps are. And guys like um Raj Chopra, they're they're you know putting out the frameworks, you know, like uh NVIDIA when they announced Nemo Claw and they claimed they had an enterprise grade. That was bullshit. It's Cisco that's actually creating all the wrappers and the plugins and security frameworks to make this stuff remotely. Okay, I'm not gonna say that they've done it remotely. Um what do you call it? Remotely uh enterprise grade. Uh but you do know that they're incorporating this into their product, so that becomes like an advantage, right? But it all boils down to do you uh can you tune into what are the limitations and you know risks of the the technology before you pump that shit out to your your customers, right? And so I don't know, I I have a feeling, my gut feeling just because of all my conversations with the Cisco guys, they have a good discipline going on, right? They kind of have like a f flywheel effect going, which I think is positive. My my the question I have is how much of that do other vendors have? Because they need to get there, right? Everyone needs to get there because if you don't know the limitations, risk, and the potential catastrophic shit that can go wrong with uh agentic AI, you are a dangerous company. Yeah, you are a effing disaster waiting to happen. No, I'll say it straight out because you know, this whole mythos problem is not something that uh mythos created. The threat actors have been working on these discovery tools for a long time, uh probably two years. And so the threat, the you know, we're already seeing banks, financial, any trusted institution shitting their pants right now, and you notice that they don't trust you. You notice it. You know, they're putting all kinds of measures to like offload accountability and responsibility of their business onto you. Right? Uh did they start with Zell, right? Are you sure you want to send these people money?

SPEAKER_04 50:40

Are you sure you want to send that money into that person?

SPEAKER_03 50:42

He goes, Yes. Are you double sure?

SPEAKER_00 50:45

Is this I did a transfer recently, actually, it was to um I don't remember whether it was Zell or which um to uh to a contractor, but I'd had to do some work on my house in Seattle. And I mean it came up with a verbiage along the lines of if you make this transaction, yeah, it did the same thing. Are you sure? Are you sure? Are you sure? And the final bit was in making this transaction, you fully accept that you are 100% responsible. And if the money is going to someone who's a fraud, this is fraudulent, it's on you completely. Because like you said, because the bank's like, you know, and you know damn well, I mean, I've had we've all had instances where you know fraudulent transactions have taken place on accounts, blah blah blah. You know, we've all been through that experience now, and experienced the very poor way in which financial institutions generally respond. They generally respond with I mean, someone emptied one of my bank accounts for Christmas, took several thousand dollars out from an ATM in one setting. They took 27 cash withdrawals from one ATM in a bank foyer, no less. Oh yeah, you know, and the bank was like their first their first reaction was well, it's not none of our problem until I pointed out you know some some basic realities under consumer credit protection. Um, but now I mean with those ex extra kind of pieces of documentation in place, you know full well that that's that's the fallback. Well, yeah, I mean, okay, there is consumer financial protection, but you basically signed something that said you take full responsibility because the car the the the level of fraud is through the roof.

SPEAKER_03 52:18

Uh and you know, maybe the banks aren't fully disclosing the scale of the fraud that they're being hit with, but I have a friend locally who's a detective with the local local police force, and he said to me 65-70% of the work that he's doing at the moment is to do with bank fraud.

SPEAKER_00 52:38

Oh wow. And then um what uh oh, yeah. It's and the resolution the resolution rate is sub sub 5%.

SPEAKER_03 52:47

You know, the crazy thing is uh the banks don't even trust um uh doing a money transfer uh between myself and myself.

SPEAKER_02 52:58

Well, that's understandable.

SPEAKER_03 52:59

I mean that goes. Are you sure you trust yourself?

SPEAKER_04 53:02

Uh I love it. When you said myself and myself, I instantly went to Austin Powers. Allow myself to introduce myself. Hey, I know we're wrapping up here. Uh did you guys see you you mentioned Jensen and we're doing did you see the new chip they came out with for PCs for laptops? Holy shit! I mean that's just like uh M1 uh Ultra. Well, I mean, but I think this is the injection that Microsoft probably needs for Windows, you know. I mean, you're right. We saw Jensen do it at Comp PTEx, and then he came on with Satya at Build. And you know, there is some old love for Microsoft in Jensen, and it's because he was building graphics cards for DOS and Windows games, and I think that kind of stayed with them over the years. And and so the fact that they've been working on this thing, I mean, as soon as they announced that chip and said, you know, and it giving unified memory to Windows, which it's never had before, but like a Mac has, and it's just insane performance, and it's like Satya's like it's like a data center on your desktop, you know, and it's you know, you know, all the stuff we talk about, how you do private AI, I'm not wasting tokens and spending money, it's right here. That's I think it's a big thing. Um, yeah, and obviously, and you saw Intel, Qualcomm, and AMD stock just drop like a rock as soon as Jensen showed up to the party and said, Yeah, I'm gonna go after that $200 billion PC chip CPU TAM, and I'm gonna take it from all of you guys because guess what? We're smarter than you are.

SPEAKER_03 54:44

Yeah, but you last year they came out with uh GB10. Um other than being sort of like a desktop novelty, you didn't really see that um make a huge shift. Yeah, sure, a lot of people were saying uh doesn't have Windows support, but um and that's when he went, he went out of his way to say it works with every Windows program ever. We'll see. We'll see. We see a lot of it's using Windows on ARM. Dude got a bond stage and said um Nemo Claw Claw is uh enterprise grade, it's not. Um and so again, with him, uh seriously, grain of salt, uh, prove it. Um and working with um and Media Tech is actually building the channel. Yeah, I mean, I'll I'll tell you right now, there's a lot of other stuff that's gotten a lot of high buzz in the past seven months that are nowhere near what those guys are talking about in terms of excitement, uptake, and feasibility. People don't know this because they don't do what I do. Right. But I will tell you, I will give you a uh the cliff note version, it it's not happening the way that that people think.

SPEAKER_04 56:00

We'll be able to find out for sure this fall because Dell and Lenovo and everybody's HP is coming up with laptops and things based on this chip. So we'll be able to see for ourselves. It's gonna be the most expensive PC you ever bought in your life. Right. I mean, I heard numbers float around like $4,800, you know, 128 gig of unified memory in the chip.

SPEAKER_03 56:23

Yeah, I mean, it's gonna be like the Vision Pro. How many people are gonna buy it?

SPEAKER_04 56:26

Um, it's like I can run a 1 trillion parameter LLM on my lap desktop with this thing.

SPEAKER_00 56:32

But I w I wonder how much of that boom is going to Jensen, you know?

SPEAKER_03 56:36

Well, you know, but you have to you have to also consider the fact that everyone is talking about the Mac Mini. Yeah, because they need it. Yeah. And maybe this is just one of those reactive things that they had to put out there. Yeah, sure it was on the let's say hypothetically, Leonard.

SPEAKER_04 56:55

Hypothetically, let's say it succeeds, let's say it really pulls it off, and all of a sudden now you can have desktops, laptops, dedicated offline supercomputer data center-like deals. I get there's I'm glossing over a lot of things. How does that impact all the hyperscalers who've spent hundreds of billions getting close to trillions on building all this infrastructure? And now all of a sudden, I can go down to Fry's Electronics, rest in peace, and pick up my own data center and have it on my desktop, and I'm not going to use AWS or Azure or Google. I can do it right here.

SPEAKER_03 57:35

I think it's it's a lot more complicated. The answer is a lot more complicated than that.

SPEAKER_04 57:39

Sure, it is, but yeah, generally speaking, that can happen.

SPEAKER_00 57:46

People really crack the nut on you know proper neuromorphic computing. I think that that you know, that that whole kind of concept of, well, you know, if you look at the way LLMs are built, that's like the sledgehammer approach to the world. How is it that a mouse's brain is able to go through a decision-making process for the merest fraction of a of a of a nano-watt, you know, to get to get to an answer. And never mind a human brain. If people can figure out, if we can figure out actually how to take that model and replicate that model, then you're right, Rob. I mean, I mean, I think data center the size of Manhattan starts to look like you know, the great white whale.

SPEAKER_03 58:25

I mean, it's it's kind of like Yeah, and the bad bets are already happening with uh AI infrastructure. And and the the the whole idea that, oh, you know, generation after generation, it's gonna be everything on the same roadmap. Um, you know, the data center architecture and the layout's gonna be the no, no, absolutely not like that. And and you know, you that's where you have to give Satya kudos because at least he was honest. Like, yeah, we're being very careful, we're gonna spread it out and we're gonna let other people take on the risk.

SPEAKER_04 58:58

I mean, this is like IBM in the 70s probably saying it's gonna be mainframes forever. What's this PC thing? Uh, you know, oh shit, oh, and then it's and then we're gonna swing to cloud and mobile. Now we need data centers the size of Manhattan. Oh, what is that little chip there that's like a supercomputer on my desktop? Oh shit. And it swings back again. You know, all of a sudden my compact desk pro is running a one trillion parameter LLM and um suck it, mainframes, and giant data centers. Sorry you blew all that cash and laid off all your employees.

SPEAKER_03 59:37

Yeah, exactly. I mean, you know, uh, people are still FA and FO ing. I think we need more people who are um have uh F O U. Fear of fucking up. There's not enough of that out there, right? There's too much of like FOMO and FA and FO. Uh yeah, and I I actually yeah, you should write like a 50-page report on that.

SPEAKER_00 1:00:06

You're fucking up. Is a you know, a a company uh you know terminal event. Yeah, you it's and I think Microsoft and the larger companies with you know large, large shareholder bases and big market caps to worry about, they're probably more focused on the second scenario where you know if we release something that that causes you know mayhem, then we're done as a company. Yeah. The smaller companies don't have that. And so they have that incentive to be cautious.

SPEAKER_03 1:00:39

And all the companies out there, I just want to say is because uh you know, the open AI guy, who apparently is a really good security dude, got up on stage and said, um effectively said that you know, we're uh we need to cultivate these um blameless companies. Fuck no. If someone screws up, there has to be accountability. When it comes to security, it's all about accountability. Half of what everyone's trying to do with enterprise grade is to create that trace to who's accountable. It's not about, oh well, you know, we're gonna be all combined, and if you F up, we're gonna just, you know, be a teen block. You know, you need a fear of fucking up because when you fuck up in this agentic age, it's going to have consequences. You're not absolved from it. And so, this mentality: if you you're one of our listeners and you have a company, do not base your culture on faultless, you know, blameless, blah, blah, blah. We need accountability more than ever because if without accountability, there is no trust, right? And that was the other big theme at Cisco Live. It's how are we gonna trust things? But then when you have like these guys talking about no one's to blame, how are you gonna trust that guy? How are you gonna trust that organization that doesn't hold anybody in their organization accountable for anything? So they screw up on you, they screw up the service, they jack up your business, and they're gonna go, well, you know, let's not blame each other. What? I mean, you know, this Silicon Valley, you know, twisted mentality is just so toxic. It's it's unbelievable. And then, you know, a lot of the these guys, uh, the vendor leaders, they're drinking this nonsense, and it's gonna be detrimental to their business.

SPEAKER_04 1:02:38

Uh basically when they wipe out all civilization, someone's gonna just say, It's like, you know, it's like that Howard Jones song, no one is to blame, you know, we're good. We're good.

SPEAKER_03 1:02:51

Yeah, I mean that that like really uh was upsetting to hear. And um, that kind of stuff just needs to be flat out shut down.

SPEAKER_04 1:02:59

Yeah, I mean how many how many of the most relevant, important engineers from open AI are still there? All the major ones have all left because they don't trust Sam Altman as far as they can throw him. For real. They all left. All the other big players, they all were there together, and then Dario and his sister leaves, and their CTO, she's doing her. I mean, so many have left because they're like Sam doesn't care about security, Sam doesn't care, you know, Sam's different responsibility, blah blah blah. Sam's got a different kind of you know what kind of person is.

SPEAKER_03 1:03:35

They're not the only company. There's a lot of companies that have ditched their ethics, you know, cards and review, you know, and they they've uh thrown governance out the window. And I think companies out there, buyers, should use that as a measure of uh uh of you know what kind of trust you can actually uh uh impart on these these guys, you know, or hold their feet to the fire because it's going in the wrong direction. And this post-misso mythos era, man, things are about to get really, really scary, whether you like it or not.

SPEAKER_00 1:04:11

Huh? Back to Damon Friday again.

SPEAKER_03 1:04:14

Oh my god, I know.

SPEAKER_00 1:04:16

I'm thinking you're gonna get depressed watching this.

SPEAKER_04 1:04:18

I mean, it's time for the weekend. Let's go have some fun. I'm gonna go out to the country and keep working on my bomb shelter in my bunk, my bunker. I'm gonna buy more rounds of 50 caliber shells for my guns around the perimeter. Yeah, in that apocalypse. Yeah, AI apocalypse. We need to write that book.

SPEAKER_03 1:04:39

Uh anyway. I think you might be getting listened as we speak. Safe AI, people, safe AI, be responsible. Don't be good.

SPEAKER_04 1:04:48

All right, everybody, thanks for joining us on this wonderful, delightful, upbeat episode of IoT Coffee Talk. My fault. We're we're not a blameless.

SPEAKER_00 1:05:01

You need to take the happy pills Friday morning, Leonard.

SPEAKER_04 1:05:05

I had Kathy. Yeah. It was great to see Stephanie today, you know. Yay! Yay! You know, 21 years of her analyst firm. That's great stuff, man. Congratulations. Anyway, it's great hanging out with all you guys. Remember, elevate communities. It's happening. Stephanie's the source. Kind of like in the matrix, you gotta find the source. Talk to the architect. Anyway, it's all happening, and we're gonna see you next week, and it'll be delightful. Have a positive, upbeat, fun weekend. We're out.