IoT Coffee Talk
Welcome to IoT Coffee Talk where we talk tech and all things related to the Internet of Things. This program was started by a group of friends who just happen to be thought leaders, experienced practitioners and business leaders in the technology and communications services industries and more. This pod will feature weekly recordings of our casual conversations on the tech topics of the day with the hopes of delivering engaging and thought-provoking insights to our audience. We hope you enjoy program.
IoT Coffee Talk
#309 GenAI's NFT Moment
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Oh yeah. Nice. Epic. That is epic. That is such a great time.
unknownYeah.
SPEAKER_04Two weeks in a row. You know, I was really feeling the karaoke coming on, but then I just held back. Just held back. It did have the urge, you know. I love the rain, you know.
SPEAKER_03Yes. Yeah, that's good stuff. Yeah. That was a great record. 1985.
SPEAKER_04Wow. Is the cult still together? Are they still touring?
SPEAKER_03I heard that it broke apart. They were going to go on a tour. Oh yeah. Yeah. What's his name? Ian Ast. Ian Ast. Nicely. Nicely. Yeah. That guy. What a blue.
SPEAKER_01I last saw him at Milton Keene's Bowl when they played live, but about him as a teenager, and he was hit by a two-liter bottle of this questionable liquid that was thrown from the crowd and just oh my god. The perils of live performance. Oh my god.
SPEAKER_04They didn't have the chicken wire up to protect them. Chicken wire.
SPEAKER_05Chicken wire.
SPEAKER_04Too funny. That's what you gotta have sometimes, right? You know, especially if you're playing country and western.
SPEAKER_01Well, know when to have the chicken wire. That's a very exactly like.
SPEAKER_03Yeah, you know, they were like the neo, to me, they were like the neo ACDC, you know. They had the great guitar riffs, right? And uh just unique, awesome vocals.
SPEAKER_01Yeah, no, and then Sisters of Mercy was kind of the darker version of the cult, and then Billy Idol was the light, yeah. Billy Idol was the live back. Cult was was regular, and and uh Sisters of Mercy was like heavy cream. Wow.
SPEAKER_04I saw that Billy Idol's going into the rock and roll hall of fame. That's right. Great.
SPEAKER_03Oh, he should have he should have been in there a long time ago. I know.
SPEAKER_01I actually had an interview with him on, I think it was an MPR recently, you know, talking about uh essentially his his profound surprise that he's still alive after I bet, I bet I mean like at some point his dad moved to uh the US to basically stop him going completely off the rails. And and yeah, it was really interesting, um, really interesting interview.
SPEAKER_02He was on uh American Idol recently, this past yeah.
SPEAKER_01Oh really? Oh wow.
SPEAKER_02Actually sang uh what was that called with the song more more or whatever that song was like Rebel Yell. Rebel Yell, yeah. He sang it actually. He came out and did a little surprise thing, and uh yeah, he crushed it. Steve Stevens was there, Steven Stevens.
SPEAKER_04Hey Little Sister Shotgun. Yeah, yeah. You know. I surprised you that whiplash smile, right?
SPEAKER_01He spent a decade going.
SPEAKER_04But what what is the greatest thing that Billy Idol gave to all of us? He gave us Generation X. So thank you, thank you for naming our generation.
SPEAKER_03I thought it was uh STD of some kind. I don't know.
SPEAKER_04Oh, ouch. Well, I may some of the people in may all have also gotten that.
SPEAKER_01You're closer to the stage than I was then.
SPEAKER_03Yeah, yeah. Good good times. No, I you know, um what was that? It was a NAM show. Uh when was it? Um 19 uh 1992, no, 1988. Yeah, uh NAM show. Yeah, I went to a party and met um met, I don't know how I ended up there. Uh but I met uh met uh Billy Idol uh back in the young. Yeah, it was like crazy. It was disgusting actually. The you know that these people partying super hard, and it's like, holy crap. What's that? Oh my god. I don't want to sit there. Yeah, you know, uh people went pretty nuts back then. It was like crazy.
SPEAKER_04I saw him at a concert, it was like an outdoor thing in the late 80s, and maybe it's the is it the precursor to the mosh pit that we got with the grunge guys? Because everybody was slam dancing, and I was just getting beat to hell. You know, you're up near the stage, and you have people wearing like these studded bracelets and oh yeah, you know what I mean? That I did not get, yeah. Dog collars, and people are all curses, yeah. And man, they're just like throwing people and just slamming, and yeah, it was pretty brutal. Yeah. So I'm so glad I did that. Not oh my god, yeah, uh, yeah.
SPEAKER_03So everyone remember, um, take us seriously at your own risk. And I want to remind everyone if you don't know already, uplink matters. Um, and so fiber to the home is awesome, and I don't have that. And so I have really jittery recording uh on my end. So uh that is the reason why uplink is important. Uplink is important.
SPEAKER_04Yes, yes we have other people on our group and our team who I know they have fiber to the home, and you would never believe it. You know, I'm not gonna say any names, but sometimes his connection is questionable, even though he's got fiber to the home. So really okay. I don't know. I don't know. You never know if the fiber to your home leads to a 28-8 modem somewhere. Yeah, this is true. This is true. You know, you know that picture they always show you, that diagram of this castle, this b blocks or whatever, like the whole internet or all this deal held up by a little tiny piece, you know.
SPEAKER_05Yeah.
SPEAKER_04You see that all the time. They just put different names on it, but you know, that's that's pretty much that's pretty much every architecture I've seen. Is yeah, it's magnificent except for this one little thing. Oh my god.
SPEAKER_03Yeah. So what's up, guys? What's up? London. London.
SPEAKER_02London. London 2026. We went live with our webpage a couple weeks ago. Yeah. Events are registrations are coming in. I actually got in talk in contact. We have a cool one of the cool panels we have is around the EU AI Act. Oh, yeah. So I've got a um EU policy person coming in to talk about that and uh experts and stuff. So that's gonna be good. We've got a uh crazy cool mixer going on the June 8th in London.
SPEAKER_04And uh in London, baby.
SPEAKER_02That's right. L-O-N-D-O-N.
SPEAKER_01You're not allowed to do Austin Powers accents in the conversation. It's all Austin Powers.
SPEAKER_02I was wondering if I should hire like an Austin Powers impersonator.
SPEAKER_01Yes, you gotta get one.
SPEAKER_02You gotta get one Burgundy, Austin Powers.
SPEAKER_01I love it. It's a continuum, you know.
SPEAKER_02Yes, London 2026.
SPEAKER_04Wow, that is a big deal. You know, it's it was kind of when you said we just went live with the web page. I was imagining that was 1995, and that would have been a big deal. We just went live with a web page. That's right. On the internet. On the internet. Oh my god!
SPEAKER_02On the web page order, you can order Domino's pizza on the internet.
SPEAKER_01Oh and we paid some agency a hundred thousand bucks to build the pizza. That's right. Yeah, yeah, yeah. That's the other part that's taking it. My intern just built an entire multi-layer website um in the last 10 minutes using, you know, insert your the name of your airplane just, you know.
SPEAKER_02Oh my gosh. Yes. London's gonna be, and there's also the show called Hardware Pioneers Max that week as well. Oh so we're Monday, Tuesday there, Wednesday, Thursday. But apparently it's like the largest tech show in the UK. So wow, that sounds like a big deal. It's really grown like crazy over the past few years. And um, so yeah, that's gonna be a lot of fun, and we're gonna host a big party there.
SPEAKER_01Where are you doing? Where you do where is it?
SPEAKER_02You because there is that one's at the uh we're at the Pullman, so we're a conference, so you know we're like four or five hundred people, but the hardware pioneers max is huge, it's thousands, many thousands, and it's at uh it's at some business. Yeah, something like that. I have to look it up on the on their website, but um but yeah, they have a website too.
SPEAKER_04Oh, oh wow.
SPEAKER_02Yeah, there's two websites out there. Amazing.
SPEAKER_04Oh my god.
SPEAKER_02Yeah, the whole week's gonna be uh it's gonna be a lot of fun. So wow.
SPEAKER_04They probably had to work with Tim Berners Lee directly to get that going.
SPEAKER_02Yeah, I believe it. The whole FTP was all messed up.
SPEAKER_04Yeah, yeah, absolutely. Those were the days. Oh my gosh. You know, here's the blast from the past, and this wasn't that long ago. Well, maybe it was. So Microsoft has always had this MVP program for people who don't work for Microsoft, but they do lots of most valuable professionals, yeah. Exactly. But to let you, but yeah, but and so to to become an MVP, you're like an evangelist, maybe you wrote a book, you're speaking at conferences or working group, but uh little groups. Uh anyway, but back in the early 80s, no, no, back in the early 2000s, actually, when the Pocket PC and Windows Mobile was coming alive and the whole mobile revolution was starting to happen for Microsoft, and to become an MVP, it was all based on news groups, NNTP. This was still a thing in the early 2000s. So there were all these news groups for Windows Mobile, Compact Framework, SQL CE, all these different things. And your ability to become a most valuable professional was your answering questions for people on news groups, and we had dedicated news readers. Some browsers like Netscape included news readers built into the browser. This is, I'm taking a lot of people back in time here, and I know a lot of a lot of you youngins out there are like, I don't know what you're saying. What is a news reader? Newsgroups, NNTP. You know, we're not talking gopher here. We're talking something, you know, talking something serious. But it was in this century, people are still using N N T P to become an MVP. So anyway, there you go. There's your little internet trivia. Okay. That was kind of random. Yeah. Yeah, that was pretty random, man. Sorry about that. That must did that hurt. Did that hurt?
SPEAKER_03Yeah, it did. No, it didn't. It was historical. I don't know how to do it. It's historical. It was, but it was yeah.
SPEAKER_02Good context. Good context. Context. Just so you know. I assume they still have that MVP thing.
unknownI don't know.
SPEAKER_02They still have it, yeah. Yeah, they did it.
SPEAKER_04They still have it and they have the because I remember seeing a bunch of old friends flying to Seattle for the MVP summit a few weeks ago. Maybe. So I think it's still a thing, I guess.
SPEAKER_01Yeah, I remember going to those there were like thousands of people when I can remember saying that it's incredible. It was uh I forget which which of the um conference units it was in on on campus. I think it might be it wasn't in the EBC, it was too, it was too big for that.
SPEAKER_04But yeah, a lot of times they would do it at the convention center downtown. And then they would shut and then they would shuttle different, you know, your MVP for a particular product, and they would shuttle you to maybe the offices, the buildings for that particular product group, to to get your briefing, to get your NDA briefing so that you knew you were special, that you were gonna hear what the, you know, and you were gonna give that crucial feedback to the product managers that was actually gonna guide the roadmap, not anyway, but you thought they was.
SPEAKER_01So there you go. It was also Microsoft developer community. I mean, there were there were a couple of factors I think that kind of brought it into existence. Not least that you could make money by being you know an ecosystem partner of Microsoft back in the day at full cloud, there was actually a whole structure that meant you could actually make a living. Um, but the kind of exclusivity access, you know, the that as a model for how to build a developer ecosystem was extraordinarily successful. I think once once they stripped away, you know, what were messed with people's economics by basically trying to take all the money for themselves in the cloud, I think it kind of fell apart a little bit. People were pissed. People are demotivated.
SPEAKER_02Well, and also the the decline of the Win32 ecosystem really was like why attached. Yeah, they did. They did. There's less of a community. Uh it's hard to get a Microsoft specific community. Now, like as a community person, I can tell you like we have an edge AI developer community out there that transcends all these platforms. But there are still, you know, each company wants to have their own community, right? Yeah, there's a Qualcomm community, there's an Intel community, like everybody still wants their own thing, even though a typical developer now is going they're going all over the place. They're at Infinian and X NXP, and they're doing this, these tools, or whatever. So the age of the kind of product-specific community is a little bit. I mean, maybe Nvidia CUDA is the last bastion of that.
SPEAKER_01Um it's Scorpion and Frog, that's the problem. I mean I think it is very clearly understood, hopefully, that you know, IoT, industrial IoT, AI, whatever you want to call it, is a is a is by definition a jigsaw of of pieces. And yet, you know, I think companies understand that and know, therefore, that a you know, a homogenous developer ecosystem that spans multiple platforms is what is needed.
SPEAKER_02And yeah, at the same time, is it's their nature, you know, it's kind of I'll I'll I won't name names, but we had so we had a live stream the other day if you want to go look at it, and it was from a company and they're a great company, and so we they had this cool tools and they and they were their chip was based on the Ethos U55, I think. And I was like, Oh, that's cool. I said, So can I use this tool for other platforms that use the U55? Because it's a pretty cool compiler. He's like, No, it only works with our chip.
SPEAKER_05And I was like, Oh yeah, I guess that's the end of that conversation.
SPEAKER_02So I think it's like it's hard because there's a gravitational pull from a business perspective to sort of own the developer. But these days the developers need to be you know going all over the place and using all the tools. Yeah, that that's that's a I mean there's a there's a stunning lack of awareness sometimes.
SPEAKER_01You think about the companies that are at the roof at the forefront of driving all of this, there is a stunning lack of interpretation of what is happening, you know, at the coal face into the broader implications. Like you look at edge IoT, we've spent the last 20 years, you know, securing devices at the edge. That's been the you know, how do you manage the fleet, how do you deploy them, how do you ensure that nothing bad is happening? And we are now busily deploying autonomous agents across all of them, you know, without actually thinking about the consequences of that. You know, you now have and I well panic you saw with mythos, and you know, you mean you can you can build an AI agent that can systematically find and string together flaws in the system that hasn't, you know, has never been under that level of onslaught in the pot. Really? You know, it's like, yeah, now we're deploying all of these agents on these edge devices. What do you think of that? What do you think is that?
SPEAKER_04Uh yeah. Some kind of sci-fi movie gone wrong is what it's gonna be.
SPEAKER_03Which is like really funny because you can see all this stuff coming, you know? Um and it it's this whole drive to innovate for safety uh gets us into these dilemmas. And you know, the real power, I mean, we've we've said it so many times on the show, the real power of AI is uh force multiplying uh the threat actors and exposing vulnerabilities. I mean, you know, this whole mythos thing is hilarious, you know, oh my god, I you know, anthropic created the doomsday whatever, you know, model. You know, the threat is already there. It's out there right now. You don't need mythos for shit to go sideways. It's already going sideways, you know, and um you know it it it you know uh somebody cited a economist uh article that came out talking about how oh you know um the AI security issue is now becoming a problem. It's like, well, yeah, your your report is quite late. You know, you you should have uh reported on this three years ago, you know. Uh there's people who've been saying this and this that this is this will happen for a long time. You've just been talking to the wrong people.
SPEAKER_01Yeah, that's the problem. Is something an unintended consequence if it was a predictable outcome of a situation? Can you actually legitimately say, well, yeah, I didn't mean this to happen, when when you look at the the probability of that that outcome happening in a given set of circumstances, it's almost guaranteed. That's an unintended consequence.
SPEAKER_03It's it's it's and it's a technical so technically sound argument, and um you uh it's there. Um all this other crap is hard to prove that AI is valuable and transformative at a ridiculous level. You know, like you have all these jokers out there like uh, you know, um they're sharing this uh uh meme where they show how um the AI guys are taking trillions of dollars of value and uh in the AI opportunity and the telcos are getting like almost nothing, right? It's like the telco economy versus the uh the AI economy, right? AI economy is whatever, a trillion bucks.
SPEAKER_04Excuse me, did you say the telco economy?
SPEAKER_03Or whatever.
SPEAKER_01Which is haven't you got a haven't you got sorrow and sympathy for the poor telcos? Because they have categorically screwed it up through every major technology iteration, and yet they still manage to make billions of dollars. I mean, it's yeah, yeah.
SPEAKER_02No, I mean, you know, because everyone still needs to make phone calls, but that's it.
SPEAKER_01It's the same economic model as it was 30 years ago.
SPEAKER_03Right. I I I think they get yeah, they they get a bad rap for sure.
SPEAKER_02But you know, the thing is is uh uh yeah, Rob said what I was thinking, but that's yeah, yeah, they're they're actually pretty not very good.
SPEAKER_01You're gonna say they're they're you're gonna say the word losers. Um they just they had the means of production under 100% control of the digital economy, or you know, have whatever tone you choose. They had they owned the pipe, you know, and and yet still somehow, I mean I was in telco at the time with the iPhone came out, and it was it was like a dreadful realization that with the advent of the iPhone, more of the worst nightmares of telcos had come real, like everything was over the top, and the whole claim that we had laid to you know the data that customers were gonna consume and the share of wallet we were gonna get, yeah, that all of that had had had gone away. And yet, yet again with IoT, they did exactly the same thing.
SPEAKER_05Yeah.
SPEAKER_01In an effort to control, it was goes back to what you were saying, Pete. Open ecosystems, everything's got to work with everything else. The telcos in an effort to control and constrain to to you know keep hold of their their their birthright as they saw it, you know, they missed it. You know, the world the world water found a way. It worked around the telcos, and um and here we are. You know, there's that's a Jeff Goldblum was one of the things.
SPEAKER_04Nature finds a way. Nature finds away, Michael. They just went right around them. And remember before that, the telco said, we are the cloud. Of course, that's our birthright to be the cloud, because all the pipes go to us. They found a way to screw that one up too, didn't they?
SPEAKER_01Look at the PowerPoint diagram. That's us. We're there. That's us.
SPEAKER_04That's us.
SPEAKER_01Why why don't you understand that? Why is nobody?
SPEAKER_04Did you see all those cables going everywhere and they all lead to us? We are the cloud. Who are these guys from Seattle? Stop that. Stop that.
SPEAKER_03Yeah, but but they are they are the majority of IT spend. They get client share. So they are the trillion dollar market. Whereas AI, the AI economy in reality, if you were to um categorically be accurate, is uh tiny, right? I mean it and people are really talking about is generative AI, and that then that's uh you know anthropic and um uh open AI, and it's like cumulatively all these guys are maybe about at best like 80-90 billion versus a trillion. I don't uh you know it's so you're double counting too.
SPEAKER_04You're double talking about who's got the money and all right, so is it as my friend Steve Jennis would remind me, he's like, is it 1850 again? Is this a gold rush? Or is it a gold rush around AI? And who's making the picks and shovels who in the end are going to be the only ones who made money while all the other people are rushing for gold?
SPEAKER_01So you say what's happening, the semiconductor stocks?
SPEAKER_04No, Genesis make the picks and shovels. They make the picks and shovels, right? Levi's, you know, Jensen makes Levi's, yeah, right? All these other players, Levi's, all these other players who are going after the gold, this AI gold, this software, this thing, you know, it's not gonna work out for all of them. But along the way, a bunch of people got paid along the way, even though, even though we may see you know massive bankruptcies in the future for other players who are you know painting for gold. Yeah. Um, yeah.
SPEAKER_03Yeah. I mean, you know, you mentioned earlier the web, you know, the web development guys, you know, you have a uh website and you paid a hundred hundred thousand dollars for it. That was like what, Sapient? There was companies like Sapient and uh Razorfish, and you know, those guys made some good money uh for a little while and then they became completely irrelevant, you know, they got folded into marketing uh design and and now they're part of uh you know marketing properties. And you know, they be get double counted, right? Into you know, whoever thought that you know be this whole web design and web development market is gonna be sort of like what everyone's imagining with AI stuff, it ends up becoming just a you know a light augmenting layer of stuff in the economy. It's not the economy, it doesn't create trillions of dollars of benefit and uh you know uh GDP expansion globally. It doesn't do any of that shit, right? It eventually just becomes marketing.
SPEAKER_04You know, just all the all the CapEx infrastructure is the are those the picks and shovels, the data centers, the well, obviously we always talk about during the dot-com thing all the fiber. We laid a lot of fiber and we thought, oh crap, that was a waste of time. We ended up using the fiber later, uh and just not in that moment.
SPEAKER_01Um, but yeah, maybe the problem is it the economics, so yeah, I think it's day anything in the data center, whether it's active copper cable, GPU, rack cooling, you know, anything that there's a massive um demand and a shortfall there. I mean, you can't buy memory at the moment, it costs any money. But that's true. They're making that there's a lot of money to be made there. But the the the economics more broadly, I was talking to someone this week who said that the their token budget, so they're used, they're using AI code gen extensively in their company. It's like that, you know, the amount of money that I'm spending on tokens at the moment is the equivalent of three or four full-time high-quality engineers. And, you know, and again, the context of the conversation was around AI governance, and it's like, well, but I would know that if I was employing those engineers, the likelihood of some significant vulnerability being introduced, or at least one that I couldn't explain, is relatively low. Whereas with the money I'm spending, yeah, I'm moving at proper at 10 times the speed because my AI can do what that team of four would do in a month, in a day, but you know, that the the cost equivalent coupled with the increment incremental risk is like you know, I don't know what's gonna I don't know what's gonna happen next.
SPEAKER_03Yeah, you either you either um you know you either uh pass on that risk by not addressing it in by compromising quality, right? You just let shit, you know, you go MVP, right? Whatever. You just throw shit up against the wall and see whether or not you, you know, don't get sued or uh be uh asked for a refund. Uh or you know, um yeah, I and so it what you what you're starting to realize even with the agentic stuff, uh agentic stuff is all becoming programmable. You know, I was at Cadence. Uh there's this one slide that um Paul Cunningham, who is leading the whole agentic uh thing at um he's like a CTO, he's leading the whole agency push at Cadence. He has one slide on the bottom that says uh agents will require millions of lines of code. And these guys spent the whole time when they're talking to my agents say, well, you know, uh when you code the agent, I'm like, wait a minute. Code the agent? Why do you guys keep saying that? And of course, I I knew the answer, but they go, well, you know, this is all like uh harness and scaffolding and stuff like that. And you know, this is all like code. And so the these things are just becoming you know, spot automations where you're going to now typically at a task level, right? So when you look at chip design and verification, these things can be very, very useful when you're looking at a particular task or step in verification and you can force multiply and so you uh you know, that particular task so you can deal with uh exponentially growing complexity, right? So if you look at advanced packaging, you look at chip design, you know, as things are going to let's say gator all around. I don't know if you guys have seen a uh uh a uh you know gate all around transistor. It it's friggin' ridiculous. And the number of steps, uh additional steps, is uh you're going from like a thousand to three thousand, right? And manufacturing these uh these chips based on um you know gate all around. This is not gonna reduce the number of people people you need. It increases the risk because of the complexity, and you need these tools to even be able to approach this level of complexity and on the other end deal with that risk and not pass it on to your customers as shit products, you know? And this one of the things that I noticed that is completely missing in almost every talk track is fucking quality. Who gives a shit about quality anymore? No one does. No one ever talks about we have excellence and excellence and quality. Everyone just talks about, hey, let's just push shit out the door as quickly as possible.
SPEAKER_02That's well, you know. Dear, I mean, I think Leonard, you're hitting on something interesting. There is like in order to keep up with the increase in complexity, you know, this is where AI and agentic AI is going to help people keep pace. So there I think there's a frankly, I think there's a little bit of a myth. I mean, I think if you look at the data on GDP growth rate versus labor market growth rate, right? GDP growth rate far outpaces the labor. So there's this increasing gap between output needed and and labor available. And it's going to get worse and worse. Actually, we talked about this in San Diego too, which is why physical AI is going to be important as well, because we don't have enough people to keep up with the pace and the complexity. There aren't enough so even though there is a little bit of a doom thing about people losing their jobs, actually, we don't have enough people to do the jobs to keep up 15 years. And so that's why that's where AI is going to need to come in. You're right, it has to be done at the right quality level. So it's not like we're going to fire everyone and just use AI. It's more like we need AI just to keep up with the growth rate. And in fact, especially in the physical jobs, there's a significant tens of millions, uh, maybe hundreds of millions, of missed uh workers over the next 15 years worldwide, right? I mean, some regions have it a lot worse than the US. So that's where we think about AI sort of filling that labor gap uh over time.
SPEAKER_01But the problem is deployment, though. But you know, and a lot of the things around governance and everything else, at least the folks that so I've spent a lot of time at the moment talking to startups in this space, and you know, when we're doing diligence to their customers, you know, the the value of AI isn't the model. I mean, if if your AI doesn't control anything, it's it's basically just analytics with with better marketing. I mean, the value, the value of an AI model in an industrial context is what changes in the physical world. And so I think what I see at the moment is an increasing gulf between the perceived value of AI, which is the promise and the potential, and the actual value of AI, which is when people aren't really deploying this into mission critical systems because there are just some fundamental hygiene factors that haven't been addressed. Lobbying it into the real world as quickly as you can. Now, you know, horrible things have happened in the consumer space with connected devices for sure, but someone hacking the echo in my kids' bedroom is an order of magnitude different from someone hacking the controls infrastructure within an electricity electricity grid. You know, it yes, there are horrible consequences the first, but the latter is, you know, there are existential risks. And because people are not addressing those issues, the gap of deployment is these things are not getting deployed, and therefore the value is not being realized in the real world. And I think you know, if anything's going to come crashing down, the GPU manufacturers and the silicon guys are going to be fine. The economic model that Anthropic and others are facing of costs is one thing, but the myth starts falling apart when that value isn't delivered because people haven't deployed because we didn't do the basic things that we known that we need to see.
SPEAKER_02Well, there is that AI fatigue going on where people are seeing AI attached to things.
SPEAKER_03That shoe company, right?
SPEAKER_04All birds.
SPEAKER_02All birds.
SPEAKER_04I actually had some all birds.
SPEAKER_02I just bought a few, I just bought a pair a few months ago. I feel like I bought the last pair.
SPEAKER_04Yeah, you did. You did. Oh my god.
SPEAKER_02No wonder they were on sale. I was like, why is there no one in the store? Um you know.
SPEAKER_01Yeah. You know the event horizon, I think, when it could when all birds turns into an AI infrastructure business.
SPEAKER_02I was like, oh, they had a shop at University Village that was quite empty, but that's right.
SPEAKER_04That's where I bought mine at UVL years ago. Yeah. I was the only customer there, and and um I guess I'm you know when you're uh when you're assembling your venture capitalist uniform that goes along with your Patagonia vests, you gotta have the all birds. You gotta have it.
SPEAKER_02Yeah, the puppy vests, the all birds. I mean, you know.
SPEAKER_03Yeah, I mean, what's going on, Bill?
SPEAKER_00Hanging, hang. I'm just I'm just listening to you guys talk about the the the bro code outfit.
SPEAKER_02Bill's got an interesting t-shirt on that I can just barely see the top of.
SPEAKER_03Oh, look at that.
SPEAKER_02Oh, mackerel, that looks interesting.
SPEAKER_03Hey man, last last week was FAFO episode. Did you see it? The epidemic.
SPEAKER_00Absolutely.
SPEAKER_03Yeah. Debbie Debbie was Debbie was uh awesome.
SPEAKER_00I mean, look, we this is FANFO Friday.
SPEAKER_03Yeah, yeah, yeah.
SPEAKER_00That's true. Yeah, it's that the culmination of all this week's fuckery just comes home to rest. Come home to rest on Friday. We'll just rip it up and and redisseminate what's useful.
SPEAKER_03Yeah, there you go.
SPEAKER_04Uh isn't it interesting how people are just suckers? I mean, Allbirds just announces we have now an AI company and their stock goes up eight or nine hundred points.
SPEAKER_00And then they then they decide, oh, wait a minute, no, we're not keep that money, let's go.
SPEAKER_04Yeah, and then people find out that it's all BS and then it drops 800 points the next day.
SPEAKER_03I mean, this is like where we were with NFTs, right? It's the same state of JPEG. Silliness. Yeah, I know. And I'm gonna pay a million dollars for it.
SPEAKER_04But I own this JPEG and it's on the it's a block blockchain, so it's legit. It's Picasso.
SPEAKER_03You know, no, this is what people need to do. If you were bought into that stuff, you need to go back and watch the videos of you know, watch Bloomberg, CNBC, and how many serious conversations people were having about absolute bullshit. They should have been smart enough to figure it out and not take this stuff seriously, and this just goes to show how stupid people can be. And I it you know, it's blockchain and Kathy always says people are stupid.
SPEAKER_04Yeah, they are. You're turns out it's true, yeah. Uh including all the smart analysts on CNBC, right? Yeah. But just like where they went, oh yeah, oh yeah.
SPEAKER_03I mean, yeah, they should go and watch themselves and remind. I mean, you know, let's face it, that there has to be some force that pushes society to continue to improve. And part of that should be being a little bit ashamed for being stupid, right? There's nothing wrong with being ignorant, right? You can always learn. It's just don't be stupid. Stupid is not virtue. Ignorance is a neutral thing. You can solve that.
SPEAKER_00Now you're now you're demanding that if they know better, they should do better.
SPEAKER_01I mean, it's a tall order. But sounds like another teacher. You can call me cynical, but I do I agree. The herd, the herd, the vast majority of human beings are really dumb. But the herd of the herd follow, you know, the analysts that you mentioned, the the market exists on volatility. And so if there's nothing driving volatility, then there's no up and down, there's no upside to trade, there's no downside to trade. So the market has a vested interest in seeking out drivers of volatility and talking them up, maximizing the impact, and then following them on the way down again. And it every single technology trend that you could possibly point to follows exactly that. There are lots of people in the analyst community who know that, you know, and in the and in the industry and the software industry, who know that ultimately a lot of the house cars is going to fall down. Um, but it serves a very specific set of purposes, raising money and and and making money to talk things up, knowing that they're going to come crashing down. Now, ideally, when they come crashing down, hopefully you've got another answer back pocket that actually is gonna you know solve some of the problems.
SPEAKER_03Yeah, but that's the case. But yeah, and it causes bubbles, and we have to remember bubbles are basically a misallocation of capital. Yeah, due to misal and uh and the reason, okay. So going back to this whole lack of discourse around quality and excellence, right? Integrity is another thing that we're losing completely. And there's a whole there's a whole mechanism that's been developed over the year years, right? Whether it's like you know, creating um the proper governance for operating markets or uh supporting markets, those those have been withered away with deregulation, right? Got Glass-Stiegel is the number one victim, right? We don't have Glass-Stiegel anymore. So now, you know, investment banks can now commingle with commercial banks, and now there's this potential to create all these super toxic assets on top of a depositors' money, not investors' money, depositors, people who are saving in the banking system. Now that firewall's gone. You can commingle this stuff. The potential for the entire financial system to collapse if there's a you we have these continued bubbles is tremendous. And the victims become depositors as well. Before they were kind of insulated, right? Because you had this firewall, it's gone.
SPEAKER_04But they just always have a problem.
SPEAKER_03The integrity part is important as well, and we're losing it. And the the what's now dominating is this meme shit. Do you know what I'm saying? And this is this is straight up gambling. It's just like, how can I create like the snazziest little uh meme video or image that can trigger uh some kind of viral uh reaction that induces volatility or maybe GameStop, buy GameStop now? Go, go, go, go, go, go.
SPEAKER_00Like that, like that bit. You you got a point, you got a point there, right? So again, this week's musings, if you look on TikTok today, you will find any number of content creators talking about the fake uh phone calls that they've received. The you know, especially, I mean, it's it's predominantly iPhone users because FaceTime is the thing. And they're saying that a woman says that her daughter called her and said, Can mom, can you open the door? And she's like, you know, what are you doing? You know, why why can't you open the door? And she was like, I left my keys, I lost my keys, and everything. She's looking at this call on FaceTime, and it's her daughter, but her daughter's at school.
SPEAKER_05Oh yeah.
SPEAKER_00Okay, and there's and this is this is just proliferating. I've watched the ramp up of all of these uh content creators reposting, posting, reposting, posting this same thing, but in different areas. A father's talking to his daughter, a mother's talking to her son, and and I'm like, okay, so what is the foundation of all of this? And it is, oh, you guys need to be careful because these calls that are being generated are, you know, but where is the root? Who is getting down to the root of this? Who is actually making sure that the applications that are being misused and hijacked? I mean, where's the where are the guardrails?
SPEAKER_01But they the but the economic, I mean the economic model, if you look at what Facebook is an example, so you know, did Facebook know that its platform was being you know harmful to to users? Yeah. I mean, but that's natural damage in the greater cause of serving adverts to people to make cash. And and you know, when you have you don't have you don't have any guardrails, you're right, Bill. You don't have any guardrails to balance that that that singular intent to maximize you know revenue and maximize profit, to balance that out with you know, the uh I mean this is why regulators exist, you know, I face societal concerns on top of the naked bear drive to generate profit.
SPEAKER_00Well, see, but this but this falls right into that bucket of tech has outpaced policy every single step of the way.
SPEAKER_03Yeah, but I think there's gonna be this natural um appetite that develops for uh trust and trust platforms and trust um you know something you know something that will foster trust, right? Because there's so many things that are attacking it um that are off, you know, like um just creating so much noise, it's really difficult to tell what's true, what's not. Then you have all the threat actors now like hacks and deepfakes have been democratized um to a level that are is really gonna be become catastrophic at some point very quickly. So going back to telcos, telcos you know, um will have a great opportunity to provide that layer of trust because what do they they own the pipe? So how do they institute and implement uh these trust services, like a trust layer that uh uh allows you know um these communications to be trusted and And uh, you know, end to end. You know what I'm saying?
SPEAKER_01But the problem the problem is that I mean, I'm so I've got teenage kids, right? So you can have the the telcos, number one, people don't trust telcos, which is just a fundamental systemic problem.
SPEAKER_03Well, that's their fault.
SPEAKER_01You know, that is their fault. But when it comes to actually constraining what your child is looking at, um, or or I mean more protecting them that they don't get stuff served up to them. You have multiple layers. You have what you know, Verizon or whomsoever provides in terms of protection, you have the um protection layers on your local the Wi-Fi systems that you can control, you have applications that you can load on an iPhone or an Android that will block certain types of content. You have you have all of these different layers, and yet most 15-year-olds are technically savvy enough at this point to bypass them with a simple application of a VPN. And most of the policemen, the other parents to have have have literally no idea what's going on. I think the only saving grace is this, and I was droning on about this the other week, I think, it's this this this rejection of AI and a desire for authenticity and a desire for real things that is, I think, prevalent in the under 18 kind of generation, whatever they are, generation alpha D or whoever lose track. But but their their response to a system which is incapable of filtering out that which is not real is to reject it in wholesale. Yeah. Yeah. And I don't know what the implication of that will will be. You know, Battle Star Galactica. Well, yeah.
SPEAKER_02Mad Max.
SPEAKER_03Battlestar Galactic. No, you know, he's talking about being becoming uh um AI uh AI Amish, right? You know, just rejecting the colour.
SPEAKER_02Well, there will be, I think that will people will ascribe more value to more analog things over time because they will be more rare and more difficult to obtain. He kind of did that today already. Um, but you know, that's that's that's the way it's gonna be. People, you know, someday when it's all autonomous EVs, then you know, driving your own car will be like riding a horse today. You know, it's like, oh, I have this thing and I ride it around this little track, and uh it's kind of a quirky, interesting little niche, and that'll be driving cars, you know, 15 or so.
SPEAKER_00So yeah. Well, that's a that's another that's another sore spot I have because just like I don't believe that we went to the moon, I also don't believe that these damn autonomous vehicles are autonomous. There is somebody sitting back there with a fucking joystick that's driving these things, and people are sitting here talking about I just got into a Waymo and there's no driver. When your driver's sitting behind a goddamn firewall, yeah, that's driving you around, and he loses connection, and then you have an accident.
SPEAKER_03Yeah, you know, you have a point, and you know, it's uh so I was at Cadence Live and they had a panel on physical AI, and there's this um one of the panels was talking about where where they are in the state of uh autonomous humanoid robots and you know dexterity functions, right? You know where they are right now? They're trying to figure out how to get robots to fold clothing. I mean, that simple to us, which is absolutely a simple task. Hold your own clothes is incredibly hard. I'm gonna go put in a load of laundry right now, actually.
SPEAKER_01Yeah, you should I mean you should talk about talk to my 15-year-old about how easy folding laundry is because he seems congenitally incapable of drinking.
SPEAKER_05I can't use humans.
SPEAKER_01I did say to the yeah, I know you don't like AI things, but what about a robot that that folded your laundry for you?
SPEAKER_04Did those Artemis guys go around the moon?
SPEAKER_01Is that real?
SPEAKER_04No, it was filled in a warehouse in Burbank.
SPEAKER_01That didn't happen either. It was filled in a warehouse, don't they fit filled? It was a warehouse in Burbank. That's what they little cardboard box be in space. That's what it was.
SPEAKER_04Tim Cook made a lot of hay out of all the astronauts having iPhone 17 Pro Maxes pictures of the moon as they were flying around. I'm like, oh yeah, brought to you by Apple.
SPEAKER_00I mean, you know, we can't fabricate any of that shit, right? Not at all. I mean, look, I look, this is a this is the beauty of of that driving. I just think that a lot of people, if they actually don't believe, or they do believe, depending on which side of the fence everyone else is on, the likelihood that they're willing to express that statement is slim because they don't want to deal with the backlash.
SPEAKER_05But me, bring it up, baby.
SPEAKER_04Let's go. He's like, I don't care if it torpedoes my entire business. It's okay.
SPEAKER_01You are the you are the new net load, Bill. You are the new net smash the machine, smash the machine.
SPEAKER_03Yeah, no, but you know, going back to the the robot thing, your uh your comment, Bill, yeah, you know, I was at GTC and you know, all these friggin' kung fu robots, they're remote controlled. And you know, actually doing kung fu is easier, make you know, teaching or programming, actually, a robot to do kung fu moves. I that's like just video game shit, you know, it's like Tekken, right? Of course. Folding, autonomously folding clothing, totally different level. But I don't think the average observer understands the distinction. They will think, hey, we're at eight, you know, whatever, uh, autonomous robots because these things do kung fu. They don't realize folding full, you know, the less sexy and impressive uh task of folding clothing is actually the tougher thing and is more indicative of where we actually are with physical AI, which is boils down to robotics.
SPEAKER_02It's a good I was gonna say some of these dexterity challenges, which sounds kind of ridiculous, but they're actually really important because dexterity is also about picking strawberries and without bruises, and you know, you can't harvest food uh at a high yield without dexterity of physical AI.
SPEAKER_01But the argument I think this part of the argument about neuromorphic computing models versus you know generative, it the world the folding laundry is a perfect example of something that is inherently unpredictable. And you could argue that the generative LM, you know, but those side those approaches to AI are really about building a much, much better memory of what happened before and applying that pattern to the future world, which is great. If you're on a production line where every single part is the same size and everything, everything is predictable. When any as soon as there is an element of unpredictability introduced, then you create a problem whereby I'm done, I'm trying to anticipate the right outcome to us or the right answer to a problem that is unusual and does not correspond to that's a it's a it's a trade balancing of risk, it's a trade of, and that's not what most most AI to today is designed to do. Neuromorphic models where you know it's like, okay, we took a cloud-first approach to AI in the first place. We assumed there was unlimited resources, we could learn the we could index the entire internet and use all of that information to to recognize patterns on a grand scale. That model doesn't apply to edge AI where there isn't unlimited resource, where there is unpredictability, where you have to factor in risks that are new each time. Um and and in the world of physical AI, that's why I think you know Yeah, and and let me let me throw something else in there that will that makes things even more complicated, is the whole tactile thing.
SPEAKER_03You know, not you know, the world is not entirely solid, right? It has there's textures, there's different, you know, I mean, certain things are soft and squishy and other things are not, but um the physics are so complicated. Uh this this tactile element of perception is like really, really hard.
SPEAKER_02Talk to analog devices about it.
SPEAKER_04And you need like and so it just I keep trying to pick up water.
SPEAKER_03It looks solid, right? It looks solid. Yeah, yeah. Sometimes it is Science Friday, people.
SPEAKER_01Well, there's not there's not been, I mean, having spent uh you know much of the the last year running a company that is focused on tact to haptic technology that that that replicates the ability to feel and sense and and so forth, there aren't very many um folks focused on that space of how do you create authentic touch, either I was gonna say or deliverer. You know, how does the robot know how hard Gilbert needs to can be squeezed before it's trying to use vision to do that, and you can't do that.
SPEAKER_03And and you know, the interesting thing is the technology is in the music industry. They've been using it for years. It's not haptic, it's actually uh sensory, you know, like for instruments. You have different levels of velocity and sensory, yeah, yeah, and you reverse that, that becomes uh and I wrote about this six years ago.
SPEAKER_01Mems ultrasonic.
SPEAKER_03Yeah, and I highlighted this as being one of the important uh technologies because no one gives a shit about instruments, they think that the music industry doesn't do anything innovative. These fuckers do like super innovative stuff, and of course, nobody you know they get no attention for being uh you know technology uh leaders, but some of these these guys in the music industry, the level of innovation is crazy, and the kind of technologies that they're using for instruments is uh it's leading edge, right? And um, but all this being said, it gives you uh a really uh good idea of how early we are with all this stuff, despite all the demos, and you know, like what is that Olaf tipping over and uh at Disney?
SPEAKER_02Well, you see the latest uh what's the half marathon they're running in China? There's a lot of interesting uh what happened? Did it robots or robots running is uh you know good fodder for um blue burels and stuff because it's yeah, Westworld.
SPEAKER_00West world, yeah. I mean, but but you know, I mean, I have to say, uh Alistair makes an awesome point. You're you're doing something, you're you're actually solving a problem. It's not you know, I mean, but the majority of the people are gonna look at it and be like, that's not sexy. Who gives a fuck? You're solving a problem that needs to be solved. Yes, and gap of the ugly, and and you're and you're doing it with all of the all of the new tech that they're interested in.
SPEAKER_01Gap of the ugly, baby. You know what I'm saying? Yeah, where they where there is muck, where there's muck, there's brass, which is the only, you know, where there's shitty, boring stuff to do, those are the things where technology excels, and of all of the things that AI is being used for, you were talking about AI replacing people in all of these higher functioning roles. It's not really the point. What the what the right what AI can do really, really well is boring, repetitive shit that requires a high degree of attention to detail, but not a great deal of intelligence to determine what's going on. Like writing test scripts, like you know, going through and and and evaluating code and identifying issues that you can then go and look at. Like it's the it's the like for the professions like the legal profession, where you know, uh a good friend of mine's a lawyer, so I shouldn't say this, but 95% of what lawyers do is is boring crap. It's like wading through a whole bunch of data and and and try you can do that with an AI.
SPEAKER_02That's right, they charge you 400 bucks an hour for that.
SPEAKER_00You can't handle the truth. We we build we build all of our legal documents using Claude at first. I turned around and gave him to my lawyer, yeah, and he reviewed all of them. And I asked him if I had to ask you to do this for me, he said, Oh, easy, $30, $40,000.
SPEAKER_05Yeah.
SPEAKER_00And and he was done in four hours reviewing everything. Yeah, I asked him, I said, Did you use AI to review this? I'm just I just want to know how we're working here. You know, but what did he say?
SPEAKER_02Yeah, I used something called Claude.
SPEAKER_00I was right the first time.
SPEAKER_01Yeah. Yeah, I did the same thing. We saved $20,000, $30,000 a month by using AR, and I always felt slightly guilty turning to our lawyer and saying, Can you just review this? You know, there was almost like a crestfall of like, all right, then I don't get to do the fun part anymore. Writing it can charge and charging you 20 hours to use a contract. You know, I'm only doing the half. But I the my guilt was assuaged by the by the saving that we were generating, of course.
SPEAKER_00Yeah, no, I mean, look, I get it. I mean, it's just it's just the same way that we had to adapt, they have to adapt.
SPEAKER_03Yeah. Yeah, I mean, you know, here's the thing there's always uh uh what do you call it? Um underserved or unserved problem, you know, that we talk about all the time. And yeah, uh the interesting thing about chip design, and I'm going back to circling back to cadence here, and this whole topic of quality, one of the biggest problems you have in chip design right now is what they call they don't call it this, but I call it this. It's the verification deficit. You have complex designs that are very difficult to verify. So what they do is they do a half-assed job of verifying the design, then they tape out and it gets fucked up, and then they lose a shit ton of money. That risk curve is just going exponential now, and so one of the biggest challenges they have is um being able to verify with and to ensure quality on the other end, so you're not doing hundreds of millions of dollars of investment in a chip design, you know, chip design getting the tape out, and it's all fucked up. Um so AI will have the ability to support the scaling and force multiplication on both ends. It's generating a bunch of um, you know, let's call it chip design slot, but it has to be verified. And then there's the whole development lifecycle or design lifecycle that needs to be adjusted to get people in there um actually doing thorough checkpoint, uh design checkpoints, right? And so the you know, just like with software development, just because you have code generated to complete doesn't mean that you have like the entire SDLC uh handled by an agent or even a series of agents, you still need to inject people in the middle to make sure that the quality is there.
SPEAKER_01Same thing with the unprotection.
SPEAKER_03Yeah, and you're hearing less and less about hey, we're getting rid of people. You're hearing more about actually what uh I think keeping up. Yeah, uh Pete, you were talking about it's like people aren't the need for people isn't uh disappearing. You actually don't have enough people. And one of the things that Jensen said on stage, I fully agree with, and I'm glad he said it. We need more software developers, not fewer. And so for the in for the people who are telling their kids don't pursue a CS degree or don't become a programmer, you you're giving your kids the wrong message. But they're probably for more chip designers, they're not looking for the ladder.
SPEAKER_01Not to roll to the ladder of being cut off. I mean, we do so take chip design. So I've spent too long uh thinking about that. So, yeah, you can go through an incredibly sophisticated simulation of what the chip's going to do. You can go to FPGA and you can run a bunch, but and then you tape out and then you and you still have to go there back in and then verify. And it takes a couple of tape outs before you're ever going to get to a point where you've got a functional chip. And most of that verification is about dealing with the unpredictable, it's it's about you know understanding the impact of of moisture and oxidation and what random things could impact this chip in the course of in the course of its life. You know, those those and identifying those issues and figuring out how to rectify them is a highly skilled task. The problem is that if you don't have the bottom rungs of the ladder where people coming out with fresh computer science or electrical engineering degrees get to do the grunt work for a few years and learn the basic, how do you end up with people who've got that skill set necessary?
SPEAKER_03Yeah, but but that's a that's a dis that is a decision that is being driven by misinformation, right?
SPEAKER_01Yeah, I agree.
SPEAKER_03Misinformation. And this is why hype and nonsense are so horrible for organizations, because you end up making the wrong decision. You don't invest in the the entry-level folks who you should be bringing into the organization because you should be hydrating your human uh human resources you and and talent and experience base. Experience base. I think that's important. If you don't, I mean if you don't do that, yes, your organization deserves to disappear from the face of the earth. Fine. Somebody else is gonna take over, and hopefully it's one of the people that you didn't hire that takes your company out.
SPEAKER_01And yeah, well, how many years have we been talking about the shortage of C C um embedded development capabilities, particularly given HAI right now? We've been to I've been talking about 20 freaking years, nothing's changed fundamentally. There isn't, there isn't, you know, companies that should, you know, with the exception of some of the larger chip companies, perhaps, but the companies that that are going to be most impacted by that shortage should be you know paying it forward, they should be investing in college-level education, they should be, and yet they they're not. And and it is kind of like this there's this kind of vain hope, I think, that AI will ride over the hill on its white horse and solve the problem. But the reality is for all the reasons that we've just cited, that that that AI is not going to necessarily not going to have the requisite skill sets to deal with highly unpredictable situations and figuring out complex solutions. It's just not designed to do that. So who's gonna say this?
SPEAKER_03Absolutely. And who's gonna code your agent?
SPEAKER_00Who's gonna code the agent? Wait a minute, are you are you saying again that if you know better you should do better? What's go back to that?
SPEAKER_01That's maybe maybe that will be the line that underpins the downfall of civilization. Humanity reached the point where it no longer it no longer knew better and did better.
SPEAKER_03Yeah, they should watch Rob's uh videos on using tech for the what do you call it, the UN, what do you call it, sustainability goals. Oh yeah, yeah.
SPEAKER_04Nobody cares.
SPEAKER_03Nobody gives a shit. I know. You did all that stuff and nobody get all of a sudden like I spent all that time making all that stuff, building building all this tech to save the planet, and it turns out nobody gave a shit.
SPEAKER_04So wow, what a bummer. Oh well. All right, try again. On that note, maybe it's time to wrap it up.
SPEAKER_03Yeah, yeah, yeah. Uh you want to take us out? Feeling positive.
SPEAKER_01You're having me really cynical enough this week. We need to there need to be more misery.
SPEAKER_03Yes.
SPEAKER_04More misery, more uh cynicism.
SPEAKER_03No, there's nothing cynical about it. Come on, we're telling people to hire more people, hire more entry-level people.
SPEAKER_04Hire more people, hire more people, get more cameras on those guys going around the moon. Or did they?
SPEAKER_05Did they?
SPEAKER_00Oh, you don't trust it unless you were there.
SPEAKER_01One of them phoned me on FaceTime from Artemis to say hi. And I mean, you know, it looked like he was there.
SPEAKER_04Yeah, he pointed his iPhone down there.
SPEAKER_01And he he only he only asked me to give him my banking details so that you know I I could make a contribution to the next mission. Was that the wrong thing? That was the wrong thing.
SPEAKER_00I mean, come on, Alistair. If you know better, you should do better.
SPEAKER_04You know, yes, and therefore, if you know better, you should do better by supporting uh IoT Coffee Talk with our charity called Elevate Communities. Really underway. Really underway now. It's really almost there. Almost there. It's it I feel it. I've seen some emails that tell me that we're there, and so that's a good thing. Um, back to non-cynical positive stuff. Yeah, maybe you guys didn't care about me saving the planet with technology, so we'll come up with something new to save. And maybe you'll care. I don't know. I don't know. Maybe you just want stock picks. I don't know. Maybe, you know, I I predicted that Commerce One was gonna go to 2000 back in '99. And then I got in CNBC and I predicted that NFTs were gonna be the biggest thing ever. Who should you believe? I saw Tim Cook in the module of the Sparta spaceship handing out the iPhones to take pictures, you know, on the movie set out the windows. You know, it was amazing. Thanks so much for hanging out.
SPEAKER_01Oh my god. That was a motion that comes in. You forgot your phone. Quick.
SPEAKER_04They didn't pay you for you gotta pan down a little bit and say that you the roller bill to roll up and roll down the window on the part of this module. One thing could go wrong. Oh my gosh. Yes, but so you know, but part of it was real. You know, we did launch a big rocket, a lot of people saw that, and that went up in the air, lots of people saw that. So you know, but after some point, I guess you can't see it anymore, right?
SPEAKER_01The death in a way would have been if they all had sun tanned elbows from driving along in the Artemis with their elbow hanging out of the window. You know, yeah.
SPEAKER_04There you go. Therefore, Bill's right. Just stick with what Bill's saying, okay? I'm just saying, you know, he's just saying, you know, Bill's making millions on TV.
SPEAKER_02He's just asking questions. He's just asking questions.
SPEAKER_00I'm just I'm just I'm just trying to be here so I don't get fined.
SPEAKER_04Thank you, Marshawn. Thank you for that. All right, we'll see you guys next week. Have a great week. Adios.
