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
#314 Forward Deployed Nonsense
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
That's a great way to get the weekend started.
SPEAKER_04Love it.
SPEAKER_01Kick Comics. Yeah. And the lyrics of the weekend with a little AC. Living EC. Loving free.
unknownYeah.
SPEAKER_01Something about monetizing telco APIs. That's in the song too, right? There's something about that in there.
SPEAKER_04Yeah, absolutely. Absolutely. You know what the secret of APIs is? What is it? Stop talking about it. It's that matters. I mean, I don't know how many years I've been saying this, but you know what kills me now is now everyone's talking about, oh, token. What is it? Token. Everything's token this. Come on.
SPEAKER_01I know that thousands of years when the archaeological archaeologists dig up our bones and they will build monuments based on the content that we shared here. Because this whole thing of just putting names on things that have been old and non-monetized for a long time, and you just change the name, and somehow that drives attention and it keeps working.
SPEAKER_04Yes, classic one. Four deployed engineers.
SPEAKER_03Yes. Yeah.
SPEAKER_01Yeah.
SPEAKER_03Yeah. Palantir really Palantir dove into that deal and started talking about forward deployed engineers. And then I'm seeing other companies do it. I'm seeing our friends at Rackspace who are partnering with Palantir, and immediately we have created 50 new, we've trained up using Palantir training to create 50 new forward deployed engineers. Are they like consultants? You mean?
SPEAKER_01Well, Rob, but hold on, hold on. Because you know it's also the mediocrity of it all. Someone saw a movie about the US military and the acronym FOB for forward operating base. And someone said, okay, I'm going to grab this military term that sounds badass, and it's been deployed, and I'm going to make it a title. And now it's like forward deployed something, something engineer. It's like sold on. You already had the engineer. They were doing engineering things, but you're going to deploy them like physically, and you're going to forward them physically somewhere, like a consultant would, you know, with the client. Dude, I don't know. I I might put it on LinkedIn on Monday to see if I get any bots trying to hire a forward a forward deployable engineer. I'm not an engineer, but that's besides the point.
SPEAKER_03Yeah, people we have to, yeah. People believe that you're an engineer, though. Yeah. No, they think you are, man. No way. You're on the show talking tech every week, and they believe that you're an engineer.
SPEAKER_04Well, you know what? They have different levels of engineers. You have to recognize that, right?
SPEAKER_03I mean, um, you know, are you a you're I'm sure that you can prove to us that you're a prompt engineer.
SPEAKER_01You know, it's funny. Even I was gonna go, you know, straight up appropriation in and talking about level. I was gonna go with, you know, how we had levels in enterprise IT, and I will be like level, like best support number one, like your printer guy. I will help you out set up your printer. Printer engineer. Um, let's do that.
SPEAKER_03And you know, yes, the printer was obviously one of the most frustrating problems in IT back in the day. PC load letter, what the hell does that mean?
unknownOh my god.
SPEAKER_01Oh my god. The people in charge, and for you folks that work for a really big company with lots of printers, and the person that decided to name all these printers across different floors so close, and then you print something and you go to the printer you think you were using, and it's like three buildings away on the next campus over. Uh, what a nightmare. Yeah, but that's how it used to be. When we had real jobs, that you used to go into a physical place and you had to print things.
SPEAKER_03Well, remember Zero.
SPEAKER_01Yeah.
SPEAKER_03You missed that. You know, early days before everything became TCPIP, which this takes us back to the dark times. It as it turns out, in the early days when you had a local area network, let's say you were running Windows for Work Groups, or maybe an early version of Windows NT, and you had a printer and you wanted the printer to be shared so that everybody could see it, you used this protocol as like SNA, which is like a mainframe protocol, in order to talk to your HP shared printer, and people be like, PC load letter all over again. Yeah, it's it's bizarre. I don't know what happened. But now we have even now that we think we've been doing printers for so long, and we have wireless printers and shared printers, and we still have problems with printers. Oh, yeah.
SPEAKER_04I have um problems with smartphones.
SPEAKER_01Because things things are are are fragmented, even the paper size, A4 for me in the US just does something different here across the Atlantic because it's centimeters to inches or something like that.
SPEAKER_03It's weird. I don't know what people on the other side of the Atlantic were thinking.
SPEAKER_04They were rebelling.
unknownYeah.
SPEAKER_03So what city are you in right now?
SPEAKER_01It's a sovereign paper format. No, I am just outside of Geneva. Literally, I can see the uh literally I can from right here. I can yes, literally I can see it.
SPEAKER_04What the hell? And you should be sitting the other way.
SPEAKER_01No, man, because have you seen these kids what they do with like a little picture and and a couple of prompts, and then they find out where you are, and then they want to take my my Pacific Northwest Cascades North Face hat, man. Ooh, wow, and this is this is limited. This is like one of like two billion that they made.
SPEAKER_03I can tell it's limited. Yeah, yeah, that's awesome, man.
SPEAKER_04Yeah, all I can say is that that's a real weird pose. It's like a champion. I really like those.
SPEAKER_03I really like Lausanne, which is kind of around the lake from you a little bit, you know?
SPEAKER_01Yeah, no, it's uh it's cool, man. Uh you know, trying to um what was I trying to do? Um, as I try to build community here, and trying is the key word. Um you know, you have I I think uh we all think of Silicon Valley because, well, that's where you know traditionally the value has been at. But the reality is, you know, all these uh you know, smaller countries do have a startup scene. It's just that they don't make TV shows about them.
SPEAKER_02Yeah.
SPEAKER_01But it but it's quite interesting because uh you see a lot of them uh are heavily localized in that they're trying to solve a problem where the majority of the customers are here. Not every startup in Europe is thinking about US and China. Some of them are, the majority are, but there's a bunch of them that are like, oh, there's this problem that the vineyards on this part of Switzerland have, and that's my focus. Um Yeah. See, and that's another thing that do you know that there's uh um Switzerland uh is a big uh wine-growing country, but they only export about four percent of the wines. The majority, uh basically the other '96 is consumed in country. And I knew very, very, very little about Swiss wines until I came here. And I'm like, oh, uh, you you guys have the wines too. But um, and there's a bunch of things, you know, uh labor costs are much higher here. Um the laws are really strict, so things cost more to produce, and that shows up at the retail store when you might find a comparable wine from Italy or France, which shares a border with Switzerland, and all of a sudden the extra money, whatever that number is, doesn't really matter for the sake of this conversation. But it's very expensive to buy Swiss wine, whether that is nice Swiss wine or cheaper Swiss wine, when you have uh Italian and French wines everywhere. But yeah, it's a thing here, man. And there are startups trying to optimize things uh just because margins are tough.
SPEAKER_03Yeah, yeah. I love that area. You when you're you go through Lausanne and you're gonna keep on going um past Puilly. Anyway, there's that whole vineyard area, you know, and it's going all the way down the side of you know, down to the lake, yeah, to Lac Lamont.
SPEAKER_01Yes, you're speaking of Lacotte. You're speaking of Lacotte, and it goes from uh from well outside of Geneva, all the way to La Veye and Montaux, and yes, it goes all the way down to the lake. It's uh beautiful, really, really pretty. Uh lots of manual labor. Uh the only the only autonomous um tractors I've been able to catch in operation are doing tilling. Uh, I'm not saying that's the only one, I'm just saying that's the only one I've seen. Yeah, there's a lot of fists manually taking care of um of the vineyards, uh, you know, funny enough, in a low-tech thing. Um, imagine your traditional um like the plastic chair that we would get at baseball stadiums in the 90s, yeah, with three wheels, and these guys kind of just go from tree to tree and kind of scoot and go over and over, and that's what you see. Why? Because the dexterity that's needed today, uh, only a human can do it. And either the robots are not available or they're too expensive. So that's most of what I see uh here when it comes to touching the vineyards. Oh, with a caveat, I'm talking about the the the organic growers, which which have different regulations than the folks that do use and leverage pesticides.
SPEAKER_04Yeah, that's interesting because you know uh when you think of agriculture and the fact that for thousands of years, or at least since modern mechan uh, you know the industrial revolution, they still haven't figured out how to create a machine that can handle certain types of props. And and without distinguishing yeah, exactly. I think we a couple of weeks ago we talked about uh tactile, and so there are all these people talking about world models. You know, Google had their uh what is it the Google IO Google IOS and they announced Omni and they said it's our world model. And I was thinking, maybe not. And so there's this one guy, he's like one uh one of those AI influencer dudes. Um he uh line Lina, something or other, uh, he made a comment. Hey, you know, Google came out with a world model, and I made a comment. Is it really a world model? Or should we bring in you know Jan Lacoon and Feife Lee into this conversation and see whether or not what they think, right? Because they're talking about world models being the next frontier, and you're saying it's here. I know that and the big the thing is all these are based on vision, but there's no tactile stuff.
SPEAKER_03That's a good point. Because Google has had internally a world model for a long time, which they used to train the Waymo cars on. And so, you know, like if you read up on all the stuff they did with Waymo, every Waymo car has already done a billion trips in every city on every part of the planet based on their world models that they the digital twins they used to train the vehicles because they're like the vehicle in the real world will never do as many trips and go as many places as it will in cyberspace, in this, you know, just like when you hear, you know, Jensen talking about their what is it, omniverse or whatever their thing is. Yeah, yeah. Yeah, um, yeah, it's kind of interesting.
SPEAKER_04Yeah, and then you know, there's that um just tactical is hard.
SPEAKER_01I I saw today uh a um uh um a video from a Chinese OEM uh where they had um uh robotic hands uh putting together the the best it was smaller, but it remember how we put together pizza boxes, you know, this kind of like a few folds and then the pizza box you stop. So they were doing that with a smaller box for smaller components. But there's a but you know, cardboard, when you have uh a good production of carbo, you you get that uh it's easy to reproduce the same type of cardboard in terms of in terms of like density and flexibility, etc. Less much with it with a natural living organic thing like a grape.
SPEAKER_02Yeah, yeah.
SPEAKER_01Um and so yeah, so I I I think we're um I mean we we've seen this, Rob. We've seen I think with uh with peaches and mangoes, we've seen the robotic arm with a suction cup, right? Because it's got a you know like a little vacuum function. So that's that.
SPEAKER_04But I think back to the example of the mechanical C here, but it's not yeah, it's not a humanoid right, but you have all these people trying to pursue general purpose humanoid robotics with this hypothesis that somehow that's going to revolutionize industry, but the productive form factor, the fact of form factor that's going to actually be scalable, right, from a productivity perspective, looks nothing like a human. You it the human hand is not perfect for everything, right? And this is where I think as this physical AI hype cycle starts to go ballistic, this is what everyone's gonna discover. It's like, oh, we don't even have the spatial stuff right. We just have all these vision models that are giving us some you know reference on how to generate a uh five-second clip without having the the you know video, the generated video go off the rails from a physics standpoint. Oh, so this is all the other stuff we need to do, you know, and here's all the other data we don't have that has nothing to do with vision or language or any of this stuff. It it's like literally IoT level telemetry that then has to be. I mean, think about it. The the simple exercise of folding laundry, which is apparently the biggest deal. That that's like Turing test for physical AI, right? That we we talked about. If you can't fold laundry, you're screwed. And they're all trying to do it with vision, but you know, we can fold laundry without vision, we can feel it. We have a sense of spatial, we have spatial sense.
SPEAKER_03I can fold shirts really badly without robots.
SPEAKER_04Yeah, yeah, you know, do a lot of stuff without any vision, yeah. And guess what the what the proof is? People who are vision impaired.
SPEAKER_03Yeah, you know, I was thinking, uh, David, you're mentioning holding the pizza boxes, and you see people doing that, and some people are really fast. Oh, yeah, yeah. And um, I think it's better for people to, I'm just I'm such a power to the people kind of deal with some of this stuff we're trying to take over. And the best example I have that makes more sense is in I used to go to a mod pizza in Redmond, and uh if it's like kind of after 11 a.m. or whatever, you know, they have a young kid with Down syndrome in there, and that's what he does, and he makes the pizza boxes for mod pizza. I'm like, let's just do that. That makes more sense for society, and you know what I mean?
SPEAKER_01Yeah, well that that's a good example of you have someone who learns differently and adapts differently.
SPEAKER_02Yeah.
SPEAKER_01We definitely need someone to do this, and yeah, like this, why not? You had me a huge uh let's go. Um we need we definitely need more of that, not not less.
SPEAKER_04Yeah, and it's uh you know, it's not like you we're you're rejecting the technology. It's uh we're we're we've we're doing our homework constantly about this stuff. The problem is is when you have folks who are imagining, have the a vivid imagination of what technology can do, and waste other people's time by telling them you should do this ridiculous, nonsensical thing and waste everyone's time, waste everyone's money. You know, there's just certain non-fit-for-purpose conversation that you can have, and there's way too much of that, right? And that's why you're seeing all of this backtracking. And you know, if for the audience, if you've been watching IoT Coffee Talk, one of the reasons why you see keep hearing this AI adoption question is because nothing the people that said in the beginning has panned out. I mean, none of it. And where we're going is with agentic is stuff that is so pared down with agency stripped away from it, you might as well call it like an object-oriented class. You know, it's just big at the end of the day, is basic software that now you're running on a GPU, and instead of bits and byte-based, it's token-based, and it's I don't know, how efficient is it? You know, you're trying to calculate one plus one, and this thing is burning tokens like crazy just to get sense.
SPEAKER_03I really think uh Cheech and Chong would have a different perspective on tokens, you know, yeah, and how they should be using go to Oregon at the border.
SPEAKER_04And then find out what the price of tokenomics is.
SPEAKER_01I think it was you know, I think that if if social media uh had arrived uh, I don't know, ten years before it did, I think we would have more old school people that knew how tech worked calling out a lot of this stuff. Because sticking to vision and tactile, uh for today, in in agriculture, there's a lot of proven uh examples where you can automate some of the wheat sapping, right? Because it is vision specific. Uh it's looking for a specific signature on the ground, laser, boom, boom. Like that proven, it works, the it's deployed, it's it's working, right? And someone says, they go to the basics. Oh, uh a vision system, they identify this thing that should or should not be there, and takes an action and it's automated, and then they change the wheat for something else. Mind you, this person has probably never been in a farm, has never met a farmer, probably doesn't even have steel steel toe boots or or hike or or nothing. And then they just say, Oh, wait, you can do this and take it and and and use it for Rob, you're gonna love this, to strip barnacles of the uh US uh Navy uh Ohio destroyers and carriers because it it sap the weeds, it could absolutely sap the barnacles. Yes, yeah, why not? And no one says, like, hold on, guy, you're going from off-ground to underwater, gonna have a few issues. Um, also, barnacles, you cannot just sap them. You actually need like a physical thing, like and anyway, and we are full of this stuff because it worked here. Someone gets on a soapbox and says that it works over there, and people going like who are that guy or am I nuts?
SPEAKER_04Yeah, and so you know what you know why uh what the red flag is when you know the shit is about to hit the fan, you know what happened? Why board, deploy, engineers? Yeah. As soon as things get ridiculous, they they they they start to basically uh find that they're standing in front of the fan where with where the shit is headed. Uh yeah, that is what they do. We need forward uh we need uh uh forward deployed engineers. I mean, think about it for a moment. Uh one of the talk tracks, one of the talk tracks uh or key themes that came out of Google um I.O. Um, we need to drive adoption because people just don't know how to do this stuff, and we need domain experts out there working with companies to figure out how to use AI and Agentic. Right? But and and that's ironic because the way that it was sold in the beginning is that this was um, you know, we're headed to AGI. Why if this thing is so friggin' smart, why can't it just understand what we want? If it can it can enable intent-driven systems and intent-driven everything, how come it can't just work? And and you know, here's the funny thing is uh when you see what's being presented on stage at uh Google IO, it's it freaking looks like magic. I mean, it's freaking amazing, right? You go into the sessions and you listen to what what's happening, man, it there's such a huge disconnect. These guys are talking about uh freaking, oh yeah, we're now trying to figure out how to do uh SDLC enablement with agentic frameworks. And uh, by the way, when you get to verification and testing, uh, we won't be able to close the gap. It's probably gonna be at about 85% accuracy at some point, but right now it's like 30. How the fuck are you gonna be vibe coding with those kind of uh efficacy stats? Well, Leonard, it just the the You just you just made this you just made this an explicit that's why you need engineers out deployed with these uh these companies because this you have to engineer all this stuff at great cost. It's not in so hi guys, sorry I was it.
SPEAKER_00Hey so so so Leonard, are you meaning that we had this dream of humans augmented by AI, but in fact what we need is AI augmented by humans, right?
SPEAKER_05Oh my god.
SPEAKER_00No, but I mean maybe that's the way to solve the investors in AI problem. We just use humans to make AI valuable, so then we solve the problem. It's very simple. We found it. We found it on IoT Coffee Talk Show, guys. The audience, this is the answer to the world.
SPEAKER_04It's a tool. As soon as you think it's your girlfriend.
SPEAKER_00We talked about the we talked about the AI divorce lawyer last week, so I've been thinking about it. It's a very important thing.
SPEAKER_01Nice, nice. I don't like that divorce lawyer. I like the alpine divorce, the one where you take your spouse and you go on a trek, and then over there you just leave them. That is a thing. But but back to this AI thing. Uh you know, those of us that were doing, you know, IoT from back in the day, uh, our the the pitch really took a while to change. It was listen, data acquisition, and with data you can make a better decision. It was it was quite straightforward. Uh, because the data was easy to explain, you know, like we're getting into Rob's uh vending machines or my water meters. It was like, hey man, we got this value, this what it means, what do you want to do? The data had context. Because it had context, people would take uh actions. We have made this jump to AGI and all sorts of nonsense when if you talk to an factory operator today, the bit of their data not having context has not been solved. Like it just it just hasn't. So when we go into uh into financial statements and you ask, okay, all these companies with AI, like how much money can you attribute to your AI strategy? And it's like, oh well, one, two, three, four percent, because we fired all the engineers and we deployed them forwardly. Um, dude, it's we're skipping the basics, we're not reading, we're we're we're not reading through things, and we're just going to the end without uh you know verifying anything. It's nuts, man.
SPEAKER_04Yeah, and you know, and speaking of the context, it you know, with these physical AI, without the tactile, without like the smell, right? Because that was the other thing, the the fart test. Uh with all these things that humans have, and you know, the more I think about it, humans are incredible, right? Even in driving. Uh without all these senses that are perfectly fused together in a human being, you know, there is no better context engine processor than the human brain, right? The human nervous system, the human sensory system, it's so comprehensive. And that is always gonna be a huge gap if you try to duplicate that with a digital twin or IoT or blah blah blah, all these different mechan or you know, digital layers, if you will. It's just you're always gonna have a huge gap, and a huge gap in capability, huge gap in knowledge, huge gap. And so that's gonna that's going to make all this stuff unravel, and we're gonna see what we saw in IoT, where people have to prioritize where they invest and where they enable, and that dictates what a solution looks like, a viable solution. Everything else is just hypothetical bullshit that eventually uh dies away as it has, right? Um and AI is gonna be exactly the same thing. There is no exception, there's nothing remarkable about remarkable about this technology or exceptional about it in that regard. From what I see.
SPEAKER_02Wow.
SPEAKER_01So I don't want to just totally bash pigs and shovels.
SPEAKER_03There you go. I wouldn't want to totally bash this AI stuff. It it it's it's actually proven to be super valuable in lots of different areas, uh, but it's not always awesome, you know. Obviously, for writing code, it's amazing, you know, it's just unparalleled, actually. You know, that it's uh great stuff for sure. You know, and back to the, you know, I want this AI to be curing cancer and things like that. You know, everything that we build needs to be for the benefit of humanity. If it's not for the benefit, if it's not benefiting people, then forget about it. If it's helping us, empowering us, that's great. If it's taking everybody's job and putting them out on the street and destroying families, I'm not a fan. And yeah, that's all I've been seeing. With and I know we live in a bubble, and so all my friends all work for giant tech companies who are laying them all off in order to have more cash on hand for data centers. So I get it.
SPEAKER_01Um, so David said good good thing that good thing these data centers don't need any fresh water or or anything that we humans really really need.
SPEAKER_03Good thing they don't need that. Yeah, no, man. So, David, you said something a second ago about when we remember the the old days, the early days of IoT, and we would just talk about simple use cases, simple, uh it's hard to even call it analytics, just rules engine, a simple algorithm, and it solved the problem. Yeah, so I just want to tell you guys there's something you know I've been working on for years, and I've been a lazy writer, but I'm finally gonna get this. Um I've got a I've got a picture here that's gonna be on the cover of the book.
SPEAKER_04Okay. Oh wow.
SPEAKER_03And finally, it's gonna be out, it's gonna come out this summer. Um I'm gonna I'm planning on having the full end-to-end draft done. Kind of uh the plan is actually over this long Memorial Day weekend, actually. Um, and so it's simple. I decided not to connect it to some IoT platform I created. So I and it's generic, you know, and so if you'll remember, you know, I've talked about the sustainable development goals, the ones that are appropriate. I don't use all of them because not all of them make sense for IoT. Um, but it's a book of use cases for the benefit of people, right? Um, across poverty, water, both, you know, on land, climate, all this stuff. And it's use cases. And so it'll probably be, it could, it could be as many as over a hundred different use cases in this book. And they're all simple, they're all designed for non-technical people. It wouldn't matter what IoT platform you're using, or even if you're just uploading a JSON file to some storage and you figured it out. So it's like and it's so kind of I talked about in the past, you're gonna need this kind of device, like a microcontroller, these kind of sensors, this kind of communications. Um, I'll talk about it generically. And then in each chapter, though, I'll say, here's some possible, like if you want to go shopping, here are some, you know, that maybe it's this Arduino or this Raspberry Pi Pico, or it's this, you know, cellular modem, or whatever, things like that. I'll have that in there. Um, I was worried about that because initially I was having a bomb for every use case showing specific ones to go by. And you guys can tell me. I didn't, I didn't want to, I thought, well, that's helpful, go do it. Just you know, like like you'd see on hackster.io and things like that. But at the same time, I didn't want to play favorites with particular brands or companies that make different technologies. I don't know.
SPEAKER_01Um anywho, I I think you could be prescriptive and just say look it up, like, hey, if you need a microcontroller, Google microcontroller like you know, I just do a search for it. Yeah. Yeah. I want an autograph copy, but by Mrs. Tiffany, because I know she's doing the heavy lifting over there. So I need an autograph copy.
SPEAKER_03Yes. And then the uh, and so I don't I I don't specify an IoT platform, I'll just say, here's all the stuff you're gonna do. Your device is gonna send this JSON file that looks like this with all these sensors. Here's how often it's gonna send it, and then here's the dumb, stupid algorithm rules engine thing that you're gonna use to, you know, like thresholds and stupid shit like that, like you do in factories, you know. Um, and then the outcome and then the action you're gonna take. You're gonna send an SMS to somebody on the farm, or you're gonna show it on a dashboard, or whatever. Anyway, super simple. The real goal, though, is to really show how IoT can make a huge impact. There's no AI across, you know, again, probably it'll be over at least over a hundred different use cases across many of the 17 sustainable development goals.
SPEAKER_01And Rob, from now on, do we have to call you Rob Tiffany Public Offer, or can we still call you Rob?
SPEAKER_03Just call me Rob.
SPEAKER_01Yes. Okay, fine.
SPEAKER_03Yeah, you know, um, but I do want to, I do want once I have a draft, I would like to maybe send it to you guys if you're interested. Yes, and call bullshit on things or find problems. So because I you know how it is when you're writing something, you're in the weeds in the forest, and you can't see the problems yourself.
SPEAKER_00And so, yeah, please we we're gonna drop it into code and we're gonna help you with it. Oh god.
SPEAKER_03Yeah, and you know what? I I I used Claude a bit to help me when I was trying to find the right devices and right cellular and things like that early on as I was building the use cases. I was like, hey Claude, I know you're gonna need this kind of mic. And so, you know, it would go out and search for me and come up with a bomb list. And so, you know, it was it was helpful for that. You know, when I use it like search-ish, yeah, you know, that I mean, you know, it it it it's it it's a good use case for that.
SPEAKER_04You just again, it just it's not that we're I'm not against AI, I'm against um irresponsible use of AI. Yeah, that's it. And and but the problem right now is there's way too much of it. And I posted on one of my friends' um uh posts, uh comment on one of my posts that you know the the tricky part, the the trickiest questions to get to valuable AI is not what it can do, it's understanding and knowing what it can't do at any given point in time. We talked about this, right, Dimitri. Too many people don't know what it can't do. There's a lot of people imagining what it can do, including the people on every single tech conference stage out there. These guys, the talk track they have is all aspirational Star Wars stuff. And and this will get if you tune into what's being said on stage, you will be misguided. And and this is the disservice that all these guys are doing to their customers at some point, right? Because then your sales force has to sell something they can't sell, or they sell something, and that project goes to shit, it becomes a I mean, come on. Yeah, I have seen this stuff before and it pisses me off more than anything in the world because I've had to clean this shit up throughout my uh consulting career, and that's why I have no tolerance for it.
SPEAKER_01You sell the strategy contract, okay, and you deliver the PowerPoint. Oh, don't even get me started with that. Hold on, hold on, hold on. Remember, sell the strategy, hear your deliverables in PowerPoint or PDF, whatever you want, and then when the strategy is absolutely failing, blame the customer, update the quote with another 30 million, yeah, and another strategy project. Yes, and then send one of your we're not gonna tell you that it's a junior guy that just graduated from a school no one's ever heard of, but he's going to be a forwardly deployed uh consultant in there, and by the time the scam is done, that executive moved on to another thing and you just get to do it again.
SPEAKER_04There's nothing worse, nothing worse than to see a sales guy get away with his bonus and sticks the delivery team with all the shit, and then you have to manage a team that has to ghost their hours because the disaster is enormous, and that project, its profitability. And if you're lucky, you can salvage it. This is this is like the reality that people just generally, especially young people, they've never experienced before. And you have to, you have to experience this because this is life with or without AI. You know, this shit happens all the time.
SPEAKER_00But beyond you know, beyond that, and you you guys heard me say that is my my my personal t theory. I think it's uh the reason why we're gonna see or we have a risk of seeing things really bad at scale, is because this is the first time where we have a digital technology for which the the inter-interface artifact is not the technology. You know, when we started interacting with technology, we had screens, we had come online, you're very limited. Now it's a natural language thing. So you can imagine whatever you can put behind. You just put intelligent language and you could say, oh, it can do that. I just type a prompt. So and again, that I think that has a profound impact and a profound danger in how we look at this technology. Because you have the illusion that you're talking to somebody and say, Oh, that person has a 135 IQ, that your AGI is 155 IQ, yep, we've heard. So you you imagine everything. Yeah, without knowing how it works internally, what are the limitations, what is the context window, what is the memory, what is the statistical nature of the thing, blah, blah, blah. Nobody dig down into that because it talks to you.
SPEAKER_04That's the crazy thing. Even the researchers don't know how it works. No. I mean, you know, what's his name? D uh D uh Dennis Have uh CB. He was like interviewed just three weeks ago. That's what he was telling everybody. Well, we don't really it would be really nice if we knew how it works and what we see inside. And we're looking at the reasoning traces, and then we're trying to figure out where it departs or where it has um, you know, that circular the circular confusion, right? The um uh what do you call it? Uh the chain of confusion. It's real.
SPEAKER_03Yeah, it's that kind of like ball of confusion from Love and Rockets. Yeah.
SPEAKER_04Uh no, this is actually, I think, far worse if you trust it. But these guys don't even know how it works. Yet we have now uh all these salespeople out there saying that it's magic.
SPEAKER_03So they all say it's magic. And as David said earlier, he used the word tricky, and it's tricky to rock around, rock around right on time. It's tricky. So just so you know, we're on DMC, baby.
SPEAKER_01Queens, Queens, New York, the house of hip hop. And without bye-bye.
SPEAKER_04Are you taking off? I have to. All right, see ya. Adios. Yeah, it's nuts, man. Hey, you know, I was like thinking back um uh you know, when you were talking about your book, Rob, and you were talking about platforms, how very early on we had, you know, I had that debate with Rick about is the platform important or is it more of a reference architecture, right? Because everyone's gonna have different bits and pieces, nothing's ever gonna really in reality coalesce into a single platform. And you're gonna be writing about reference architecture, right? If you're gonna keep it agnostic, you're gonna tell more people what the bits and pieces are, that is a reference architecture. That's right.
SPEAKER_00But this is this is more use case, but this is more use case centric, right?
SPEAKER_03It is, it is more use-centric.
SPEAKER_00Yeah, so so it's so it's more kind of a dictionary of the potential of the technology. So it's the re it's is the is the functionality platform, is I mean, no, it's uh it's the applicability platform more than the technology platform itself, right? Which is, I think, what we're missing most of the time, because people don't necessarily I want to make sure I'm doing the right thing here.
SPEAKER_03Um I mean, I originally had it, you're gonna use this platform, and it's gonna go into this platform, and here's the analytics and magic's gonna happen on this platform. And then at some point I was like, I really need to finish the book. Uh the use cases, the because the book is actually it is like a handbook, though. You will take it and you will be able to take action as a person, as an NGO, as a whatever.
SPEAKER_04Um I see what your strategy is.
SPEAKER_03Yeah, it it it's it's got every step. I'm gonna go take care of this issue of poaching animals in a wildlife, and I'm listening for a shotgun, whatever it is, you know. And uh, you know, I'm listening for a chainsaw because people are cutting down trees in the Amazon, whatever it is. It's step by step. It's a handbook to actually act do the the only part that's left out is I just didn't force you to use a particular platform. But I know that every IoT platform will easily be able to do what this thing is. So it's like a cookbook.
SPEAKER_04It's like a recipe for all these recipes that you're gonna recommend for different treatments. That's right. That's right. It is a bunch of recipes, you're right. Absolutely. But even like if you go to cooking school, they give you a reference architecture for how to think of cooking, right? How to solve I mean, there's like certain basic rules, like there's certain things that you ideally don't want to mix together.
SPEAKER_00Yeah, because again, I I think that you know one thing that might be interesting is kind of an applicability map. Because I think in the case of IoT, what what I personally perceive, and and I'm I'm not as deep as you've been in all your career, Rob, but what I'm perceiving is that I think there are a lot of domains where the practitioners in the domain or the people that do the stuff actually don't know that there is technology that could help them significantly. So don't imagine. I give you a stupid example. I uh you know when I do the cooldown, I watch gold rush. You know, it's one of the stupid stories. And they have this very complicated, expensive uh equipment and they generate you know millions of dollars in gold. Honestly that ends up into Nvidia chips, by the way. But uh at some time you say oh it's traumatized. But always this big machine blowing and cut. And I always wonder why I don't have a bunch of little cameras and a guy in the room just watching that and you're not pushing a button on the radio when there's a problem. So I'm pretty sure that because they probably don't know that now you can do that with Raspberry Pi and cameras for 30 bucks, a device and a tiny little mesh network, and just deploy a bunch of these because they don't need to be hardened, they cost 30 dollars. So you just put five. And in a way, they only have as gold in it's in the Yukon. So it's only in the in the summer season because in the winter, so there's the weather is good. So the point is that there's probably a lot of area like that in agriculture where the guy runs a field and he doesn't know that there's those tiny little things that can help him.
SPEAKER_03Yeah, that's so true. You know, it's funny is um when I showed, you know, when I showed the picture to my wife about this, and she was looking at the way it looks in the font, and she goes, This looks a lot like, you know, she's a fifth grade school teacher. This looks a lot like books that we would give to kids to do. I mean, kind of the hey, let's go. And she was like, I wonder if you could create a version that's all pictures for you know kids in elementary school or middle school or things like that. Because it's simple.
SPEAKER_04My son would love that. And you maybe you can license it to uh code ninja, yeah, some of these um what uh after school education programs, you know, like my my son, he's totally into all this stuff, and he actually got invited. It's like weird. They have uh some sort of show being done by Netflix, and he's the youngest kid to be invited to this thing. And I'm like, how the hell that happened, right? But he totally digs this stuff, yeah. And um, I don't know. I I've shown you, I've shown you, I think I've shown you or fent you pictures of some of the stuff that he's designed when he was like two years old.
SPEAKER_03It was like freaking freaky stuff, right? It's gotten freakier. Yeah, kids could totally do this stuff. Yeah, I even had an angle. You don't know that I will follow this angle, this narrative, but you know, the whole thing, this sustainable development goals, and everybody agreed to it in 2015, and we were gonna have them all completed by 2030. You and I, we all know that none of that's gonna happen at all. It has not even close. And in fact, you know, what's been most painful for me, actually, there's not many people care about this stuff anymore, I feel. But you know, the the AI, the data, all the tech companies all abandoned their sustainable goals, all their net zero commitments. Everybody's just blown it off. They're like, no, we're going after this gold rush of AI, and I don't really care about saving the planet. And a lot of people don't even believe it's real anyway, and then a lot of people don't believe that anything needs to be saved. And so, you know, it's kind of disheartening to me. And so there was part of me that thinks, well, you know what? It's the adults who fucked it all up. Yeah, and so maybe this book is for you kids, and it's gonna have to be you kids are gonna have to save the earth because clearly the adults have not shown the wherewithal to do what's right because they haven't done it. We're gonna get to 2030 like in 15 minutes from now, and not a single one of those goals is gonna be accomplished.
SPEAKER_04Well, that's probably why Mike Schmid uh Mike Schmidt or Eric Schmidt got booed at the University of Arizona command yeah graduation commencement speech when he did was doing that. Yeah, I I he would that that dude is he's lost it. He has no EQ whatsoever, has no I saw the uh I saw another one.
SPEAKER_03A woman was uh the same weekend, she was doing a commencement at Central Florida and she started to communicate AI and they all started booing her, and she was so surprised. Like, what's going on? Um this younger generations are pushing back hard on on AI and and its impact on the environment. They're disheartened because they're all graduating from universities and they can't find a job.
SPEAKER_04Yeah.
SPEAKER_03I why would we be surprised that they'd be pushing back on it? It's impacting them directly negatively.
SPEAKER_04Yeah, and then even even the the generations uh before them are pissed off because you know, they can't buy homes.
SPEAKER_00Yeah, can't they can't uh well the the gener the generation before was impacted by globalization where we outsource all the jobs sorry, and now the new generation and now the generation this generation is impacted by us replacing them by machines and by AI. So I mean it's if you think about it, it's it's fuel for revolution. So it's fuel for you, you're gonna reach a point.
SPEAKER_04It's yeah, you know, it that something is not right, and um the other thing is when you look at the economy, uh so you have like the agriculture economy, you had then had what the industrial economy, the uh information economy, right? The or a service economy, information economy, and I don't know, some jokers are calling it an AI economy. But what's what comes next? Bullshit economy. So everything that you're your influencer, your economy produces is just a bunch of nonsense. Uh synthetic to fuel synthetic uh impressions uh generated by uh humans as well as you know humans having relationships uh you know with um imaginary AI bots, and then who knows? They might even gin engagement by having bots fall in love with each other just to create an on-player character uh but that's the that's the that's the key.
SPEAKER_00That's the key, Leonard. We just need to find a way to start giving money to those bots, so then they're gonna start spending it between each other to do something, and maybe we'll be the middleman in the middle making the money. So maybe that's the idea. What are you talking about? But think about it, then we we here's the great idea. We we we we we we uh uh we we we outsource economy and money making to all those AI agents. Let them do an economy between each other, generate the cash, and we just take a tax. Like the Apple model. We take 30% tax. Well, you know, here we create hey it's the creator tax. We created you, we give you uh data centers to run, you're you're making money doing things, and now we're just taking 30% tax, and you go do your stuff and we just retire. You know, here's the solution, people. Here's the solution, people.
SPEAKER_04Here's the ironic thing about the whole maltbook thing. Open claw wasn't the big deal. It was maltbook that was the big deal. Nobody talked about maltbook anymore. They talk about open claw. Why? Because you can sandbox stuff and now you can go and stick this in your enterprise and be safe.
SPEAKER_02Yeah.
SPEAKER_04Yeah. Right. That's just they completely ignored the thing that was much more important is what was happening with Maltbook. And what what can happen with Maltbook in your enterprise?
SPEAKER_03I mean, it is the height of silliness. You would create a social network for agents to be on and have oh my god.
SPEAKER_04Oh my god, that's now they have like agentic uh agentic commerce, which is all a big deal.
SPEAKER_03So now maybe maybe like agenic like tender and agents hating each other and swipe right or swipe left or something, you know. Why not? It's just it's beyond silliness. Oh my gosh.
SPEAKER_00It's actually it's imagine Rob. It's it's it's it's kind of the idea of the matrix, but for AI bots. Yeah, we don't harvest energy, we harvest money that they do with their interaction in their whole virtual world. This is super we need to build that, Rob. We need to build the infrastructure for the matrix of the AI bots, give them a virtual world where they generate crypto money and we just take a tax on it.
SPEAKER_04But see, you know, the the that's where I think sovereign currency is going to become uh important again because uh crypto it supports that model that you're talking about, right? When you ever whenever you hear about these um chatbots going bonkers, it just opens up a friggin' wallet and then you know starts doing whatever monetizing or transacting using crypto.
SPEAKER_03Remember how ridiculous? Remember how ridiculous we thought it was when the NFTs became a thing? Yeah, and it's like seriously, but so many people are like, we're all in. I just paid for this JPEG for 10 million dollars, and it's mine. I put it on the blockchain, and it's only gonna be mine. I'm like, dude, you're an idiot. You just bought a JPEG. Well, it's unique, but it's unique.
SPEAKER_00I mean, it's mine and it's on the blockchain.
SPEAKER_03It's amazing how stupid people are and how they all get hyped up and oh my god, I've been just dumb shit. Oh my god. Yeah, it's crazy.
SPEAKER_04No filter, that's the problem. People just don't have filters these days. Yeah.
unknownYeah.
SPEAKER_04I mean, it does get you in trouble, and then you can't have any of the serious conversations about how to actually move the ball forward.
SPEAKER_00And actually, it's it's a shame because it's a shame because the NFT technology to me has a very promising aspect. One aspect, for instance, is that if you pair NFTs to existing physical assets like art, you can actually build a model when whenever the art the art is sold from one person to the other, there's a certain portion that is collected and that goes back to museums or or or you could or you could add to your point of sustainability, whenever something is exchanged, there's a portion in the smart contract that is collected to actually exceed. Because the the problem, the reason why nothing is happening on the whole sustainability thing is because it doesn't make any money. It's just a hard fact. So yeah. So there was some very potential, and that's that's the saddest part of it, is that I think there was a lot of interesting potential with the concept of NFT to attach that to physical objects. But yeah, we just used the hype, it went up, some people made money, a lot of idiots lost, and now it's gone.
SPEAKER_03So some kind of sad ape or something, some kind of ape with some deal with NFTs, and I don't know. There was you're right, it just got ridiculous. And you're right, attaching an FT to a real piece of art, just like we should have a digital twin attached to my a Tesla or whatever of machines, you know, that says this is what the machine is digitally, here's the real one. Yeah, but but the problem getting the owner's manual for your car.
SPEAKER_04The problem is that you you assign you allowed it to become the currency uh of the value of that piece of art rather than having a beam mechanism for provenance. And that's a huge different thing. Oh, but this is I mean, you know, the the whole thing. So they hit when they associate the crypto element to it, that's when everything goes sideways, and then it becomes like a joke. And and you know what? We've been talking about this on IoT Coffee Talk forever.
SPEAKER_00But the the root, the root cause the mother, the mother of all those problems. I mean, sorry guys, it is in the uh when we stopped with the gold standard. As soon as you start disconnecting the representation, the monetary artifact from the physical backup and start trading it and doing derivatives of it, you open the Pandora box and you end up with all the problems. The derivatives of monetary instruments is the problem. If you want to solve the world, that's the problem. You have to regulate that extremely carefully because otherwise you have derivative of derivative because the NFT is just another derivative at just another level, which is even more complicated.
SPEAKER_04So there's a proof point to that. There are two countries that realize that very well India and China. Yeah. They are buying gold like crazy, they know at the end of the day they're going to have to do that.
SPEAKER_00And the only yeah, and the only crypto they do is actually similar to the NFT, is a stable, the stable coins. So it is they are really inventing the gold standard, is what is really and it should be actually at a geopolitical level very concerning because if they do that well.
SPEAKER_04Yeah, yeah, yeah. No, I mean like what Iran's doing, they're going to bricks and asking them to basically flex on uh you know not using the US dollar anymore. Right? I mean, these are all things I mean, we've i in line with the whole crypto um perspective we've been debating here. That is one of the key points that we've got to do.
SPEAKER_00Yeah, but it's it's actually what what we're saying is that it's I would say the double worse. Because if it's replacing the US dollar from another currency that has the same mechanism with no real backup, it is yeah, it is just a power Swift. But if it is replacing the standard of the dollar by a crypto thing that is backed up by real things, but then that's even more and even worse for an impact.
SPEAKER_04That is the interesting thing, because that BRICS does not consider crypto a currency. Exactly.
SPEAKER_00It's it's a it's a mechanism. Yeah, it's a mech it's a mechanism to exchange and follow the provenance of real value in an extremely immutable way. Which is only just the same thing happened to NFT. The concept of cryptocurrencies is a smart and interesting interesting in its immutability and fundamental security. But it needs to be backed up by something because it's value you're supposed to collect. If you just put it on the wide as, oh no, it's not connected to anything, yeah, then it's pure speculation.
SPEAKER_04And it's just I mean it's it's all of the aircraft carriers and nukes, right, Rob?
SPEAKER_00Oh yeah, but yeah, but I have aircraft carriers and nukes. Yeah, but the yeah, but the problem is no, and I don't disagree with that, but the problem is that you then depended on another system. You have a system that is dependent on another system, and that's very tricky. It's not dependent on intrinsic value, it's dependent on the system. Oh, we're we're we're we're we're um yeah, we're all walk around. Yeah, we're on the meta, we're on the meta, meta, meta today.
SPEAKER_03That's right. You're so right about that.
SPEAKER_00Uh my guy. Cool, guys. Yeah, yeah.
SPEAKER_03Yeah.
SPEAKER_04Let's wrap it up. So remember, take it seriously, at your own risk. We're just decompressing. I don't even know if we're having fun anymore.
SPEAKER_00I always have fun. No, but my my attitude now is it has I have to take it the fun way, otherwise, yeah, otherwise it's it's too depressing. It's uh we we it's a comedy, it's a musical, and uh yeah, and we we we as we said in France, we're in front of a big cliff, we're very, very happy, and we're just about to make a big step ahead.
SPEAKER_04Yeah, and you know, in wrapping up here, I think the key takeaway for everyone is that if your vendor shows up or your solutions uh provider shows up and says they need to deploy forward deployed engineers, that is a red flag, and you should probably run away at that point because your vendor has no idea what they're doing, and they realize they need to get an AI scientist of some kind uh on site with you to conduct their experiment on your business. So remember that. Forward deployed engineers. Yes, they are just they're not even consultants at that point. They're exp they're gonna OJT on the experiment they're gonna do on your business. So remember that. If they say they're a consultant, that's better because a consultant ideally has experience and they can help you.
SPEAKER_03At least you know what you're getting with a consultant. I'm here to extract money from you in exchange for PowerPoint.
SPEAKER_00Yes. And those PowerPoints are c are all generated by cloud nowadays, all generated by cloud. Yes, remember, increase increased revenue, decreased cost. Remember, increased revenue, decreased cost.
SPEAKER_04Right, right. Yeah, and so if you made it this far, that's awesome. The other thing you have to know is that Rob is writing a book, really wonderful book, with apparently a uh fifth grader level uh book cover graphic on it. But he is going to create a comprehensive cookbook of IoT solutions to help address the thing that nobody cares about now. Nobody cares, which is sustainability and and uh the SD, uh what is it, the SD, the sustainable development goals. Development goals, yes, that nobody cares about. But eventually, probably this year, because we're gonna have a you know a super duper El Nino that's going to fry the planet or inundate the planet, you will probably give a crap by the end of this year. Uh or you know, starting in July, I think is when they predict the super El Nino. And so, yeah, remember, um, IoT and technology used with proper purpose, right? Yeah, can help save the planet and your home.
SPEAKER_03There you go.
SPEAKER_04And keep your family safe. And so this is why we are uh established elevate our communities. So check it out. Um, I think we're almost ready to go, right? Now we're gonna do the work that Stephanie Atkinsons did to uh get that whole uh whole uh project um revamped in a nonprofit, and um that those are the things. Hey, maybe your book can be the basis for everything.
SPEAKER_03Yeah, the cookbook would be helpful for what we're trying to do. Oh, definitely.
SPEAKER_04Yeah, yeah. So and remember to uh follow us here at uh IoT Coffee Talk, like, share, and um comment, you know, tell us what you think. And if you want to vent, vent, just say all kinds of stuff in the comment section. We'll love it, and we'll respond and we'll talk trash with you as well. So um, but uh thank you for watching, and uh we'll see you next week.
