mergeconflict409 === [00:00:00] James: Hey everyone, before we get started, I want to remind everyone that you can see Frank and I live. Well, minus the live part, but you can see our faces live on screen. Not actually live, but recorded this podcast. You could actually hear me say these words, like, but you could see me too. If you head to youtube. com forward slash at merge conflict, F M Frank, you know why? Because we're almost at 500 subscribers. Which means like we could almost monetize soon. [00:00:29] Frank: Uh, amazing. Uh, we, we had to start our own channel because we got voted off of your channel and, uh, we've had to rebuild our audience and we're doing it. Thank you everyone for subscribing, hitting the like button. What are all the YouTube things you're supposed to say? Uh, do those. Thank you. [00:00:46] James: Yeah. Share, share it. You know, if you're listening to this, Just go and subscribe anyways. Now I will say we're, we're close on the, there's, there's like multiple, there's like 8, 000 different ways to even try to monetize now on YouTube anyways. But we're, we're one of four requirements done. So when we get two, we unlock some tiers and we get all four. Then we all unlock the full tiers of our money making scheme to make at least pennies off of YouTube. [00:01:11] Frank: It is our goal. I mean, one penny would be a good start. Um, you know, they make the gates sound so clear, but I've, I've passed these gates before on other channels of mine. And they still said, no, your channel seems a little derivative, but at least here we give our heart and soul and passion. Every episode is unique because honestly, only about half of them do I know what we're going to talk about coming into it. It's [00:01:36] James: very true. Uh, Frank and I tried to record this specific episode on Tuesday and I was so tired. That I basically fell asleep. Uh, and we, we did talk for like an hour and a half. Heather, I came out and Heather was like. You guys didn't record a podcast. What do you mean? You were been in there for an hour and a half. I was like, we just stared at each other for an hour and a half. I don't know what to tell you. We just, we're just like, I don't know what to talk about. And then I basically. True at [00:01:58] Frank: all. We did, I believe a half hour Patreon episode also. So we're, we're just promoting things here at the beginning of the episode. Let's promote our Patreon also. Hi, everyone. We have one of those. [00:02:10] James: Yeah. If you want to support us, you're like, wow, we, I hate YouTube. I'm going to not subscribe and not hit that like button. You can go over to patreon. com forward slash MergeConflict. fm. See what we did there at MergeConflict. fm and you can become a patron. You can give it a free trial. If you want, you get bonus exclusive episodes. You can also see our face there for the bonus exclusive episodes too. We try to put out a few of those every single month and you get all the podcasts early, as soon as we possibly get them. So give that a look, all the links in the show notes below. And with that, Frank. Talking about not talking about anything else that the internet is talking about right now. We gotta talk about AI devices. [00:02:50] Frank: Is that what we're doing? Okay, because I, I, I, I, I've been, I've been staying out of the AI devices thing, honestly. Um, I, I think, I think AI is one of those cross cutting features. It's going to be in everything. And I think it might be a little too early for AI devices. You know, I've been working so hard to get AI running on phones and stuff like locally, you know, they can be internet devices too, but it's early days. It's the wild, wild west out there. I keep saying that with AI, um, but it is, it's the wild west. Uh, Is there, is there news, is there news that we should be talking about? I know I made at least one tweet because you suckered me into looking at one product. [00:03:32] James: It's kind tweets once, then that's probably what we're going to talk about on the podcast. Well, no, I think here's what's fascinating. I want to talk about, even before we talk about the devices, here's why I think I was, I am interested. I'm always interested in new hardware. I actually sent you today, The new Anchor, Solix, X1, like PowerBank, it's the whole house, like the Tesla wall, you know what I mean? With the solar panels, dude, the things. I am just always intrigued by new technology. I always have been. I know you have been too, because we're just tinkerers. We love the latest bits and we're willing sometimes to shell out 3, 500 to try those latest bits. Frank, that's you. [00:04:15] Frank: Not, not often, hopefully. Hopefully. Hopefully. [00:04:18] James: No, none of them, but you know, we both, we both went in on the M1, uh, pre M1 days of the developer box, you know, from, from Apple. I had, um, some early Google devices from going to Google IO, uh, the, the good, the Google Q audio. I was a big kickstarter of trying all sorts of electronics. I can tell you almost, in fact, none of those things that I kickstarted are near me. But they were cool. I mean, maybe one or two, but like, I like to think what's new, what's outside the box. You know what I mean? You do inspire [00:04:49] Frank: me with your Kickstarter. I wish I did more Kickstarters, but the one I ever did was a complete failure. So I've, I've, I've totally dropped the ball on them. Uh, I, I do love your passion for tech. Usually our passion for tech involves, uh, Nintendo knockoff products also, but hopefully we're not going to be talking about those today. Uh, it's funny, like in the AI device world, I, I've been a sucker for my own self because I like to do that. That wasn't a good phrase. I'm a sucker for my own self. Just note that one. Um, I like to buy little IOT devices that would be like additional hardware for like a Raspberry Pi or even a smaller chip, things like that. These are basically GPUs, but tailored to solving AI specific problems. Um, they're usually very restrictive, restrictive to the point where I never seem to use them. Like, you have to compile your models down to them. And so I actually have my own little box of AI devices that I've Not touched in a long time, to be thoroughly honest. Um, but I, I can't help but to buy them sometimes because I, I want little AI devices. [00:06:02] James: Yeah. I want little smart, little awesome devices that are doing really cool things. Even if I have to interact with them or not interact with them. It's why I've been intrigued by robots, right? I know you're intrigued by robots. I've been intrigued by having little robots do things that I don't want to do in the house. It seems that I just end up doing them anyways, but I'm still intrigued by them. And this technology seems to get. Better and better and faster and faster and smarter and smarter. And now with generative AI and chat GPTs of the world and the Geminis and the clods and, and all the things, uh, the Chuck Norrises of the world, you know, all of these things that are out there. people want to do more than give them. An interface on your computer or on your phone that you're typing to. Right. I think that's really what the catalyst is, is what else could you possibly do in this scenario? Like, and even chat GPT, right? There's these extension points and, uh, there's a whole bunch of other things going on. We just saw GitHub announced GitHub workspace is workspace. I want to say, have you not seen this? It's magical. Oh, you got it. I signed up for the beta, [00:07:09] Frank: but I haven't gotten. And yeah. So I've only seen screenshots. Yeah. It looks like, uh, you just give it all your code and you're like, fix it, make it better and, uh, magic. That's that's what I understood. [00:07:21] James: You guided along the process. You prompt it, you prompt engineer at doing the work for you. Uh, yeah, it's, it looks really, I'm really excited about it. But again, there's all these different interfaces, right? And there's, there's a bunch of different cases for them. So when I see new, exciting hardware, like you said, I've, for the most part, been in the space of smart home and game related hardware, mostly that I've kickstarted even in front of me, right? I also have, obviously I was really big in Elgato products into like these light bar products that I reviewed on my channel, just anything that enhances my life. And I think. That's why phones in our pockets are the most sort of valuable thing nowadays. Cause they enrich our lives in many ways. They also don't enrich them in other ways. Like, you know, things are bad. Like social media. Purely good. Purely good. They're purely, there's no evil. They're purely good. 100%. But I think, you know, when you look at the device range, like obviously tablets, you know, and PCs, they enrich our lives in different ways. Some more than other based on who you are. So I think like. The interesting part is there's a lot of good on phones, but a lot of things that I don't like on phones too. Like, I don't like that it's always buzzing, always doing stuff. There's like so much with it, which is why the dumb phone revolution, you know, is trying to come back, for example, but to me, it's like, what can enrich my life? Is it a little robot that's sitting in and mopping the floor? Is it something that I'm attaching to my bike to make it smart? Is it a little screen? Is it a new video game thing that's emulating stuff? Is it a balance bot that is just balancing for no reason? Frank, I don't know why it's balancing, but it needs to balance for some reason. [00:09:01] Frank: Well, you know, that one's a great example because my balance bot is actually an AI. I didn't program it to balance. It learned to balance. And so you'd really have to ask it, why, why balance? Why do you balance? And it's going to say to minimize the R score. Dummy. Um, I, yeah, um, I, I, I do say the phone is purely good because anything bad about it is what humans have done to it. The technology itself, I think is. And how you use it. Yeah, exactly. But it can be a lot. It is a lot. Um, I hate the notifications on my phone. I'm sorry. All my, all my friends know I'm terrible at responding to text messages and everything. Except for you, James, because we're professionals. We get this thing done. Um, but everyone else I'm really terrible with because it's too many notifications and all that stuff. I, I, I think that's why we all like the, um, Amazon, Google, uh, uh, Apple family of dinguses, uh, because they're fantastic. It's something you don't have to have in your pocket. I went so far talking about technology gadgets. I bought an, uh, Amazon dingus that goes in the car. Cause I still want to talk to the, the Amazon dingus whilst in the car. Um, just honestly, like, you know, okay. That was. Pre Chat GPT. Nowadays, like, Chat GPT is just so outdone it before. I, I would like the, those dinguses to catch up. Cause they are, they're, they're a generation behind now. Um, but yeah, those different form factors, they're, they're important. The phone can't be the only thing. And honestly, it can't be the watch because I hate putting on the watch. You know, like a necklace or a Star Trek communicator comm badge, or, you know, hilariously enough, I was just watching. Um, Zoolander. Oh yeah. Yeah. He had the, he had the tiny little flip open phone with a tiny antenna. I was like, you know, I make my living through a touch interface device and everyone go buy my apps on the iPhone. Oh gosh, that tiny little Zoolander phone. It didn't look so silly this last time I watched the show. [00:11:09] James: Uh, no, I, I mean, I think it's fascinating. I think, you know, when we look at interfaces and what an AI actually wanted to be, it's sort of this combination of. Many things, you know, I actually don't believe that the current interface and some of those new products that are out there and how they're, and what they're even capable of doing today is maybe the problem with it, but I think it's really a kind of combination of everything. Right. I think if you look at Ironman, right. And, um, this is done that fireman. Okay. You [00:11:36] Frank: lost me. Okay. We're okay. Jarvis. We've gone Jarvis on this. Okay. I [00:11:41] James: mean, Okay. So here's the thing. When we think of these devices, and I think why the, the phone is really powerful because the phone, you tell the phone and you connect to all your services, you put everything in the phone, right? So my life is in this phone. I can get a lot of, basically, I can turn on the sauna. I can adjust my heat inside of here. I can basically do anything. Now the interface to do that, I can ask it anything. I can open a browser. I could ask Bing. I could ask chat GPT, I could do all those things, right? The interfaces are what sometimes half the problem, which is why you like dinguses. And if all the things can connect to the dingus, then that works. Not all the things connect to the dingus, but it wouldn't be great if there was this sort of. You know, Jarvis around us that, that knew the context that could construct stuff, um, and knew about our lives and could schedule and the whole, the whole thing, right? Imagine if you're putting on your, um, vision pro, but it was, had the Jarvis and you opened and imagine Jarvis was in there and you're like, Oh my God, this is amazing. So when I think of it, it's like, really all the things that I can do on my phone. I want to be able to do that, but I want to do it in a visual way, a streamlined way. I want to do it in a way that sort of. Is more natural than grabbing a phone. Cause right now grabbing a phone, opening an app, it's not that hard, right? It's a little intrusive. It's not like it's hard to do. Right. So how would you make that better? And then also something that is better than the phone, because it is in a virtual space around you, right? You can more easily interact with whatever you're looking at. And that's sort of the dream. And we're very, very far away from that. Right. But that's why I think these devices. Interest me, and I'm not sure if the device is what went in and we're talking about two devices that I'll talk about at least, which is the HumanePin and then the RabbitR1, uh, which do very similar things in different form factors. I'm happy that they exist and that someone's trying something, but I'm not sure what they're trying is actually a problem that needs to be solved. Does that make sense? Does this make sense? Why I've said the Jarvis reference when you watch Iron Man, Frank, you watch Iron Man and you saw Jarvis for the first time. What did you think? I want that. Oh, I want that. That'd be great. How cool would that to be? Yeah. To have Jarvis just hanging around. Speaking to you. [00:14:03] Frank: Well, Jarvis was two things in the movie, which I found very interesting. Jarvis was sadly the home butler. I feel bad that Jarvis had to do that. True. But you know what? Our home AIs, they're going to have to be butlers. Um, but Jarvis was also, um, uh, technical. Assistant, which I always found intriguing from a conceptual level. Uh, when Tony Stark was doing his little designs, there was an AI assistant helping them along. Tony's like, I want him, uh, gold and AI actually does all the work. And that's kind of the future. Uh, so I think a lot of these devices, we're talking about the modalities of the interface of communicating with a computer. You want many of those. Um, like, like the phone is easy. But, uh, it's a little screen in your hand and it gets, we're humans. We, we, we recognize patterns too easily. It gets boring fast. Uh, if I have a question, I want to ask it to the Amazon Dingus. The chat interface, I, ChatGPT, like everyone's first goal is, of course, this should just be audio. You should use your own human voice and it should speak back to you in a vaguely human ish voice. It's kind of an obvious thing. It's from Star Trek. It's from all the different sci fis out there. I'm not even convinced that that's a perfect. interface for it either. Like, it's slow. Speech is slow compared to reading. You can read a lot faster than you can speak or listen to tolerate someone speaking. So, uh, when friends come over, they always crack up because I have my Amazon dingus speak at a very rapid pace because I have zero patience. I'm like, spit the information out. Don't, don't waste my time here with all your Yeah. Um, And then, uh, you bring up like virtual reality of, I'm so disappointed how little VR, or sorry, uh, how little AI there is in my VR world. We gotta match these two things together because they aren't fit for each other. Yeah. In, in that world you can present. Any modality, you can present a little rectangle that you're talking to. You can have a speaker system. You can have a virtual actual assistant person or a little robot or something. Who cares? Um, that world presents a lot of opportunities for AI and we haven't barely tapped into that. Uh, so I, I, I'm a big believer in that there should just be multiple ways to do this. I don't want to have to go to one central terminal to access my AI. It should be everywhere I go. And a lot of different forms. [00:16:35] James: No, that, that is a great point, um, on it. And I think that it's right. You're right. Is one of the drawbacks I've always had is I want, I like precision. I think that's the thing that hinders me in some of the, in some of the AI's, uh, chat experiences that are out there, no matter which one you're using. They're very verbose. They like to tell you everything about everything that's going on. So I sure James, I would like to respond to your last point. Yeah, exactly. Well, I think, you know, it's like. Uh, I was watching a demo of the RabbitR1 and like, every time you would ask it something, it, because it's slower, it would respond to you telling you that it's going to do the thing. You don't need to tell me you're doing the thing because it's told you to do the thing. You know what I mean? Like when I go to Google or Bing and I type a thing in, it just goes and does the thing and then responds to me. And those interfaces have gotten so good. With the glance ability, right? When I go and say, what is the population of the Netherlands? It's like a 17. 7 million. You're like, oh, cool. Like, and then they'll give you a little bit of different additional information, blah, blah, blah. If I ask that to a bot, it's going to be like, let me tell you all about the history and blah, and like this, and that is going to be a whole thing. And you're like, and then the, the, and then no matter what, if you're reading that or you're. NET MAUI, James Montemagno, Xamarin. Forms, Xamarin. Forms, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, NET MAUI, James Montemagno, Xamarin. Forms, And then point it at it and say, what is this? Right. Let me do this. And it would be things that they know what it is like a Dorito chip. Like you don't need to like, let's be, let's be real. Right. I did enjoy that. And I did like this part is for the Abbott Rabbit R1. I think it was the verge, maybe it was in gadget. They were walking around DC, I believe, and taking photos of different monuments and say, Oh, do you know what monument this is? Can I get some more information? Oh, that's cool. Not in DC. I could. I know most of the monuments in DC, but not everyone would. So if you were visiting DC or you're visiting, we were just in Vienna. Every five seconds you turn to some beautiful thing and you're like, what is this? Right? That would be. Uh, pretty cool. Um, the problem with that is that then this thing is yelling at you back, like you're saying, right? I just would rather have something glanceable. That's easy to do. I'm just out and about walking around. You don't need to yell at me. I guess if I had earbuds in maybe, but let's say that that did work good. You had earbuds in and you're walking around. It's like a personal assistant walking you through a guided tour of a city. That's pretty cool. You could just also build an app for that. Um, You wouldn't need any app. The app is the AI, right? It's the whole thing. There's a whole point of it. So that is kind of a cool experience that when I was watching some of those reviews, I'm like, Oh, this is cool. I also don't need to tell me that this is a little robot thing and this is a Dorito and this is a AirPod. You know what I mean? But it also has to be like 100 percent accurate, you know what I mean? That's all that matters. [00:19:41] Frank: Oh, it's just so 2016, the, the object recognition thing. [00:19:48] James: Do [00:19:49] Frank: you remember at our old meetup? I had a demo doing object recognition with some of the old early networks. Um, I believe we pointed it at a clock and it said it was a monkey or something. It was not 99 percent accurate. Um, actually none of these image recognition networks are. Um, it's tricky. I, I think that that was one of the first tasks we trained AIs on. And they got really good at it because for a while, honestly, it was the only task they were getting good at for a while at the size of networks that we were doing back in the day. And so I've never loved the tell me what object this is networks. They've always been like, like, Hey, a solution in search of a problem. Um, because humans are really good at object recognition. It's, it's one thing we excel at. So, I, I think the big use case for all of that stuff has always been for the vision impaired. Um, night time, you could argue too but these things aren't good at night time so they're not, they're not good at that. Um, but for the vision impaired, but even then, uh, single object detection, it's, it's, it's, It's not useful. What you want is [00:21:00] James: context. [00:21:01] Frank: You're in this environment. It's this kind of building. There's these doors, et cetera. The floor is clear. That's [00:21:05] James: what, uh, seeing AI, I remember seeing AI, which has been around forever, you say 2016, right? It's like, I remember those early demos where it would be like. It would, you would take a photo. And again, this is a, like basically a talking camera for people that are visually impaired. And it would be like, there is a woman playing fetch with a dog in a green field with a, you know, outside, you know what I mean? Like, it would describe what's going on, descriptive. That seems Like a great use case for it compared to, Oh, I'm at this museum. What is this thing? Oh, there's a plaque that says what it is, right? You know what I'm saying? Like, [00:21:43] Frank: right. Okay. Uh, I mean, the statue thing is cool though, but I'm a little doubting of their accuracy because I mean, Uh, at least in America, you see one guy with the tri cornered hat. It's one of 16 people. That hat was popular for a while. The AI is going to have to make a guess. It's hard to know these things, but maybe it says like, uh, you're in Boston on this block. It's probably, uh, the statue that you're looking at. Uh, [00:22:10] James: and it's actually a word problem. There's this other aspect for, I kid you not though, I was actually trying to do this recently. Uh, we were in a hotel in, in Vienna and there was like this, I think it was in Vienna. Yeah. And there was just like this picture of this guy. And we're like, I wonder who this guy is, right? Like who this person is. So I took a photo. And, and Heather and I were there. I'm like, Oh, I bet I could put this into, uh, chatgbt or into Bing chat. Right. I bet it would tell me the thing. Copilot. The problem is it's a picture of a person's face. And it's like, no, can't do that. Okay. You know what I mean? It's like, it's like, Oh, I can't do anything. It's like, Oh, we can't, you know what I mean? So, so the privacy concerns, security screens, all this other stuff. Oh, that's interesting too. Oh yeah. I think so. Can't have people. Even though it was a [00:22:51] Frank: painting. I get why they wouldn't want to dock someone in the modern world, but if it's a painting, chances are, it's not a modern person. [00:23:00] James: Then we reverse image search and we try to kind of find it, but then couldn't. So [00:23:05] Frank: I was going to say, so how did, because that's our, Old standby technology. How did reverse image search work there? No good. [00:23:12] James: It found some similar photos. I don't think it was the same, like similar paintings, you know, of, of this person. So it must not have really been like a person, you know what I mean? It's like in a hotel or whatever. [00:23:24] Frank: And honestly, right now, um, uh, the way all these, uh, sorry, behind the scenes, the way all these networks work right now with images is called this clip embedding. One of the embeddings, go see our tokens episode. Um, And that embedding is not trained to distinguish between faces. It's more trained like how to recognize faces, not distinguish faces, if you get my meaning. So that would have to be added into it to add some more context to the pictures and all that kind of stuff. [00:23:58] James: Okay. So these, okay. These devices they're doing that stuff. I mean, I think that's one point of it. I think that that's probably the worst Part of it, to be honest with you, I feel as though between the long text, between the accuracy issues, right? Some of this will just get better over time, right? The, the chat GPTs of the world and the things they will trim it down and it'll get more accurate, right? [00:24:25] Frank: Let me interrupt you real quick because it's a feature I don't even take advantage of. Enough myself with all the good AI chat apps. They let you set a system prompt or a pre context, whatever they happen to call it in your app. And it's funny at one point, the chat GPT mobile system prompt was exposed. Cause this is something OpenAI doesn't really want to show you exactly what it's doing. But very clearly in their system prompt, they kept telling me, I keep your answer short and distinct or like right, right to the point, get to the point. Cause we're on a little mobile device. People don't want to read eight paragraphs to find out there's eight slash nine planets in the solar system. [00:25:07] James: That's true. That, you know, that's, I think that's an interesting part too, is like. The idea that you could give this thing additional context, right? For example, if you were sending this image or asking questions, the prompt under the hood could say, Oh, James is walking around in Vienna, Austria, and it would know my exact location, right? Of where I'm at. It could send information creepily up. Maybe it could be, you know, more vague, but it would know that it could then prompt and give it the context. Um, Uh, more information instead of here's a generic photo. It could be like, I'm at the Hotel Bristol in Vienna, right? And here's a photo that here's a painting that I found of a person. Tell me who, who's in this photo. [00:25:51] Frank: Context. I think the phrase is context is for Kings, but I'm going to switch it context is King. Um, someone told me at Facebook once, uh, we were, we were talking about. Um, face recognition, so not detection, not is there a person in this photo, but who is this person in the photo? And there's a lot of people, human beings on this planet and just given a random photo and if it's told to identify which person, human being this is, it's probably not going to get it. It's too hard of a problem. Way too many people out there. Could be anyone. Lighting could be off. Who knows, but context. Um, you give it your social graph. You give it one or two nodes out from your social graph, and all of a sudden, an impossible problem. They said it was at least two nines, like 99. 99 percent they could hit whoever that was in your social graph. So context matters, for sure. [00:26:51] James: Think of it like this is we just landed back into America and we have the global entry. And now the global entry is vastly changed. Before what it would do is you'd, you'd put your passport in, it would scan it. You do fingerprints. Cause you had to get fingerprinted before a person would verify you. Now you don't even take out your passport. You walk up to a machine. You look into a camera and it gives you a thumbs up and you just leave. Like pretty much there is a person to revalidate. That's it. But context, right? How many people are in the global entry system? A lot of them, right? How many people are in the global entry system landing in Philadelphia in this gap? They know, they know, so they can trim down and now there's what? A few hundred people, if you, a thousand people, maximum of the busiest airports tops. So context, right? And that's what these devices, and that's what the whole thing is. A lot of them are missing. It's the context that I think is so important about it. And even in the, the, we were talking about the tokens and all the stuff, the, again, if I'm asking. And that's why I do love like, you know, where actually, when I think about how I use AI today, I use it in, in work context. Like we have it built into teams, we have it built into the office suite. I use it there the most, right. And also get a co pilot, but, but I would say, like, this is a me, like the one thing I do is like these AI recaps. It's like, I'm in a meeting or I missed a meeting. And it was recorded. It has the context of the transcription. It has, it knows who's speaking. There's all this information. And it just was like, here's everything that happened in a concise way. Bulleted point. In fact, what it will do is it'll do bullet points. And then if you want more about that bullet point, you can do a dropdown to the bullet point. But it'll give you like five bullet points and then action items. Right. And it has all the people that are in there and you can time code it. So great. And it's like so awesome because the context that's there. Right. And I think when Tobe was doing the demo and then I've done the demo again, which is. About basically creating your own get up co pilot with semantic kernel and, and, and, and open AI is you give it the context, you give it the code and you say, look at this specific code, this as the code, right? And then you, you, you go and you use that a hundred percent. So that context there is so important. And I don't, I don't know if these devices, and then maybe that's the problem of these devices. It's trying to use the entire context of everything, um, that's out there. But. That does seem to be one of the biggest problems is that's why Jarvis is so good. Right. Because Jarvis knows the context that Tony Stark was working in, but also then can reach out of that context if needed. Right. If Tony was like, Hey, I want to build. NET MAUI, James Montemagno, Xamarin, Forms, NET MAUI, Xamarin Forms, Dot Net Maui, Xamarin Forms, NET MAUI, NET MAUI, NET MAUI, NET [00:29:44] Frank: MAUI, NET MAUI, NET MAUI, NET Tony Stark could have afforded to train his own network. So that thing's probably just tuned exactly to his data. You know, you, you wouldn't need to worry about how would other people feel about these responses. You're like, no, no, I wanted to talk this way, exactly this way, know about this information. He would have enough money to actually train one that could do it. Uh, I, I think that, um, the, the, the context thing is going to be important with these devices. And I, this is why I'm excited about Apple getting into this game a little bit. They're, And I don't think it's going to happen on what, May 7th or whatever. That's not going to happen. But at WWDC, I think that they will have a product because they're in this enviable position of they're running the phone, that little device that we keep complaining about, uh, that has all the information in the world, every bit of information about you, it's right on that device and they have access to it all because that's how the security system works. And so it's, they're in a great spot. Uh, to, to fill in more of that stuff. Now I have noticed that like, ChatGPT, and I don't agree with this. I don't like it. It's, it now says it's using memory from your previous conversations. And I kind of liked each conversation being its own sandbox, because I know that low level, that's kind of what's happening each, that's the thread. It's building up this context in this world. And so I can, I can play games with it. Like. Don't pretend I'm not me. Get me out of my bubble here, you know? Talk to me like I'm a five year old and explain this problem to me. Um, I don't want it's memory of me. Sometimes you do. Sometimes you want it to know, Oh, Frank, he lives on an island and he has a lot of black mold he's trying to get rid of. Here's a bunch of suggestions and that kind of stuff. Other times I, I, I'm too embarrassed to tell people about my black mold. We're not recording this, right? [00:31:37] James: Uh, it's okay. We had to remediate plenty of black mold. So if you need black mold help, I can help you out. Wear a mask is number one. Uh, [00:31:45] Frank: I [00:31:47] James: think I might be sick for the next couple of days. Nah, it's all right. The spores, you know, it depends on how severe. Did you take pictures? Cause if so, send it our way. Um, but we did remediate some, some plenty of black mold and we did have one bad spot. We actually ended up. Removing the drywall itself. Burning it down itself. [00:32:05] Frank: Burning. Okay. Yeah. , [00:32:06] James: uh, and we didn't burn it, but we remediate it. So there's, there's spray you can get like Fred Meyers. Oh, I see. To one to. So we did that first. Then we actually cut out the drywall and we put new drywall in because one wanted to see is the black mold behind inside the wall. It wasn't, thank goodness. Yeah. You have to, due diligence, gotta open that wall. Due diligence. So let's talk about the device form factors. Frank, we have phones. This is a form factor. We've got two new ones. One, the AI pin that you clip it on, you talk to it, you tap on it. It has a little projector. I kind of like the projector. That's kind of cool. I'm 100 percent here for this one. [00:32:45] Frank: A little projector. [00:32:46] James: This is pure Star Trek. That's very Star [00:32:48] Frank: Trek. I'm not fashion forward. I don't really want to wear a pin necessarily. Could have been like a cool clock necklace or something like that. I would have worn that. A pin you can put on a necklace, a chain. So you know, it might go there or something like that. But then it starts to feel like a police security camera. I'm talking in circles. Either way, I like this one because it reminds me of the Star Trek communicator. That's why. [00:33:12] James: Yeah. I agree with you. I think that it is, it, it is nice. There's a lot of extensions like that you can do. You can clip it, you can put it on your belt, you can do whatever It's like yet one more thing to attach to your body. You know? That's the other thing. It's like I'm putting on my, I already hate putting on my watch, hate take, taking off my watch. Mm-Hmm. especially to charge it. If I ever have to take off my watch to charge it and I'm not going to bed, then I just get real upset. 'cause I'm like, it's a watch. Like you should be able to be on a muscle. I should be able to tell time for at least. 18 hours. [00:33:45] Frank: Sorry, sorry to interrupt you, but yeah, mine started dying around 6 PM and so I never wear it anymore because what's the problem? If you can't get through a day, you're not a watch. [00:33:54] James: That's a hundred. It's very, very true. It's very true. You have to be from, from when I wake up to when I go to bed. Job [00:34:01] Frank: one, while [00:34:02] James: I'm [00:34:02] Frank: awake, tell [00:34:03] James: me the time. Tell me the time. It's all, you have one job. One job. I, it's either, like, you know, You put eight lives on [00:34:11] Frank: these things. [00:34:12] James: Oh gosh. All right. I, I like the, I like the concept of it. I think, obviously, it, it even kind of confused me why there's even a camera on it and more things. I'm not positive. I don't know. I think with this device, the one thing that would be kind of cool is like a voice memo notes, right? Imagine just like, Oh, I have an idea. I'm just like, you know, you could talk to it and then you could file it somewhere into doing something again. That's not really AI. Like this is just a pin that does other stuff and AI is built into it. Um, I like the idea of if the pin knows, I think that'd be cool if I could be like, go to my pin and be like, You know, order me an Uber to go home. And it's just like, Oh, cool. I ordered it to you. Right. Again, I don't even know if that's needed though. That's the other thing of these actions. The question is, are the actions needed? Like when I have my phone now, if I don't have my phone, the question is. Are you going to have these devices without your phone? The problem is that these devices are living in a phone exists world, right? So in a world that I'm there, am I wearing this pin around all the time? What am I doing with it? Am I speaking with it? Is it just in the house? I don't really know. I like it though. I like, I like the concept of me. Like the glasses, like the meta glasses or the Google glasses, like something non invasive that kind of can do some cool stuff, laser project something on it, right. That's there. And I think the vision pro will get there eventually. Um, you know, it'll, it'll trim down over time and they all will. Yeah, it, it seems cool. But then, yeah, I think the other problem is, of course, it's like talking to you, but I liked, I like, I like the idea that it's not. A thing that goes in your pocket. I like that. Yeah. I, I, I think we're all [00:35:58] Frank: tired of putting things in our pockets. This is so like right now it, I think it's still pretty boxy. I, I don't, I'm, I should have brought up a photo of it, but it's still a little bit boxy, I think. And so it's still a bit of a mass that you would put on yourself. It's really that thing of, okay, so every day for the rest of my life, am I going to take this thing off at night off of whatever outfit I'm wearing and then put it onto the next outfit that I'm wearing the next day? Am I willing to commit to that lifestyle? And I think it again goes back to the dingus problem. Do we want these passive devices where there's something kind of always around you? Like, uh, you get in your car and now your AI is following you in your car or your out working in the garage, a not normal part of the house. Is there something there to talk or to monitor you and to listen for you? Or do you have to carry the device around with you? These are like the two things, like, so either the environment is equipped with all this kind of stuff, which takes money and time and hard to upgrade and all that kind of stuff. Or we have these personal devices that follow around with us. If it's going to be a pin, I would want like five of them. So I can just have that pin on multiple jackets or something like that. I put the jacket on or it's, or it's built into the jacket. You know, this is my AI jacket and I'm almost more comfortable with that. You know, talking about the vision. I keep thinking like the vision should have just been a helmet. It should have just been like a motorcycle helmet because those are comfortable. They're because rest on your entire [00:37:40] James: head. [00:37:41] Frank: There's no like wiggling around. I mean, okay. Some motorcycle helmets really compress you in, but a lot of them are comfortable and it should have just been that I should just pop a helmet on my head because it's easier. And you know, I won't be opposed to some of these things are AI. I, I want comfort. I want ease and comfort. For most of that. I just want the passive environment experience. And so if I'm gonna physically put something on myself, it better, gosh, darnt be easy and the battery better last a few days at least. [00:38:10] James: Yeah, I agree. Alright, let's talk about, uh, second device, the rabbit R one. This is from, um, what is that Rabbit? What's that one? , the head name That's from Rabbit, uh, teenage engineering. Oh, that's who like designed the hardware. Yeah. I love it. Now this device. Not a wearable, but it's a little, uh, thing you hold in your hand. Pocket, [00:38:38] Frank: huh? [00:38:39] James: Where does, where does it go? Where do you put this thing? Pocket? I think back pocket, front pocket, A pocket. It goes. Well, you already have, you have keys in one pocket and you have your phone in other pocket. You have a wallet in the back right pocket. So this is back left pocket. [00:38:55] Frank: Do I have a phone anymore? [00:38:57] James: Yeah. 'cause I guess these [00:38:58] Frank: don't have cellular, so I guess I still do have a phone. [00:39:01] James: They have cellular, but they don't have the ability to make a phone. Well, actually that's a lie. The human don't need the phone. Humane pin. the humane pin can make phone calls. Yeah. [00:39:12] Frank: Don't need the phone part. Honestly, don't need it. So that's fine. If it's just the cellular network that I connect to a hundred percent fine with that. But [00:39:20] James: I kind of, [00:39:22] Frank: okay. [00:39:23] James: Okay. So, so yes. Okay. So here's, okay. So, so I want to double click on that really quick is yes, but I'm going to not yes and you, but I'm going to say yes, but, cause I agree, I don't actually want to have my phone with me, but what if that device that is replacing your phone. Can only do Spotify music, maybe do Uber, and you can ask it chat GPT things. Is that enough for you not to carry around your phone? [00:39:52] Frank: I thought, I thought you were going to say it could do iCircuit also. No, no. Cause the, the phone is so general purpose, I guess that's the problem. They're competing with a very general purpose device here. Kind of like how, I guess the iPhone computers, huh? Because it was mobile and computers were never mobile. A new device like this has to compete with the phone, which is really tough. [00:40:15] James: That's the hard part, right? I mean, the iPhone was a, you know, competing with like the iPod , you know, I guess, and then it destroyed, I guess so, yeah. 'cause we were, [00:40:25] Frank: we were all carrying Walkmans around, I guess. Is that what we were doing? Yeah. Before the iPhone? Yes. Um, that's correct. And so, yeah, I guess it did replace that device. Uh, but most of us have whittled down to a single device. So there's not much to replace, other than the device itself. [00:40:43] James: That's true. Yeah, exactly. Right. And so this little Rabbit R1, super cute, has a little scrolly wheel on it. It's got a little screen on it. It's got a little rabbit. Super cute. We, we [00:40:54] Frank: opened by saying that we like technology and I can't see anyone who, who doesn't look at this and say, that's just nicely designed. I, you have to say it's cute. They went with a wonderful aesthetic, a very strong color and a black and white screen. I love simple simplicity like that black and white screen, uh, looks fantastic. And it has the greatest scroll wheel ever. Just a giant wheel that scrolls. I have absolutely no idea why it has a camera on it. No, cause it's the vision stuff. We already talked about it, but I don't need the camera, but otherwise it's a beautiful little device. [00:41:32] James: It should, these devices allow you to take photos with them and then upload it to their cloud. That's the mistake is they're trying to be general purpose. Don't be general purpose. Like that, that's your goal is to, I already have a general purpose device. That's way better at everything that you're about to do. Be really flippin awesome at something that I want. [00:41:56] Frank: They really have that. They have like a camera app [00:41:59] James: where I take pictures and [00:42:00] Frank: upload them. Take a photo. Yeah. Waste of time. Waste of time. Yeah. They should have. [00:42:05] James: On, on, on the, on the, on the humane pin, you can take a photo and then it will project the photo onto your hand in green. At super low resolution for you to view it for some reason. [00:42:18] Frank: I think the problem is a lot of engineers are my age and we grew up with Polaroids and fun, wacky technology. And I think, look, I applaud fun, wacky technology. I, that was the thrust of my tweet today, by the way, get in full circle here. I love technology. We opened with, we love technology. Polaroid, great technology. Taking a selfie with a bunch of my friends and then projecting that selfie onto the bar table. I'm going to go with a pass on that one. But give me, give me a week and I'm going to brainstorm and I'm going to come up with things that I could actually do worthwhile with the projection. [00:42:59] James: Yeah. It's not that, no. Um, you know, I, I agree. Like the R1 device. Super cute. I feel like they've made some really bad software decisions with it. Like they deactivate the touch screen for basically everything besides typing, which doesn't make any sense. So you have to use the scroll wheel, but apparently the scroll, which I like, I think is awesome, apparently it's hard to use cause there's not enough input. You know, on it, I think it's trying to So you can distinguish itself, between, from being a phone, but it's okay. You can have a touch screen. Nintendo Switch has a touch screen. I mean, I use it all the time. Just [00:43:35] Frank: use it. Why would they disable that? Yeah. You're letting your designers get a little too far ahead there. Um, Touch is the new everything. Little babies know how to use touch screens these days. So you have to do touch. Is that why? Runs ants! Okay. So yeah. Tell me James, I, I, I, I just want to kind of cut to the chase here because I have not watched one negative review on this device. What's all the negativity from, why, why do people hate this little 200 orange thing? [00:44:07] James: Everyone hates both the devices, um, and I think it's, and the, the, I've watched many review videos. The gist of it is this. Is that. They just don't really seem to work very well. I mean, that in a nutshell is it there. They don't seem to work really well. They don't seem to be solving a problem that needs to be solved and they add more frustration than not, um, for the most part. Uh, And I would say this, they also aren't finished. Now we started with Kickstarter and some of the early things. These probably should have been Kickstarters because here's the brilliant thing with Kickstarter is when you kick are a Kickstart backer, you are expecting to be the people that are beta testing the product. It is a, for the most part, right? Right. Rev zero. Right. And you know that what you get may not be the final design at the end of the day or what you get, it's going to be early. There's going to be tweaks with it. My, you know, my, when, when you buy something like a coffee machine or like I bought like the, the kettle, uh, from fellow or whatever is their first product. Like it's going to change. It's going to take longer. It's going to iterate. They might iterate with you. Right. They might send you a, be really cool. It's like, I send you multiple products and get multiple feedback right on it. Like hard, but yeah. Yeah, I beta multiple software, hardware and software for, for work. I have a device right in front of me that is, when I got it was unreleased, right? From the world. We got the Apple devices that were unreleased from the world. We got to try them out. We get to give feedback on a report, all this stuff with it. You get that data. You get the, the, to see how people are using it, that feedback. And now these products, rather they're just releasing the products in the wild. And I think the problem is the factor that both of these devices came out kind of in this unfinished state where they had a lot that they were promising, but didn't deliver. One, stop promising things, just say what you're going to do. But also what we're seeing is this is. This is, we're actually seeing this with more hardware and software and games and everything like this is people are putting them out. They're not finished and they're expecting us to pay for an unfinished product, which again, I actually don't have a problem with by the way, but just be super clear what's in there and then pick and choose your delivery mechanism accordingly, because I don't mind. Opting in, we've seen super successful Kickstarter projects that have hundreds of thousands of users, right. And deliver on a different schedule. You can do that. That's totally fine, but you gotta set expectations accordingly. So that's the gist of, uh, What the reviews are. The one, now the problem is once MKBHD sort of did this review and then a few other ones did the review, now you just see, uh, everyone's stacking. Right. Yeah. On, on it, right. So now it's more of a meme than anything in my opinion, if we're talking about at this point. Poor company. [00:46:53] Frank: Okay. Yeah, [00:46:55] James: but it is what it is, you know, I'm sure that the difference is like the humane pens, like 700 plus you have paid 24 a month for the subscription. and Uh, it's a cellular connection compared to the R1. That's a cute little device for 200. Expectations, which is why they did it that way. Right. [00:47:11] Frank: Yeah, and I think 200 hours is a good price point because it's a good chunk of change. You don't want to be buying one of these devices every month, but you can afford to do it every so often. And so it's, I'm still surprised because I think it's a good price point. You can go buy a Raspberry Pi for off Amazon right now for about a hundred bucks. So they're only asking, call it another hundred, that's without a screen, I would say even. Um, [00:47:39] James: yeah. [00:47:40] Frank: So I think you're still getting a pretty good deal from a computing device. It is too bad that they're nerfing like the screen and things like that. Let's dig into unfinished product. Cause I know this is, it's a plague of the modern era, ever since the like, release in MVP, the Minimum Viable Product, we, uh, there's a little too much beta going around. I think it's especially a problem with, uh, video games right now, like, they're bad. They're just, uh, whatever. Everyone makes fun of cyberpunk, right? I'll just jump on that bandwagon. It took like a year to clean it up. Uh, even And Sky. That took like a year or whatever to get polished. So is it just the modern times? Do we release things like that? Um, I don't know. It's tough. Okay, so they have one feature. There's a beautiful gray button on the RabbitR1. You're supposed to press it, and you're supposed to talk, and I guess it's supposed to respond. Does that work in these reviews? [00:48:41] James: Does that much happen? I think that that mostly works. Yeah. That, that functionality mostly works. Yeah. More so than more, more so than the, uh, humane pen, the humane pen. Both of them seem slow, but the humane seems plan seemed really, really slow. And I'm not sure why, what the rabbit R one is. Maybe they're doing more of a wifi or something. Cause that's like how it's out of the box. It seemed like from yes, that, that worked well. I actually think that these cameras that it put, they put on this. Wow. Is a good. feature like we've talked about for walking around taking photos and getting insight into what what's in there. I think that's actually a detriment because then people just take photos with plans. They're like, well, that's not that you're wrong. Right? I go, that's not correct. That's not correct. No, that's that's not a whatever, not our present. So they're able to easily debunk a five this stuff compared to this. However, I'll say this. If I kind of, right. Yeah. The, the, the. The thing that I see most of these reviewers doing is that they're asking questions about stuff. You know, what I'd be interested in is ask it stuff that my dingus can't answer. I can ask my Google assistant to like, I can ask it questions. It'll just Google it for me and give me the answers. Tell me to, you know, Oh, I need you to, I don't know, write a, write a poem about this thing and then save this as a thing. I don't, I don't know what I need it to do, but like. What I would want it to do is maybe like, Hey, can I'm, I'm going to do a voice. Here's what I would want it to do. I'm going to do a voice memo. I want you to transcribe that. And then I want you to like, rewrite that with AI to make it like way better. And then write a summary at the top. Right. I don't know. Something like that. Right. I don't know what I want it to do, but maybe something like that. That sounds cool. [00:50:26] Frank: Gosh, I would love that. Like here, I'm going to, I'm just going to talk for about 10 minutes. Can you turn this into a blog entry? [00:50:35] James: Yeah, totally. But, but these devices, the AI devices are just basically, I feel like in the demos of the reviews, people are just Google searching with them and that's not really what, that's not the power, it's not powerful. [00:50:49] Frank: Yeah, I, I'm always a little depressed at what people ask AIs. I mean, okay, so the AI presumably is not running on device because these are not powerful enough. It seems like they're, they're barely powerful enough to light up a screen and talk to the internet. And so we are back down to how fast can you talk to the internet? How fast is the server? How fast can you get the response? [00:51:11] James: I think they're both talking to open AI, but they could be talking to multiple services. Yeah. [00:51:18] Frank: Oh, I'm sure they are because, yeah, that seems like all the vision tricks they're doing. Those aren't. Not all of them are off the shelf. Um, so they're just dumb terminals. And so it makes sense with their 200 calling these AI devices, they're not doing any AI, they're just presenting a better, simpler interface to these AI devices. Solving tasks that they think that we're going to do. And the problem is we are so headed for another AI winter. There's going to be such a crash because what AI is capable of right now is here and people are hyping it up to here right now. And it's just not good. I like this device because I think I understand the limits of AI right now. And I would use it within its limits. But the moment you take a picture of. A random piece of artwork in a random museum. It's just going to fail. It's, it's too hard of a problem. We're not there yet. The databases aren't that big. They're not trained to recognize that stuff. And so I think that there's a hype cycle here. I see this thing for what it is. It's a cute little orange terminal to a bunch of web services. And their web service is controlled by this rabbit company. I think that's a little weird. As a developer, I much prefer a programmable device. That's why I'm always going to have the phone. I want to write my own apps. I want to talk to my own services, but for a simple little terminal that can talk to an AI by the click of a button, I absolutely see no problem with this thing. And I wouldn't mind having one in my life. I think that they need to. Improve their feature set. That's easy because it's just a dumb terminal. [00:52:57] James: Like [00:52:59] Frank: that's, that's all services. Like, what is this thing? It's a simplified phone. That's all it is. And we're just going to refine it to these few tasks. And I don't know why people are getting so hyped up about that. It's a 200 internet kiosk. [00:53:17] James: I know. Listen, I think on the Rabbit R1, I totally agree. I think that for 200 bucks, you know, 200 is still 200. But at the same time for a device that has a screen as a stuff, I've paid way more money than for something that does probably way less at the end of the day. Right? Like I still think it is nicely designed. It has a cool screen as cool enterprise. I can see it doing more down the road, hopefully, but I agree with you. I, I think I'm not mad at this device as long as it does, you know, work or enrich something. I think it goes back to the original question is like, is this doing something that is going to enrich my life or make it easier than grabbing for my phone? Or would it allow me to like, leave my phone in the bedroom charging? And then I can just have this little device to do what I need it to do, you know, in general. [00:54:04] Frank: Cause I'll tell you, like, when, Uh, WhisperNet came out. It was, I think it was Google. It was an open source neural network that broke every, uh, broke all the state of the art accuracy challenges for speech to text, a solved problem. Microsoft solved this one 10 years ago. It's all worked pretty good. We've all had Siri transcription for a while, but here on the street came WhisperNet, it's a new modern network and it's available, we can all just put it Every journalist out there is like, Oh my God, I need this. I need every meeting I'm ever in, every interview I ever do. I just want to pipe it through WhisperNet because it was that much better at dictation than all the others. And this would be a perfect little device for that. Like the little handheld, uh, Tape recorder. They're not tapes anymore. You know, audio recorder. Those aren't going away. Like our, yes, our phones can do it and all that kind of stuff. But sometimes it's good to have a dedicated device. And I could very well see one of these little AI recorders being, this is what I do. I sit on the table. I have a really good microphone and I have maybe a little screen showing the transcription. This is all I do. Uh, but I do it well and I'm cute and I'm orange. [00:55:25] James: And, and then it does AI things for you automatically on top of that, right? Yeah, exactly. Like it really does like summary. It does TOC, like really figures it all out for you. I would, I would love that. I think that would be super useful. I, I often dictate to Word. Um, when I'm doing like reviews and connects and things like this, and I'm writing big documents and stuff like this, then, then now I can go in with copilot and clean it up and things like that. But yeah, it'd be fun to have little things with me around and imagine like at dictating, imagine like dictating the world around you, like you're on a travel trip and you're like trying to like, remember the moment I'm just going to speak about what I just had. And then imagine if it's like, Oh, I'm AI. So I'm also going to like. Underneath, like enrich this note with like all things about Vienna and all this other stuff to give you more context. Like there's cool things that you can do. Like, again, I think that's the power is the context, right? If I give you the context of 10, 000 words and you take that and you do something with it and enrich, you know, enrich it and make it easier for me to digest things and do a little summaries up top, all this other stuff, like that's cool. Right. I think that's. You need at the end of the day else, when people say ai, I mean, we have AI with this all the time. I, I talk it to a dingus, it does something that's ai it, it's, it's, it's doing something. I don't care what it's doing. Yeah. It's ai, it, it's, it's, it's all ai. [00:56:43] Frank: It's all ai. I used to joke, like, um, at least in the seventies, AI was every interesting computer program problem. That that's what AI was like, what, what is ai? Because we, we confuse it definitely today with machine learning. Um, not all AIs require machine learning, especially if you look back in the seventies when we were solving, uh, pathfinding problems, graph searches, things like that. [00:57:11] James: Uh, [00:57:12] Frank: the good old like traveling salesman problem, things like that. Those have always been considered AI problems. They are advanced problems. They are not pure calculation. They are, you know, computers are good at calculating. Honestly, that's logic and calculate. Those are the two things they're good at. These are problems that cannot be easily solved with logic and calculation. They need more interesting solutions and therefore AI was invented. And then machine learning came along and now everyone thinks that machine learning is the only way to do AI. But no, uh, we've been using AI for a very long time in computers and they've been assisting us. Now it's just, now they're speaking English. So it's a little more awkward, um, but that we've been using this stuff forever. Yeah. We've had Drake and dictation since the nineties. Yeah. Had a cute little devices. Yeah. They do that one. [00:58:08] James: So I hope that this, here's my takeaway. I hope that these two AI devices, AI devices, aren't the end of AI devices. And I don't think they will be, I think they'll only be. The beginning, I think that Apple, Microsoft, Google, they'll do cool stuff with the phones and the computers and all the stuff, the OSs. But I still think there's a need for some new interface of some source to do the unique things of AI that, that we want to bring around with us that maybe won't be in this. I think having device, the, the phone will even infused with AI will be a generic device that does all the things. But if I can get something that is really good at doing all the things The thing, right? Um, I think that would be great. And then that's why, you know, some people game on their PCs. Some people like consoles, right? One's a generic thing. You can do all this stuff. You get to tweak it. You get to do all the stuff. Some people just want the thing that does the thing. Don't have to worry about it. Right. And that's sort of where I hope these devices go and I hope we see different form factors. I think we both agree. Neither of these device form factors. We love or hate. We're, we're seeing to cool. Seems good. I hope there's more different things. I hope there's crazy things that I can't even think of that we, that, that happen in these form factors of different modality of how we're communicating. Like you said, like the other thing of interest is like, you know, what if there was just a humane, the humane pen is now not something you'd know. You no longer like, um, you wear, it's just actually you. You magnet all into all every room in your house. You have to buy like 20th. Like, imagine you have this little projector screen in every room. yeah. I was like, that'd be, that'd be kind of cool. I'm just saying like a little, a little Jarvis in every room. I'm [00:59:51] Frank: here for projectors. Yeah. I 100 percent think they should be projecting, like, uh, that, I, I cannot stand listening to the dingus tell me mathematics stuff or numbers. I, I, listening to numbers and mathematics does not work in my brain. I have to see it. Show me that thing. And if every time I asked it a question, it projected the answer out onto whatever surface that happened to hit. [01:00:18] James: I'm fine with [01:00:18] Frank: that. [01:00:19] James: No, it's like we've always seen those smart kitchens, right? With the projectors coming down, trying to help you chop up food, do a bunch of stuff. There's a bunch of ideas. I love it. But I think making that tech that can be integrated easily throughout any home that's not custom built for it would be kind of cool. Um, anyways. Frank, that's an hour on these AI dinguses. I hope we get more. I hope there's more Kickstarters. That's what I think. Again, I think we'd be, people would be looking at these devices completely different if they were just Kickstarters going through it. [01:00:48] Frank: And also don't forget you attach some motors and some wheels to these things and you have, you know, Basically, proper robots. So, let's just cut to the chase and put some wheels on these gosh darn things, okay? I'm willing to spend 300 if you put some little wheels on this puppy, and I have a little robot that follows me around and tells me interesting things about how to Make chocolate chip cookies. I don't know what I asked last. [01:01:18] James: I have a little, little, little, little robot, you know, here's what would be cool is, you know, Sony used to make these eye bows, which are like little robot dogs, they're very, very popular. I'm just saying, infuse this AI technology into a robot dog. [01:01:33] Frank: Yeah. We don't need something that goes in our pocket. We need something that follows with legs or motors or [01:01:38] James: tracks. Robot dog goes in your car with you. Bam. Yeah. Yeah. RobotDog can go to any hotel, no pet fee required. When it's behaving, sometimes they're stubborn [01:01:50] Frank: robot dogs. [01:01:51] James: Robot dogs. Uh, there we go. Listen, companies out there of the world, if you're not infusing AI and robot dogs, I don't know what you're doing. Come on, step up your game. All right, that's gonna do it for this week's Merge Conflict. You can check us out everywhere on the internet. YouTube. If you want to watch us talk about these AI devices, we don't own any of the devices, but you can't see them, but you can see us. YouTube. com forward slash at Merge Conflict FM. Just go and search it on the YouTubes. Patreon. You want to become a patron, support the show. Boom. Patreon. com forward slash Merge Conflict FM. Or just share this episode or the podcast with friends, family, loved ones. And if you're on Apple podcasts or Spotify, hit that subscribe button or whatever podcast app you're on. If you've got a review option, leave a five star review. If you like the podcast, that's good for this week's Emerge conflicts. So until next time, I'm James Montemagno. And I'm Frank Krueger. Thanks for watching and listening. Peace.