AI Judges Surprising Case + Data Centers in Outer... | MCAE

This is your captain speaking. This is time. I'm going to ask you to pass the air seat belts. 25G South. I'm waiting. Teelo, you're going to take off. Oh. Here we are. It's recording day. Here with me is Jim Roundsville and Gus Stein, the sub billionaire podcast. I'm Bill Alderson. Here we go right into the topics of the day. Jim, take it away. Well, I crossed another threshold today. We're AI judges entering courtrooms, robots moving into global out event security, like down with the World Cup, some serious infrastructure conversations about taking computing off-plant. So that's what's in it today. I know Gus, you're going to cover the third section bill you're going to do. The second one. So we'll get into those in just a moment. So welcome back to Morpheus Cyber. I'm Jim and today we've got four PAC segments. So gentlemen, let me start with this one. And it may sound like science fiction. It may have been written 10 years ago, but it's not. So what if I told you that an AI judge just processed 10,000 traffic violations in the time it took for you guys to go get your coffee, which we all just broke for and got our coffee earlier? Anyhow, it might have been more fair than even human judges. That's the great thing about this thing. So that's the upside. So welcome to the surprising real world of AI courtrooms. So I'm not talking about advisory tools. I'm not talking about some legal assistant that's better or faster than something else that we had in the past. It's an actual AI judicial decision system. Pretty cool. And already operating in most parts of the world, or a lot of parts of the world, not most, but a lot of parts of the world. So now, before anybody starts imagining minority report or the new movie Mercy, where judges are basically sentencing people to death, that's not what this is about. So we can slow it down a bit. This is about replacing. It isn't about replacing our Supreme Court or any of our major judges that make those kinds of decisions. Instead, it's a bit different. In fact, far more disruptive is essentially because it's something that will now take care of backlog court systems, human inconsistency and document bias in our justice system. We're talking about things that are more like civil cases and that kind of deal. So not government efficiency, it isn't the government efficiency hawks that are really saying, we need to do this thing. It's not all about that. And, Rick, it's actually the civil rights groups are saying this needs to happen and understandably so. So what do you guys think about this so far? From what I've heard, studies are showing that human judges give 23% harsher sentences before lunch. A fair two-division fatigue. So, I mean, we already know statistically that human judges have bias and they also have the human element. They get tired. So, yeah, what about these AI systems that are going to be unbiased? Cool. Yes, you got a call to that? They're not really unbiased, they'll build. Because it all depends on their training. There's bias, early bias in AI. However, I still think it's a good idea. As long as it's, and we talk about this a lot where they are. As long as the AI is taking care of kind of the document review and the different process and fill that out. And as long as anything important, I think a lot of these countries that are already doing it, they have a certain all-er number. If it's over $700 or whatever, it's going to have human eyes on it. If it's a traffic ticket, is that really a problem? So I think that that could be game-changing in a positive way. Because everything's backlogged right now. I got a speeding at last year. You know, I live in California. I got a speeding ticket. And I tried to go pay it. They said, we can't pay it yet because it's not in the system yet. So I put an outlook calendar. I checked it every single month because I didn't want to get light fees and everything. And finally, after about eight months, I called them up. And they said, listen, if the officer does, if the process doesn't put a claim in by two months from now, you're off the hook. So I started rooting for it and I never paid it. The state calculated it was capable of collecting my $400 fine. It's so funny. Help with that. So this story is really more about the fact that our court systems worldwide are drowning in backlog cases, right? There's over 600,000 civil cases backloved in the US. The federal courts alone, 600,000 civil cases. Some jurisdictions have been facing over two to three years of delays. Now, that's just nuts. You get accused of something or you have a dispute and you want it settled quickly. And two to three years later, you're still waiting on something to happen. And then when COVID hit us, all kinds of bottlenecks happens that still are not unresolved. And then we're not just talking from civil cases, legal cases, but look at all the tax system. That's backed up too. That's a whole nother subject we need to cover. But right now we're to stay with the judges right now. So we all understand that state courts are underfunded. That is a problem. And then we are in a litigious culture, right? That is just so sad. Everybody sees everybody at the drop of a hat. So that's backing up things. And then small claims disputes are they're waiting months just to get a resolution on something. It's really simple like a property line dispute or something. It's simple, something that could have been resolved quickly. So here's what makes this really uncomfortable. It's like Bill was saying judges have biases in them. So if one of the things he mentioned is about, you get harsher sentences if the judge gives you a sentence before he's eating lunch. That's a that is a known statistic that's actually been proven. And there's you get a lighter, more lenient kind of judgment after maybe a positive support outcome like your team won the Super Bowl. So the judge is on a high the next day. So he's feeling good. I'm feeling merciful. I want to do good things for people today. And there's about 65% variance in sentences for identical crimes. And that's really sad. And this is why we see, you know, civil rights people, people, organizations very much want to see this happen. And black descendants received 19% longer sentences for the same offenses. So when you add all this up, yeah, I think we're ready for a change. This is not a conspiracy. This is data. This is real data that they've proven is happening. So our goal should always be fair and equitable treatment under the law. But unfortunately, that isn't happening. And so hopefully an AI system will be able to do that because AI systems already show less than 5% variance compared to a human at 65% variance. So you already know you're going to probably get a fair outcome from an AI system and you will for a human judge. So the question is, is the system, is the justice system the next domino to fall? You know, just like the finance and credit scoring industry, right? That needed to be under some scrutiny. This needs to be under scrutiny as well. You're good. Yeah, if Estonia, which I think August said had 1.6 million, three, reference, small country, those judges have a 12% appeal rate for the AI judges and a 31% appeal rate for human judges. So in the perception of the people, it's already showing that people would rather be judged by an AI than by a human judge. Yes, you're going to say something. Well, I'll just as I'm going to say, what Bill said, I mean, there's a lot less appeals, which that's the proofs in the pudding, right? I mean, if there's less appeals, it must be doing a better job. Well, in other countries like China, they've been using AI since 2017. So they've been out there away out of us. But their system is definitely a lot less nonsense than art system. You don't see the backlog in China. It's pretty clear it out fast because they have heard fast rules. Yeah, for sure. They have roughly, they are now handling, AI is handling 4.5 million cases as of 2022. 4.5 million just in 2022 alone. So boy, there's a big change happening there in China that we need to see if they have to take place in the United States. Now, I'm not saying everything's rosy. We need to make sure that we're making sure our systems are fair and that they are not judging cases like murder cases and things like that. Those are a whole different element. So you're now, I don't know. Yeah, go ahead. What do you think? Yeah, the whole is a fear. It's always a scary thing. The scary thing, you know, you put governments in charge of these things and we think, oh, it's not going to be a problem because they're just going to handle the parking tickets and everything else. Next thing you turn around as 30 years later and here's our front of a computer. Yeah, mine already. Or a report and mercy comes to life. Yeah, that's the cultural fear, right? People are thinking like, well, I've seen those movies I don't want to have. I can minority report. There was a pre-crime system that was already predicting when a crime would occur, before it even occurred, and then went and grabbed the person that would have done it and then put them in court because you already were going to do this. Why didn't do it? But it doesn't matter who you were going to do it. Anyhow, so mercy is the same thing. They kind of took it to the extreme. It's like you get pulled in. This is a movie that's playing in the theaters right now so you can go see it. So the system pulls you in and says, like, you got 90 minutes to provide a defense. If the end of 90 minutes, if you've not provided it, having good defense or dead, I'm going to kill you at the end of 90 minutes. It's very efficient. But totally wrong. So total loss of human mercy, total loss of discretion, total loss of empathy. Those are the things that we, that is calling human fear, you know, public fear about systems like this coming into place. But justice isn't only math, right? It's got to be, it's got to have some predictability, but it has to have mercy discretion in empathy as well. We're going to see a lot more of this. Why? Because they have billions of dollars of venture funding going toward this right now. So we're going to see it in the future. And now you know what to look for. Well, as long as we don't outsource moral judgment, now I think we'll be ahead of the game. Proficiency, yes. Capital punishment. Absolutely not. Can I be? We blur that line. Then we're just stepping into the world of fiction, become this fiction like mercy and minority report actually become a policy. So I think we can prevent that. As long as we know what the downsides are in advance. Okay. Onto our next segment. Data centers and outer space from energy limits on earth to moon base computing. O M G. First of all, out there in space, they can build components that don't need air conditioning because it's already cold out there. And then the other part of the equation that's pretty cool for that is there's a massive amount of power required to run these things. And guess what? Out in space. If you're honest, if you're a satellite or you can put a space station out there that does all your data processing and a company like SpaceX is perfect for this because they can relay data around the world to all of their satellites, thousands of them, and they can communicate from anywhere to anywhere. So you can set up your satellite so that it's always in the sun. And it's always in the sun what happens then? Well, you've got 24 by 7. What? Power from solar. And that negates the need for what? Batteries. You don't have to have those big old batteries that are dangerous to shoot off into space. And so there's a lot of advantages to putting these data centers in space. And with Elon having this massive new rocket that can take up like, I don't know, 600 or more satellites in one shot, they can put all sorts of stuff out there in space, have a data center in space. Now no doubt somebody's going to complain because maybe space is going to get hot. What do you think, Gus? I don't know. It's crazy to me, Bill. I don't know. There's so many issues here. We see this over and over. There's something seems impossible and then it isn't. But what about the latency? I mean, okay, so you got the satellite collecting 24 to 7. And I get that. And then you have some sort of space station on the moon, presumably, right? And that's nice because you mentioned there's no atmosphere. So there's nothing holding the heat. So that solves that problem, which is about 30% of the cost of a data center is cooling. So that makes our latency from the moon to the earth is like 1.3 seconds. And that's great for doing stuff like, you know, large language model training or things like that, but not for communication. Yeah, they're going to have to use Leo, low earth orbit satellites. And that's exactly what, you know, Elon has like thousands of them. And Elon's not the only one who has those Leo satellites. There's a lot of people who have those and are building on that and increasing as they go. So there is the possibility that our next data centers are going to be in space because there's intrinsic value of cooling and intrinsic value of 24, 7 solar and a whole bunch of a whole host of other things where you don't have to have batteries. Batteries can explode. They can get too hot. And hey, space is a good place. Yeah, makes makes a lot of sense. I know nuclear are coming on, but online eventually will make a dent in that too. But right now, a typical data center, AI data center is 30 to 40 megawatts per facility. The bat right now, data centers are taking up 1% of the electricity grid right now, what 1%. They anticipate by 2030, you know, that's only just a few years off, four years off, they'll be at 4%, 3 to 4% of the national grid, electricity grid. So some things got to happen. So comedy show nuclear, we talked about fusion not long ago. And so fusion reactors coming on board, I think Microsoft had in a contract with a fusion company would bring on, come on board by 2027. And that's only, oh no, 20, 2033, I think it was 2033, something like that. But it was so it's not many years away. So if fusion comes on board, that's going to change the whole discussion. If more nuclear plants come on board, that changes the discussion as well. So you know, I understand that there's unlimited solar energy, natural cooling, and the other thing that may be driving the thing going into space is that you don't have to, you don't have to contend with zoning, you don't have to contend with environmental permitting. No community pushback because you put a big huge data center in my community. There's a lot of those exactly. We're back, we got a problem. Yeah, for sure. But like if Musk is the only one or that could really do it, I mean, think about it. Yeah, he's already got all his satellites out there and he's already now talking about putting data centers in space. Who can really keep up with him? Really? Who can really keep up with him? So it becomes a monopoly now. So he's got that ability to just throw up a AI data center in space and he's got enough money to do it. He's going down. He's going down. He's going down. He's going down. What's that? The cost is going down to dramatically, like 90% the last couple of years of, you know, of putting this stuff up there and that makes sense. But I wanted to go back to what Bill said. So, so you got these low level satellites and I get that and the latency super fast. What about that server that might in the freaking moon getting to the satellite? I mean, if the, if the data centers on the moon, which I think is what we're talking about, there's a latency issue. And I'll tell you something. Do you know that the temperature fluctuates on the moon from 260 above to 280 below? So, there's no that. Yeah, but with the lack of, of, of atmosphere, it's less of an issue because the atmosphere holds the heat in. So it still dissipates, but it gets hot up on the moon as well. So there's a lot of, there's a lot of moving parts here. True. I think these are two different issues that we were talking about musk because musk just amounts that he's abandoning the Mars colonization effort for a while. He's going to concentrate just on putting the base on the moon. And so he's not trying to put data centers there to shoot back the results from AI back to the earth. He's just trying to colonize that, that, that effort first before even thinking about going to Mars. So I don't, I don't know that those two issues are the same. I think the data centers in space are effective. They probably will go into place. Nobody can stop him really. So it's going to happen. It bussards them. But we should shift over to the whole concept, the, the comment that he made about abandoning the Mars effort and going to the moon. So go ahead Bill. So Gus, there's a lot of processes that we can put on the moon and let me explain some of them. When you are training a new AI model, that's a localized low latency need, right? You're training locally. So you can do that on the moon. Why? Because there's no interaction with somebody else at that point, right? No inference. It's just now let me talk to you about something very practical for a moment. I have a GPU and one of my computers and then I have a bunch of servers who have no GPUs. And I have designed a system where my GPU can, and the machine on my GPU can provide a service that my VMs who have no GPUs can go access the service on the machine with the GPU. And the amount of time that I save on certain processes are like from, let's say, three hours to do something down to eight minutes. So three hours without a GPU eight minutes with a GPU. You can still send that data to the moon. Do the process on the big GPU capability or that sort of thing. Get the job done and the 1.3 seconds is nothing in comparison. So they're just going to put the workloads, the computer workloads where they don't get affected. We all know by being network analysts and working with analyzers all of our life that you never want to take your database and put it a long distance away from where it's going to be queried from because that's where the problem is that latency going back and forth there. So I think the same thing is going to be for jobs that we do on the moon. Hey, training for AI models takes months. Oh, we just remembered up there and yeah, that it. Bang away and yeah, sounds like we're watching the birth of off planet industrialization. So any process that more effective to be off planet will be sent there. And all of a sudden whoever got there first and built a livable place that you could actually run these systems has the advantage. Like you said, the one second, one, one second delay is overcome by the fact that the processing power results are better because they're in a cool environment, low cost and got to this, maybe got to this, the result faster. Who knows? So that's pretty interesting. That's a different, different way of looking ahead and thought about that. So what do you think? What's up next, Gus? Oh, our next segment is something that's pretty interesting. It has to do with AI and it has to do with the last couple of years. We've seen a lot of lawsuits that centered around do AI, scrape different resources, music, different articles and books for their, you know, for their information and there's been some lawsuits. But now things have shifted a little bit and the difference is that AI now has the ability to sound like people to look like people and potentially replace people. So there's a Google notebook. I'll notebook LM and it's generating voices that sound almost perfectly like famous people. And there's a host for MPR, host the morning showers names, David Green says one of the voices sounds too much like him and he's suing him. So at the same time, Hollywood is pushing back against the same sort of of AI thing with C-dance 2.0 is a product that you can go text and you can produce high quality movies and leave these shows and commercials and stuff like that. And Hollywood's really pushing against this and it's real similar because things shifted. So this is what I'd like to throw out to the group. You know, if you have a machine that can replicate your voice without your permission, you know, is that innovation or is that theft? It gets fits really unique. If your voice is extremely unique and we know who all those people are, James Earl Jones. Yeah. Yeah. John Wayne. I mean, there's just a bunch of them out there that just you can tell when you hear that voice and you're using it for that very purpose of what that person represented. Yeah. Then I think they've got a legal case. Like for instance, if you're going to use John Wayne's voice for cowboy commercial or something to do with the outdoor and horses and maybe something of being rugged and being a tough guy, you know, yeah, then you've already made the inference by doing that. So using his voice would be, I think unfair in that case, he's not alive to defend himself. But in the case of someone who is alive, there's plenty of people that are still living that have great voices that I would think would be an invasion of their private use of their voice. So I get it. Our own podcast just to let everybody know we generate a test or what I call a mock podcast. And we use 11 labs of my voice, Jim's voice and Gus's voice. And it's kind of like a mock podcast based upon the topics that we are talking about today. And it takes all of our background information, send it over to 11 labs. It pops in my voice when I'm talking, Jim's voice and Gus's voice. And we banter about do humor and that sort of thing. And it's somewhat of a training ground for us. It essentially allows us as podcast host to hear in our ears a an expert capability of doing humor and going back and forth. And then when we listen to those, it's like muscle memory gets developed in us. So that when we're on the air, we can actually mimic some of the expertise. And sometimes I laugh at those things and they sound better than we do. Yeah, those are good uses. So in this case, let's bring it back to kind of the topic here about being legal president. Does this create legal precedent for somebody to use your voice to profit off of it? So that's that's kind of what the question here and like we were saying, they're taking into court. So I guess we're going to find out, aren't we right now? But right now we don't care if anybody uses our voices at this point because we're not famous. So it doesn't really matter, but I could see someone who has made their whole life trying to create their persona. And then someone takes it and tries to use it because it sounds so good that their pro their podcast benefits from it. And they get more views because of it. Sure. I think you've got a legal position. And also they're doing additional things like Matthew McConaughey is protecting. Oh, right, oh, right. Now I don't know if I just violated the law. I'm not selling that. But basically, you know, these guys are copywriting their saints and their voices and their inflections, right? And that's that's what we're looking at here in the future. But we all know that someday in 10, 20, 30 years, yeah, it's not going to be so much in my opinion. It's going to be by the wayside. Why? Because those people will be gone. They're whatever. There'll be new people, but but they're cooked. Here's the here's my question. You know, isn't this just the nature of progress? You know, when when digital photography came about, you know, codec and bankrupt, you know, boo, right? Dark room and pulling is all over the country. Loss their jobs and everything. Why? Because it was cheaper. It was instant. It's a little more, you know, what about Blathbuster? Blathbuster got put out of business by Netflix. Like we talked about another way. And I used to own some pay phones. That was an investment. You guys are old enough to remember my own 10 of these pay phones and then so phones came along and, you know, you don't see pay phones anymore. And I think that, you know, every major technological we create winners and losers. And, you know, I just nobody's protected the blacksmith, nobody protected codec. I'm not sure that we should protect the Hollywood actors. Why are they in somehow a protected breed? Yeah, well, right. I think the other extended part of this argument too is that using the the CDAns tool doesn't just threaten actors or performers. It also threatens camera crews, lighting technicians, editors, production studios, post-production pipelines because it can actually take text and generate full professional looking videos from text prompts. So it takes out a whole bunch of people that may have big concerns that their jobs are out. And we keep hearing this every week. Another technology comes out and AI is probably going to take some jobs out. And I guess we have to be realistic about all of that, right? Because AI can do things a lot better. We know it ourselves. We've watched, we use these tools and we know that when I had cropped AI to do certain things for me, it did a much better job than I would have done in much less time. Could I've gotten there eventually? Maybe, maybe I could have. But the fact that it does it so quickly, I can't compete. It doesn't make sense to stop and not use it. And that's what's going to happen with the public. It's not a matter of, oh, well, whose human rights are being or human, of effect on humankind. Yes, we don't want, we don't want a negative effect. But there's going to be a point where we just can't ignore that we have to use it because it just makes more sense. It's more effective. It produces better content. Everything about it becomes better and better. And we had, we had to talk about this the other day about, hey, but what doesn't the value of human created content become even greater? Yeah, probably just. I mean, if you still could produce after it, that it will survive. And that's why we are here and we are not using our personas or our clones, right? Because we feel it's very important to have the human element, even though we're enabled by AI, we're not AI. And that just shows all of our foibles, you know, and as I age, I forget words. It's like, whoa, where did that word go completely out of my head? Yeah, I would know. I don't know anything. Yeah, I know. Yes. Yeah, there was a, there was a big argument about, hey, AI is training off of other content that was created by somebody professionally, right? So that was a big up for a few weeks and a few months ago, actually. Now, I think that that huge amounts of content was being absorbed by AI that everybody saying who legally owns that content? Who, what about the creator? Who did this? You know, so it's the same argument that they're saying, but a little different in the sense that they could create their own custom content. But then now they're actually turning it into a live motion video, but they are mimicking someone's else's persona. And so that's where this, this is kind of different level argument. The other argument was just pure content alone, you know, the written word that's out there, you know, scanning thousands and thousands of documents. Now we're talking about, okay, now you're generating something that sounds like somebody that we all know. Just going to be some lots. So like there is on regular life, the land or there's, there's, there's copy rights. And I think there's going to be some friction that in general, you know, technological advances has casualties. It will open up new opportunities though. There'll be new licensing contracts in the marketplace that will start to emerge and people will have to license that capability. Like or license that human voice or that human contact that was created by somebody. There's going to be some point where there's going to be a way to say, but identify that was that person you're mimicking. You owe him some compensation. Here's the, here's the licensing agreement done over and done. We've already gone through that with Apple music and other music digital music. You're not selling the album anymore. You're not selling that vinyl that cassette tape or in Gus's world, the eight track tape. You know, you're, and you caught me or check it for track or track even before. So anyway, hey Jim, listen, there's no more talking about Jim. Did you get a new dog because I heard in the, in the, that there's some new guard dogs out there. You want to talk to us about it? So let's bring our conversation down from outer space and get it back down to the ground. So picture this. This year is going to be the world cup in Mexico. So you're going to there. You're celebrating your team's goal. And then there's four legged robot dog walks by past your seat, scanning faces with AI powered cameras. Is that science fiction? No, it's going to happen this year. In fact, it's happening right with, for the first time with autonomous robot dogs. And they will be patrolling the global sporting event. The question isn't whether this tech works. It's whether or not they're ready for robot security guards to become normal. So Guadaloupe Guadaloupe City Council approved roughly 2.5 million pesos. And I want to say that because I want you to convert that. I'm converting it over to US dollars. That's $145,000. So 2.5 million base is $145,000 US dollars. Right. And they're buying these four legged security robots that look like, look like they're made by a unit tree. I kind of looked around and tried to see if anybody could tell me who made these things and, and kind of the consensus is it's probably unitary go to platforms that were manufactured by a unit tree, robotics in China. Because Mexico buys a lot of stuff from China, as we all know. So they've decided to make this purchase from them. And FIFA, I guess, is paying some of the support in the security budget, right? There's a security budget for the event. So they're able to now acquire these things because they've got a budget for security. Now, what can they do? That's what we're going to talk about a little bit. But so this is no longer some sort of tech demo. This is real life global scale deployment. One of the largest sporting events on earth. So we're going to find out what these little dogs, these robotic dogs look like. And the fact that they went out and bought from a unit tree makes a lot of sense. Because, right, there's another, there's another four legged robot they could have bought from Boston Dynamics. We've seen that in the news for almost two years now. We've been watching these. They're probably cost 145,000 a piece. No, they're 74,000 a piece. That's what I could find. 74,5 per each one. But I don't know how many of these these dogs they ended up buying for the 145,000, but there's probably quite a few of them. So I think they're affordable. They're developer friendly. They're security-acable. They're not military robots. They do not have a gun. They think they're not armed. So that's a good thing about that. And so they have this ability to have a little container or brandy on their... I was wondering what other value could be because they carry around extra ammo or something for the actual security guard. Maybe so. But anyhow, they have a top speed of about 11 miles per hour, which is faster than a human's probably going to want to walk. But one thing I saw on the downside of it was it only had like about a two to four hour battery life. So I got to run this dog for about two to four hours and then you got to go plug them back in. You better have a lot of them. Because it's pure purpose is really to go and document. And changeable batteries. Yeah, maybe so. At the same time they have to... I read 90 minutes. Was it exact? I read 90 minutes was the limits or there's a couple different numbers floating around. But I thought, because a soccer game is about two hours. And then a lot of the security is prior to the game. And I wondered the same thing. Did they go on shifts or can they change the battery out? I didn't see that as a slight limitation. But if you think about it, their ability to communicate and to be able to assist the human security by noticing different trends and stuff like that. Maybe they can signify it a bomb. I don't know. But they can recognize different behaviors that are happening in the audience. And they might be able to prevent some sort of terrorist attack or something like that. Although for me, that when you're talking about a robot, there's several different aspects to a robot. When one robot sees a threat, boom, instantaneously, all the other robots have that same threat knowledge so that they can focus and triangulate on the threat. And I asked for that knowledge for a different perspective of where they are. So that relevant... This is something that's happening over here. Well, gee, that makes me look over... It's going to make them smarter. Yeah. Well, let's talk about operational years possibly. Ways they might use these robot dogs at the event. One of them could be perimeter patrol or parking garages, loading dock areas, restricted access zones. They could do night patrols because they have light art. They know where they are. They map out where they're at. They send their locations. They have a camera. Multiple cameras on them, not just one, but multiple cameras on them. So they see almost 360 in all directions as they're walking. Whatever perimeter they're expected to go through. I have an idea. How about those TSA agents who stand at the exits just to make sure you don't enter the exit? Could we... Is there a dog? Yeah, there's like several of those at every airport. And so think about what that must cost. Times of thousand airports. You got a thousand people, three shifts a day, standing to make certain that nobody enters in the exit. Perfect place for a lethal robot dog. And you're going to be profiled next time you travel. That ball guy wants to put a thought of work. Well, the second... The second... Great, probably good use of these robots would be suspicious package, Rikon. So like if there's somebody left a bag that's just sitting out there at this event, which is always a big problem whenever they see a backpack that got dropped, they don't know what's in it. So this dog could actually approach it, determine whether or not it is suspicious in any way based upon whatever sensors it has on it. And it has the ability to retrieve it and move it. So it can stream the video, it can give time by time for the experts to get in by keeping other people away from it, bringing notification that he found this device or this this bag or backpack or whatever it is that's sitting there. So that's a good use, especially the residual aspect of it. If they decide, you know, there's just they've got to get this thing moved somehow and they don't necessarily want a human to touch it. That might be a way to do it. So for them to be able to also travel a lot of distance in a short amount of time means that if there is a problem, they can actually bring the likely hood of a success into a better time frame. So in other words, if a robot actually can sit with all of its sensors, see something that a human didn't see in a shorter amount of time, it can prevent a catastrophe. So that's a real good use of it. One is there. Oh, sorry. Great. So I've got two more. But go ahead. Okay. So there is another element of this. And that is cost. You're paying your security people 25, 30, $50 an hour for contract. What do you think these robot dogs? I think that what's going to happen is you're going to just be able to just like you hire a security company to have a guard station here or do this or that. You're just going to hire them by the hour and they're going to come in and do their duty. And at an hour, they rate that's lower than the human, right? Gus. Absolutely. I remember when you were, you had all those celebrities in your, in your bars back in Minnesota. And the prince would come in. He'd have his own security team, right? Couple good looking girls and Ed is brother, but that was different. It was that first security team back in those days. I'm sure there is now, but we had too many apples cops at the front door. Is one of the not legal in every state to have cops front door. We had, we had them and it was, it was a nice thing. But you know, I could tell you stories where a robot would have been able to have a little more disciplined. I could tell you stories. Tell me that. Let me finish up the last couple of usage that would might be seeing these dogs used for the world cup. One of them is crowd monitoring. Bill, you can talk about at the airport, maybe whether you can direct traffic one way or the other, you're allowed to go into the, the, the, the, the quick lane, the fast lane, or you don't have a credential so you have to go through the slower lane, those kinds of things. So they're like edge zones that it can do. So that's, that's a good thing that they could actually do. But the biggest thing I think that might be helpful is that they become psychological deterrence because people don't know what they can do. They really don't know what they do. They don't know what kind of senses are on them. I don't even know if they have some sort of weapon on them. I mean, we're saying they don't because that's what they're trying to promote is, you know, security, but yeah, calm and belief that these systems are in no way dangerous to humans. So what at the same time in the future as these machines become more and more prevalent and we start to get used to them, they can equip probably all kinds of things. Maybe they've got some sort of a chemical, you know, I'm trying to think of like pepper spray or whatever. I don't really really know. I don't even want to go down that road. If we were me putting them together, I would electrify them like an electric fence. So somebody comes over and start calling on them. They get a shock. Yeah. And so now we come to, now we come to economics where you were at, you were talking about how expensive a human is as opposed to having a robot take its place. And so yeah, definitely these, they figure that these little four-legged robots will operate for about $5 an hour over the life cycle of their, you know, their entire life cycle. So it's yeah, considerably cheaper $5 an hour versus even a low paid security guard probably in the 20 to $25 an hour range. So yeah, I think where do they go for the smart break, Jim and the water cooler. Yeah, it's the recharge break for them. Yeah. And you know, they always talk about robots can go 24, seven capability, you know, even the ones that are going to take probably do jobs and factories, but they still got to get recharged. I guess maybe some robots could be fully wired all the time so they don't have to be recharged with there in a factory. Yeah, that could definitely happen. As well, but the ones that need to be roaming around and as in perimeter work like a security guard, they're going to have to be recharged. So there'll be some downtime. But if you got enough of them, you just rotate in and rotate out very simply. So Gus, what was your favorite segment today? Well, the one I brought up of course, you know, the one about Hollywood. So it's us protecting them and I'm interested in my investment in the payphones. Yeah. How about you? Yeah, I kind of like this robot dog situation because robot dog is a perfect pet too, right? Because why? Well, you can put some fur on it, but it doesn't have to have allergy kind of fur. And they don't have to come over and lick on you all the time. And you don't have to clean up after their poop. To me, robot dog garrison today, bros. They're Ed Pony. They aren't really isolated stories. They're all indicators that hey, eyes moving into governance into the mainstream, into infrastructure, into our culture. Right. Robot. Yeah. Robotics. Yeah. And physical security now. And so it's all happening all at once. That's pretty much progress at breakneck speed for us. You know, we're seeing things happen. Look at in our generation now, well, for our generation, look, we've seen everything since the beginning of the internet all the way to where we are today. It is just really amazing. But things are moving at breakneck speed really, really fast. Now it's hard to keep up. So anyhow, thanks for watching everybody. And as always, remember to like, comment, subscribe and have a great week.