KID COMPANY & UARE.AI: WHY YOU ARE THE BEST AI WITH ROBERT LOCASCIO

Thanks for having me.

Cool.

Yeah. I have three young kids, 10, seven, and five. And you know, when you have kids in your life and devices and everything else, you get a front row seat to the horror that you hear about. You know, when I didn't have kids, I was like, 'Yeah, that's kind of interesting.' But then when you have kids, you see them with this behavior when they're interacting with their devices. And what it comes down to is there's a ton of people on one side that are making content that doesn't really fit to normal family values, like don't harm your kids and don't put bad words in front of them and bad pictures. And there's people on the other side just trying to get clicks and scrolls and we're fighting that every day. And then obviously the device itself has its own addictive behavior, you know, doom scrolling and everything else. So from all that emotional, it's really an emotional thing as a parent because in one way I like my device, you know, it does good things for me, but when I hand it to my kids, you feel like a pain. So, I felt like it'd be great to create a device that would allow us to safely provide content in the form of AI generated content that could allow children to be their best creators and not just like zombies, you know, scrolling stuff. So, that really was what I set out to do a little over a year ago. And we just launched the Kid device during Christmas, so during the holidays.

Yes. Yeah. This is the conundrum is that general models like a ChatGPT and everything like that were set up to also create engagement. So you know they're always like hey great you're awesome you're wonderful. And like when you, in our case we made models just for four to 12 year olds. They're just kid models. So we didn't have this corpus of like adult land and we're trying to now like play around with it and kind of make it kid. We started with you know for kids safe for kids. So, one of the big things we wanted to make sure is when a child is having a discussion with one of these like we have little buddies we call them that they are able to offramp to a parent quickly when it comes to something like hey someone hit me today or I don't feel great or whatever. It's not like oh really let me just get into it. It's like really that's good. Okay, let's why don't you go talk to mom and dad. And if you say well mom and dad I don't feel safe. Great, to talk to someone at school, but you got to offramp them pretty quick versus trying to create this weird relationship with them when they're vulnerable. And we're one of the first COPPA certified, which is this certification by the Federal Communications for AI because we started with the idea of being certified. We didn't build something and then throw AI into it. We're an AI native device and a lot of these other devices were teddy bears and they decided in order to sell a lot of them during the holidays they would put an API in there to an AI model like one of the Chinese models or one of the US models and they just toss it in the market and you're going to have problems.

So there's really, I'd say there's the two biggies are a child can't create or be exposed to content that can be harmful to the child and the second one is you're not allowed to use the data set to do anything to understand the child and target them with advertising or manipulate. So these are the two big fundamental things that you know the people who put it together which was the government lawyers and then the people in the industry of child products. Obviously like YouTube Kids will never be COPPA certified, you know, and they don't want to be so it's a small group of us that get the certification because it's hard to get and they really go through your stuff and they have to certify and they run all different scenarios and they look at your services. They look at the platform architecture and then it gets approved.

That's right. And you wouldn't be able to do maybe the last part of my point is like you'd have a really tough time doing that if you weren't set up originally to achieve the certification. That's it. And that's what makes it like when I started to build the product like before we got started I brought a company called PO. PO is one of the companies in the world that certifies for COPPA compliance. And we said like what do we need to do? Like it wasn't like we built it and then they came in and they're like because you're right if you already built it like trying to undo a bunch of stuff is going to be very costly. In our case we built the models the infrastructure everything from day one and the most important part is the engineers understand it because they were part of the process. So for instance we'll look at data on the device and we'll do things but like there's a line where even the engineers say like we're not allowed to look at that data, we're not allowed to use that data and we understand that. So if everyone in the org understands what it is to be certified, we want to maintain the certification. But in this case, other people they just came late to the party, you know, they like I said, they're not getting certified, they don't want to be certified, it's expensive, it's time consuming, but we wanted to do that to set a precedent.

Yeah. And I look, I'm authentic to the brand because I got three young kids and I don't want exposed to junk. And what you saw there was during Christmas there was a bunch of products that came in predominantly from China. There's one from India, but they dropped in these bad products and then there's four senators right now who basically were like, you know, they want to create now they're trying to pass legislation to put a moratorium on chat buddies and chat bots for the next couple of years on toys and because of these couple of companies that came in the market. Matter of fact, toy fair was last week. It was this past week and these specific companies, there are two or three of them that were banned from the toy fair. Like toy fair didn't want them. So I know the toy association is pro-legislation which I think is the bad idea like putting a ban on the stuff. For instance, ours will educate kids. Like there's so much you could do with these things. And we're not just about chatbots. You also can create music and you create images and stories with our product. It's not just talk with a buddy, but like banning it is such a, what are we doing? You know, like we're better off certifying it and then the consumers know, mom and dad know this is okay. This thing's safe. And then, you know, and COPPA compliance to me is it needs to be more refined for AI. I think they can do better at like defining more AI use cases, but it's there already. It has a precedent in the market.

Yeah. I feel like there's a bigger problem like we know kids and devices and social media and all that is the problem and obviously what that's doing now is that it creates a reaction to anything new in the technology space but we didn't solve the bigger problem and I think it's the same thing Dalton which is we didn't create a competitive framework. I think the best thing is they didn't want to. And social media obviously like they did it in Australia recently where at 13 and below you can't have access to social media by the way which I heard is a joke because kids what they do is they put makeup on stuff or they make themselves look older because it's a facial shot to see if you're below the age. And kids are like literally I heard like just making themselves look older and they're gonna, kids are kids like you know what I'm saying? Like there's a candy store. A kid's going to get to a candy store whether you like it or not. So I just think the best frameworks are to create frameworks that allow competition to happen in a way that potentially could allow the market to choose a winner or a series of winners that align to what the market would ultimately want. So why do parents, the question is why do we as parents allow our children to be in front of a device and be hurt, be harmed potentially. And the answer in many cases is that parents want a break in their day. They want to keep their kids entertained. There's only so much entertainment we can do when we have to play with them personally. So, we get to hand them a device and we're willing to trade off whatever that does to our children for our own freedom. And that's really what it's coming down to. You know what I'm saying? Like I'm a parent. I'm not judging anyone. I'm a parent. But like this is what's happened. So somehow we have to bring something else into a family's life that can make it as entertaining, exciting, but safe, you know, which is hard.

Yeah. Are you comfortable doing that live on the show?

Here's a little, so this is a Kid device. On the bottom it's got some buddies and then in the middle here you've got you can make images and stories and music and you can make buddies and then there's a whole educational area where you can get content. The educational content's really cool. Like there are books in here, but so you can see there's different books in here and I can talk to a book say here and I can listen to a book. Come in a second. So it's now powering up. Okay. And so I can listen to this book. 'When Princess Elsa and Princess Anna of Arendelle were little girls.' So you can read, there's also that book can imagery. And then I can talk to the character. 'Greetings. I am Elsa. For a long time I was afraid I share what I've discovered.' Yeah. I wanted to ask you a question about you and your sister. How close are you guys? Come back in a second. 'Anna and I, we were very close when we were little, but for a long time, my fear made me keep her at a distance. It was hard for both of us.' So, think of that. You could talk to this book. Princess Elsa and Princess Anna becomes very personal. Yeah. And interactive. And you could ask a lot of questions. I mean, as a kid, you could spend hours just asking random questions about shoes, environment, family. And then I know that there's like board games, books, you can make music, you can create stories like you write your own stories. I think you can make images. Let's make a story of Rob and Dalton. They both have brown hair. One's hair is a little, Dalton's hair is a little curlier than Rob's. But they're going to, they're flying up to like Mars and they're going to go to this, they're out there trying to find the secret treasure that's in Mars. But then there's this like evil doer, Zorhot, and they need to defeat him. And so, that's what they're going to do. Okay. So, the device will, we can keep talking. It'll go ahead and then make the story. And that's what it does. So, it's really cool stuff, you know, it's like it allows the child to really have this very interactive experience, creating their own stories and everything like that. So, we really created it for something, you know, very unique and different. You know what I'm saying?

And the reason we did this is that it all starts with the device because the front door is the device and you know if a child can take the device into their, instead of borrowing, they're like renting a device from their parent all the time. They take it back most of the time. Here, they can have their own device. They can go in their room and all you know as a parent that everything's cool. It's safe, you know.

Yes. That's correct. And that's just we programmed it. You know, most people won't do that. They want to have, they'll start creating a creepy like a weird relationship with the child, which is not our case, you know.

Not our MO. Getting advice from an AI when they're getting something that's tough in their life. I want them to talk to me or they could talk to someone else, but at least they are given the route and they'll follow that route, you know?

That's correct. And by the way, it goes back to social media. Like I saw it with my kids and even like my 5-year-old, I remember when he said the F-word one day, he just like he said the F-word and was laughing hysterically and I was like what where'd he get that from? Because I'm not saying that to him. He literally is watching YouTube or YouTube Kids and they go 'that's effing funny' and they laugh. That's what they do on these shows by the way. They have kids hitting each other. They have a lot of abusive stuff. They do a lot of cursing, but they wrap it around comedy. And a child, my 5-year-old thinks like when if someone says the F-word in a little show on YouTube and everyone's laughing hysterically, my kid thinks it was a joke. He thinks, 'Oh, that's a punch line to a joke.' Like, that's the end of a joke. You say the F-word. So, when he said it and he saw the reaction of myself and my wife, he all of a sudden was like, 'Whoa, what do you mean?' Like, and he was laughing. He thought we were going to laugh. I'm like, 'What? What are you doing?' And we scolded him for like you can't say that you know it's a bad word and he was like his mind for a second I could see it was like what are you talking about like it's funny everyone laughs when people say that especially by the way if a 5-year-old or he was four at the time says the F-word it's kind of funny it is funny to see a little child say the F-word for a second you have to laugh because you're like how did that happen you know it's like a...

So, same thing with the, it's the continuation of the thinking by these same companies to continue the thought about how do we get people engaged? How do we create an ad unit? How do we create this? You know, now ChatGPT just came out with their adult version. Of course, you know, of course you could sell porn and advertising. It's like I mean we can do better. Like AI to me is a revolution. Like we can do better. We can come up with new ways to make money that don't have to be so basic, you know.

Yeah. It's like look the world like you said it's we're in massive change right now and there's an opportunity here to change things massively. It's all I can say. And it's going to take a group of companies that have a different vision and a different understanding of how to use this technology and break the mold. Like we have the opportunity to break the mold of what kind of hurt us. There's also some amazing things we learned by having the internet, but there's a lot of stuff that we need to break the mold on and change. And you know, guys like me, there are people like me out there with startups that like have the will and skill to go do it. We're going to try it. It may not work, but we're sure it's all going to try it, you know.

Yeah, I gave it a shot, you know.

Oh, I turned it off. Sorry.

Yeah, I mean when I left my previous role at LivePerson, you know, the company I built was a big company eventually and public and all that. I left and I thought, you know, the alternative I was saying the alternative to AI and the current construct is large language models, large data sets that are basically stolen from us. I mean, you know, it's theft. They've taken our data sets off the internet never asked our permission because it's public. Used it to train their models. They need a lot of power to do that. So we know they need like giant data centers to do that because it's the aggregated global world knowledge of billions of people and they need power to power all that. So that set up a game of 10 companies in the world could play in that game. If you don't have $50 billion to invest, then you're not in the game, right? So you got to be building these because these models need that much compute because they're taking that much knowledge. My sense was that there could be a world where we have these small models that represent each of us that instead of aggregating the knowledge on runtime that we could allow everyone to have, hundreds of millions of people could have their own individual AIs, their own individual models and if we need to aggregate a couple of them so let's say I need scientific knowledge so I need 50 of the best scientists coming together with their AIs working as a group fine but it's you owning your data you owning the way to share your data you owning the way to monetize your data and that's the platform that we're building at UR.AI. And it's just an alternative you don't need large data centers and energy you just need your small data set, small models and reasonable compute and eventually it will go onto a device, that's why I made the Kid device, it'll eventually go onto small devices that we can run our entire model, our personal model, our personal data sets on small devices, even our current mobile devices. So that's the world I kind of see versus this giant world of large language models and stuff.

Yeah. Think of them like they're today everything's a chatbot. So even OpenClaw that we've all seen recently that was like you know you run these independent chat bots or they can work together and stuff there you can you know once you get the data set in place and these the AI has that knowledge then it's about you can bring them together we call them roundtables on our platform but there's like round tables they're independent the most interesting thing is that I can chat with my AI and it can help me my own AI can help me but then I can bring in someone else's brain into my AI for a second and maybe I'm asking I want a collaboration with my AI and someone else's. So this is not by way science fiction. This is what we're going to be releasing shortly into the world. So this is the world I see that it answers the existential threat question like everyone's about like AI is an existential threat to humanity. It's going to take our jobs. It's taking our identity. It's going to do all these things. And I just think if we control our data sets and our models, we can answer all the questions about why AI could be something that could help us be our best selves, not take our jobs. Everything's about job taking right now. Take my job, take my life, robots. I was at a conference the other day and Peter Diamandis was speaking at this conference and he was talking about Elon Musk. He just did an interview with and they were talking about that doctors will be gone in the next three years. Elon Musk he asked when do you think a surgeon will be gone because he's got his robots and he said oh in the next three years there's no need for a surgeon. And then they cut away to a robot doing surgery and they had the first robot doing some surgery on someone and it worked. So, and then they asked, 'How many people in the audience are doctors?' And like, you know, there's 400 people, 50 people put up their hands up. So, I was like, 'Oh my god.' Like, come on. Like, what are we doing? You know, like, so I raised my hand. I had my hand up. Finally, Peter stopped his speech and he came down and I said, I turned to everyone. I said, 'All of you have the data sets that they're stealing from us. Every doctor in this room, if you're online and you're putting your stuff to the ChatGPT, they're taking that data and they're learning your business. They're learning your intellectual property. And you are the guys who have the power. Stop putting the data in the models and the rest will be okay. And you'll own your data, own your model, and you can own your own AI and you can continue doing surgery.' And it's like people are like, 'Oh, yeah, right. I forgot.' So, but it's dystopian what the Silicon Valley crowd, although I live here now, is thinking about. It's very dystopian.

No they actually went to of course that's where it went and then Peter said you know then there's going to be universal income where everyone gets a certain amount of money and then everyone who's born gets to pick their purpose. I want to build a company. He's actually doing this. They're going to have a video contest of living your purpose. And I believe in living your purpose. I really do. But it's so like it's like, yeah, you're not valuable at anything. So go find your purpose. Well, my purpose is to be a surgeon. No, no, no, no, no, no, no. That's not the good purpose. Go pick another one. Like be an artist. Like he's basically got everyone painting watercolors and all the real work's being done by machines and the robots like where in the world do we live you know?

That's what we're doing. So basically you do two things, there's two models. There's one a voice model today and video. So it's like you as a, you're talking and we can, voice is more important than video by the way. The voice is where all the emotion is. So we want to capture the emotional range of a voice. So there's a model for that that we have. The second one is we call it the human life model. So when you upload data into the system we look at it and we basically break it into many different pieces. We actually show you now like all your value system what you get excited about like your emotional state. So when you put data in, even if it's factual data about some work that you're doing, it'll still try to break it up. But as you put more and more personal data in, it really gets to know you as an individual. And then that basically we put the data set together with the voice and now we do video. So we output video. So you can look at videos of yourself and you can create out of your data set ways to engage people in your life including yourself. You can engage it by chat, talking, video, all of that, imagery. So, we're abstracting all the knowledge in the brain and as you put it on the platform, we create ways in which you can share it with others. And if you're a professional, a doctor, a lawyer, accountant, we have people like this, writers, you can then share that with others so that they can engage you. You get your time back. So, they can ask you questions while you're sleeping. They can ask you questions that they normally would call you about or email you about. And all of that is personalized to the user. So who's ever using your AI, they get a very personal response based on who they are.

Yes.

Well, yes and no. Because the traditional step is that it's like an avatar like we're talking here. What we do is we actually put you in scenes so you can create the imagery the movie you have in your head you can provide that that gets created by AI based on the story and you can maybe sometimes you're in that story sometimes you're not it's just imagery of people or places and then you can interact with that or you can share that with someone they can interact with that story with that and we call it a thought what I've done is I've broken everything down into thoughts because that's how we work. Like we have small pieces of data that we've collected and we keep collecting and they sit in our brain and then we aggregate it sometimes to get a full reply to a question to ourselves or to others. Well, we make that so that you can take a thought and bring it out of your brain and show it to others. And then you can show that through videos and imagery and people can then chat with that and get a deeper dive on that thought. So like I have thoughts about like raising money when you're an entrepreneur and how to raise money. I've raised money. So I could put that thought out and you can see a little video of me like doing something and then you can flip it over and then you can chat with me and say, 'Well, I'm trying to raise money myself and I'm doing this and I actually have a fintech and I live in New York. What would you recommend?' And I would recommend for you how to raise money based on my thought because there's a big data set behind the thought.