AI & The Future of Play Ft. DeLonn Crosby

"Are you ready to revolutionize early childhood education with tech that empowers rather than numbs?" On this energizing episode of the AI Rebels Podcast, hosts Jacob and Spencer sit down with DeLonn Crosby, founder of Say Kid, to explore how technology can transform learning for our youngest minds. DeLonn introduces 'Toy Bot'—an innovative, screen-free device that fuses voice technology with tangible play to spark creativity, boost executive function, and forge genuine human connections. Through a live magic trick demo and an in-depth discussion on design challenges—like making cutting-edge tech affordable on platforms such as Alexa and ensuring it’s developmentally appropriate—DeLonn makes a powerful case: if we’re serious about transforming education, we must reach kids early with tools that empower both teachers and families.

https://www.toybot.com/

Transcript

0:19

hi everybody welcome to another episode of the AI Rebels Podcast we're very excited to have Dylan Crosby Dylan thanks for coming on super excited to be here the lawn you're the founder and you know idea in inventor I mean there's so much that goes into what you're doing at Saykid with the toy bot can you walk us through maybe your elevator pitch what what is it what are you working on what's what's the big idea yeah you could say crazy crazy person too um so uh it's it's actually one of those things that's kind of hard to to understand till you sort of see it but um at a high level we know technology is not going away right um but at the same time technology was never really made for young kids uh Jacob I know you have a young child right so so touch interfaces aren't natural and these right kind of factors right they're not really playful and and screen time you know we know kind of can hurt development so if we know technology is not going away and it's not helping then you know what do we do so what say kit has done is we've designed basically a um we call it a toy bot but basically what we do is we combine voice technology with tangible play to help kids learn in a more actively engaging way without screens I love it love it is this a good time to do a quick demo of the toy seems like that might be a good segue yeah let's do that Spencer Jacob do you guys like magic always love it I actually totally tried to become a magician when I was young so love it so um this magic trick has changed say kids trajectory um so I'm gonna do a magic trick um so here's what we do we make plus plus robot call it a toy bot and we're just gonna play a quick game who wants to play like me I do alright alright Jacob alright so um Ziggy open magic card trick so I'm gonna shuffle the cards got them right here and you're gonna pick one welcome to the magic Card show can you hear can you hear it okay I will ask you to select a card and use my imagination to guess the card you select so I'm gonna shuffle dealer please shuffle the cards see me shuffling Claire please take a moment to inspect the cards are the cards shuffled yes Alrighty dealer hold the cards face out player just say when please select the card when after you've selected a card say I believe in magic I believe in magic great I was hoping you'd say that dealer place the remaining cards face down player place your card face up for dealer to verify dealer when you're ready ask me to reveal player's card let us know which card was picked you selected the 6 of hearts did I guess the correct card yes excellent what lucky guess thanks for playing with me see you in a while crocodile what in the world that is the shortest thing I could show you that kind of shows you how it works interactive conversational that is wild no that I mean I'm not a kid but I want to keep playing haha so what we discovered the reason it's it's like I wasn't joking it literally changed our trajectory so No. 1 that's the first thing I ever made I'm not a trained developer but um years ago uh when we were just sort of trying to figure out how to make this technology I was working with a couple of train developers and it was taking us a really long time took us like three months to make our first game and it was crappy and so I just spent a weekend learning Javascript and and sort of wanted to make something that I just to see if I can make something and I made that game tested it took a video of it and shared it on this platform with all these developers and it kind of went viral and uh and then I went back to these developers and I said look I built this in like a weekend like we we have to go faster so it's what it's what sort of change how we built software and then secondly um over the holidays I was literally in a mall like we do a lot of things that don't scale and it's really really hard to get people's attention so what I would do is I'd say hey do you wanna see a magic trick and like you know 100% of kids wanna see a magic trick and you know 50% of adults or so will be like and I I probably wouldn't honestly and you know inevitably they'd come over they'd see the magic trick and then the adults would be the ones that would be hanging around for 10 20 30 minutes just to wait for me to explain how it works but in the meantime we're talking about education we're talking about screens yeah we're probably playing a board game with their kid like and by that point you know kids are this is this is off the chain um kids love it and so that's how we sold most of our Tobas oh my heck that's awesome so what does what role does AI play in Toybot exactly yeah so I would say the most the most common question I've gotten over the last year has been what do you think about AI and um here's how I think yeah I'm sure cause I see it differently than what I think most people see it as so a lot of people think of a I they think of kind of generative a I like chat G B t right and I see the world almost like as a spectrum right so you have sort of generative a I which is the idea of kind of teaching a machine or training machines to sort of think and speak right or you have sort of on the other end of the spectrum kind of receptive AI where you're teaching a machine to listen and understand and that's really more of what Alexa is so what we do is we teach a machine to listen and uh and understand and then everything we everything sort of a childhood here is is programmed by an adult or human sometimes it's kids actually so it's not always an adult but I think that's the big difference and the reason that matters is um I I think we want kids to learn how to think and speak so I I think it's I actually think uh generative a I for all intensive purposes is actually developmentally not appropriate for young kids um because I think like if you were to sort of play this out a little bit um super easy for a tool like that to almost inhibit kids creativity and thinking yeah critical thinking and you know there's just a whole bunch of stuff and you know we haven't even talked about um hallucinations and where it can become really or like or like what I would call it uh sort of uh unhealthy relationships um you know that's yeah you know that's all stuff that like is not going to be good and I think it can be really really harmful and so we do the opposite of that wow that makes a lot of sense yeah no I've I've I've wondered that cause I'm you know I'm a huge proponent of all types of AI I love it all um but I have had some of the the same um concerns I guess I'd call it or or you know have wondered about similar things about you know generative AI AI and and young children um uh cause overall I'm you know opposed to AI censorship but then you know you you take the case of like a small child and it's like uh there's there's a lot of potential here to help children um but yeah as to your point there's there's no good way to currently you know well let's guard rail it I think it's helpful Jacob were you gonna jump in uh no you go I was gonna say something else it's helpful to sort of think in terms of first principles so if we were just ask ourselves what's best for a child which by the way this is literally what drives us so um what do we know about what kids need to succeed and how you develop that right so number one uh early child development is really important it's the most important period of development so so whatever we do it's gonna have a sort of a um exacerbated effects we have really really thought about what we do but what we know is you know kids um need to build executive function skills these are sort of working memory and cognitive flexibility things the ability to pay attention by the way and and I would say that like most technology is kind of hurting it's kind of hurting those skills but kids also need to develop real human connections with real people and um you know we try to everything that we do this is everything we do encourages kids to think so there's no screen so kids actually have to pay attention they have to actively engage if they don't just stops so we're we're without even trying we're building executive function skills every game we make can be played in a group why cause we wanna strengthen those real human connections and many of the games gets get kids moving in some way um and I could give you examples of all of all of those but like even I think one of the hardest things for people to understand is like wait a minute how could a robot strengthen real human connections and before I like share how we do it I I I wanna get your thoughts like how could a robot strengthen real human connections yeah I'd love to actually um because this is this is a talk that Jake and I have talked about a lot on here is that we do believe there's a ton of potential in AI to strengthen and and enhance social connections if used in the right way um so I think that but uh what you're talking about is is a perfect example actually of of what Jake and I have talked about in the past of of you know um creating a new experience that that prompts humans to you know interact right that wouldn't have existed without it um so for example another guest that we had on here Cody I what was his last I can't remember his last name right now I apologize Cody um noby dot I 0 um and he has a kit community management tool built with AI um but he he built went about it in a different way than most people I see who have you know just a a discord bot that you know interacts with people or whatever right um his is still has a lot of language models used and used with it um but his focus is on uh you know if if a particular user say user you know asks a question and then you know they don't get an answer this bot will go in and be like hey you know this user asked this question this other user knows about this topic really likes talking about it I wonder if they just missed it and so they'll you know tag them in and say like hey you know check this out things like that right um where it's it's kind of filling in uh yeah anyways yeah those are my thoughts yeah yeah Jake go ahead no I I I just think this touches on everything you're saying Delon that AI I mean it's currently right maybe someday it'll be different but right now it's a tool it's not it's not acting of its own free will and choice here so I think for a robot to increase human connection it would have to be built for that from the beginning which I like hearing that you know toy bot and say kid that's you're starting with those first principles in mind when you're building this cause it's far too easy and we've seen this with screens in general you know shows and the whole purpose is just like keep your attention child hears colours flashing and everything um so I think it would have to really kind of go against the grain of what's normal right now like I think the majority of toy or kid centered AI or technology is purely attention grabbing like that is the core purpose it's like you will cocomelon on steroids yeah this technology will watch your child and make sure they don't move basically um and it's like you're taking that flipping and sonic flipping it on its head and saying no this is going to be a catalyst you know for movement for connection yeah that's that's that's right and you know I think a good way or at least one way I think about it is I would say most technology is kind of designed to shut kids up and yeah a lot of kids a lot of kids need to feel heard you know when you're 4 or 5 and so let me share a really quick story about how we got into this um so years ago um like I by the way I didn't just wake up one day and say I wanna make a talking robot um what happened is years ago um uh my son cow was like this really happy four year old kid love school loved his teacher named his fish after his teacher and then one day that teacher left which by the way um when you get into this Jake into early childhood you're gonna find oh yeah turnover it's crazy um cause it's a broken system which is another conversation but um ultimately the the teacher left and the new teacher that came in just didn't have the same tools or the sort of training to relate to kids and I didn't know what I didn't know back then I literally thought okay we're paying this childcare centre a lot of money I trust that they sort of know child development and I don't think people realise how not awesome it can be and so I would just say like 500,000 kids are kicked out of preschool every year that's nuts to me nuts and I that's crazy it didn't make any sense to me it just didn't make sense so I literally went back to preschool but and I would spend you know years studying and I didn't want to sort of like I don't like excuses and so I literally wanted to try to be the best parent that I could be after after that happened so I would spend spent like years you know digging into science child development etcetera and what I realised is it wasn't necessarily that teacher's fault that she didn't have the training or the tools because she didn't make a lot of money and what I Learned is that you know it's really hard to sort of just like um be an early educator right like you've got 20 kids lots of movement it's it's just it's a tough sport and um what was missing was this teacher's ability to sort of connect with kids that's the thing and so I'm gonna give you an example of how we do this cause I think most people wouldn't pick up on it unless you were in early education so think of the game uh would you rather remember it from like you know when you're a kid um yeah but we we made a game literally it's called would you rather show me and uh and and we made it for that use case and so literally imagine kids sitting in a group teachers facilitating it and the robots like hey would you rather go to school on a train or a tractor uh kid says train and as a teacher you now know something about that child oh I love trains my first time on a train I went to like it's a sort of a natural organic way to sort of understand a child's interest um and kids are pretty smart they know when when teachers are trying to force it um and so this is almost like a playful way to sort of learn more about a child's interests and that's sort of a vehicle to engage kids and then you know the next question might be hey would you rather stand on one foot like a flamingo or walk backwards right and if you got 20 kids if you're an early educator and you've spent eight hours chasing you know four year olds around and you've got 20 of them standing on one foot you feel pretty good you know you did that and so we now we wanted to be a tool for the teacher to sort of um better understand kids interests yeah but also like be a tool for her to sort of help manage the classroom and make it more engaging and that's how we do it yeah that's powerful and so is this being is this currently being used in you know by teachers and early education the classrooms things like that yeah we were with early childhood um centers were teachers um we don't uh actively go at sort of K through 12 cause I think it's a broken system but we have a lot of K through 12 teachers that I use the product um oh nice okay and so but but ultimately I feel like the best way to sort of move the needle in education is reach kids where they're at which is often at home and so um our primary focus point forward is going to be how do we reach families um and so that's that's kind of our focus our focus primarily right now is is reaching kids where they're at which is often at home and so um cause I think that's the best way to actually move the needle in education for sure yeah and so so that's what we really are are focused on now um and you know I don't you know ultimately what we're trying to do is we're trying to help adults be the best teachers that kids need us to be and what I've Learned is that it's super difficult to just you know go to your average parent I have three kids like life is busy I don't necessarily wanna hear that I'm not doing enough or that I should learn more about executive function skills and take like it's it's it's hard so what we almost try to do is is we try to make it as easy as possible for for adults to do that so for example you know one of our most popular games is called the Clean Up game and literally kids teach the robot how to clean up so it's a parents win kids win um and you know it's it's a you know no screens no fights no crying um but yeah and and you know beyond that I mean you can play board games so like like what's your favorite do you guys play board games when you're grown up yeah I growing up like played it yeah when I was little though it was like sorry played a lot of sorry and UNO it's a good game yeah you can you can play games like that with the toy bot which sounds crazy and so a lot of people think oh well okay you can sort of kid will play a game like that with robot but but that's not just it you can actually play a game as a family so like we made a game recently called um Lost Toys it's like our version of clue and it's instead of sort of like um like you know have it for trying to figure out what person killed some other person with what weapon you gotta figure out what robot lost a toy which toy was lost in what room of the school so we just made it more like developmentally appropriate for kids yeah less murder centric that's probably a good thing totally yeah haha wow okay so I mean how does it how does it work like what's the framework what's it built on when a child starts up the toy bot like what's going on on the back end yeah so in addition to sort of reaching families where they're at I think it's really really really important to be able to reach all families so what that means is you have to find a way to make it inexpensive right so we're not the first to make a a sort of an interactive robot but I think we're the first actually make one that everybody can afford and how we've achieved that is we actually built an Alexa so um if we may yeah if we're not the largest developer on Alexa we're certainly one of the largest um we uh we won the grand prize in the Alexa Edtech Challenges for a literacy game but what we really do is we build on top of Alexa and we deliver so think of what we do is is making conversational games that's a good way to just think about it conversational software yeah how kids learn and then we deliver that software through the form factor of a plush plush toy to help kids learn in a more actively engage way with without screen so that's that's what makes it inexpensive interesting okay was that challenging to get on the Alexa platform or was it pretty easy haha uh I mean I would say everything we do is hard honestly yeah yeah there's nothing that I would say is like oh that's easy um it's not challenging to actually sort of get on the Alexa platform it is very challenging to make a really good game um yeah I can see that for a bunch of reasons I mean one of the reasons is I think like most technical people and um I'm generalizing here but sort of aren't great at sort of the the sort of like um social development and social development is really important uh for young kids right so so the most valuable skill set that we have as an organization or capability is the ability to speak child you know anybody can sort of make a game but actually make a game and use sort of the sort of a child's language that's actually incredibly difficult I know it sounds yeah it's incredibly difficult yeah so I'd be curious to hear um and this is just kind of a minor question uh but why why the form factor like a little robot um as opposed to like you know an elephant cat was there any specific reason for that or it's just like you love robots yeah no this is I mean so uh this is gonna be a theme but every decision we've made has been through the lens of what's best for a child so yeah early on uh so first of all we didn't wanna make it human right because right as soon as you get into sort of human type things it's nothing it's not gonna end well um moreover we just wanna get into the sort of like um blue for boys pink for girls kind of thing yeah totally we could have we could have done we could have done like sort of some animal type thing but also um what's really great about a robot form factor is you can some you can sort of like program a robot to not know something what I mean by that is to start the start of any game starts with the robot having a problem it'll say hey I'm a robot my motherboard's broken uh I forgot you know the my letter sounds and letter blends or hey um my memory card is broken uh I forgot this can you help me or like my battery is running out of power I don't know enough I don't have enough energy to um you know solve this problem that usually relates to a learning domain so cool yeah creativity I like that a lot yeah so that's the thing about using a robot form factor and persona I think that people kind of underappreciate how much little kids love feeling like they're helping um or you know I should rephrase that little kids love to help uh sometimes sometimes there's they simply feel like they're helping sometimes they're actually helping um but and and so I really like that that that the hook is like you know uh something that kind of engages that that instinct um that's super interesting very very pro social ha ha ha yeah totally yeah I mean I I just on a kind of a sidebar here I I think what we've seen consistently with all the guess we've had on which now is going up on 50 with a lot in the pipeline like we've talked to all kinds across industries and all the I maybe not all Spencer you can correct me but it seems like every single founder we've had on that is seeing success has started with they've started the organization and the project with you know why they've found those first principles like you found and we've had it others on we've talked to others maybe not even on the show where it's like they're just excited about AI and they wanna build something but they lack that purpose and that drive so I mean I I'm loving it seems like every question we're asking you're like well this comes back to our first principles that we harken back to you know the robot the the dialogue everything and so just props to you that's it's amazing yeah and and you know when we uh first of all I appreciate you saying that Jake um when we first got started remember this technology has has evolved but um we designed the voice so um you know that's yeah puts a lot of you have to really think critically about what what do you want the voice to be yeah do you want a lot of people early days thought oh it should be as natural as possible I don't think I don't think that um for the reason I said about why it shouldn't look as natural as possible so another reason about making another reason to make it a robot is that we slow down the voice and we actually try to make it a bit more robotic so we don't get in this kind of weird situation where kids start to think it is interesting yeah yeah so we're really intentional about everything even even the colors are are a certain reason um you know the original version of The Eyes were actually kind of a bluish color cause color theory would say blue and orange go go great together and then I sort of stumbled upon this research it's called the Brown Eyes Blue Eyes experiment I encourage you all to look into it but long story short I read this one okay yeah uh it's one of the best examples I've ever seen to help kids build empathy and um I just it was sort of the light bulb went off even though color theory says you should make these colors this it wasn't going to be right to make a blue a robot with guns oh yeah yeah I see interesting so okay if we one question that I had as I was thinking of this idea comes back to the data what data are you hanging on to do you hold on to any like is that is there that back and forth with the robot or how does that work yeah I love it so we don't capture any personally identifiable identifiable information I think a lot of people here um think of social robots a I Alexa and I think they haven't like to think like oh it's capturing all this data right so we don't we don't capture any personally identifiable information P I I um what we do capture are sessions so okay how long session length you know our kids um it's basically we do it for a couple reasons one it's it's basically like quality control did yeah did a child finish the game did they not where did they end um so you can see what game they're playing yeah yeah like okay why aren't kids playing this game we thought that we thought kids would love this game why why are why are they not playing it or um kids aren't finishing this game why are they not finishing oh maybe there's a bug or um you know kids aren't playing this game again maybe you know there's just different things that you can look up yeah yeah you know we don't unless we have a purpose you know we're not going to capture data just to capture data so I'd be interested in hearing is there ever um any plans to open up like a platform for like you know user generated content of some sort you know so that uh you know a kid and their parent can create an new game of their own or is the the intention to always have it kind of be obtur like where it's like you know you guys have control of the creative experience um and you publish it to you know the robot yeah I mean honestly if uh if you were to come to me and say and you were to be the best in the world at making conversational software to help kids learn it's a great make make here you go make a game um but I most people haven't spent the the energy to yeah understand child development and there's a there's a lot of ways that you can do this the wrong way and I think they're fairly intuitive ways so um so you know that again and a lot of this came from just mistakes right so um how we how we ultimately evolved to having kids play the role teacher we made a game a really long time ago it was about manners terrible game um and there was a bug and I remember my daughter being like is it trying to trick me like it was it like gave her the wrong answer and at that moment I realized like we have a really uh we have a responsibility to sort of get this right um and so that was sort of where we're like we're not gonna have kids we're not gonna be a teacher it's not it's not a teacher uh that would be the the the worst thing we could do it's really designed to help teachers and where possible we wanna have kids play the role of teacher because like you said Spencer kids love it they have agency and it's one way where kids can really feel successful cause a lot of the kids today are put in a classroom they're told to sit down and be quiet and listen and I think kids like I said before kids really need to feel heard and it feels pretty good like you said to play the role of teacher yeah yeah yeah I'm curious you've mentioned several times that um which I agree with in so many respects that the education system especially K through 12 in most areas is pretty broken um which I think is not super debatable at this point like I think there's been a lot of research a lot of problems have been identified um I'm curious if the uh the mission of Say Kid encompasses that like do you foresee also trying to work with like older kids or just helping education general or you like no we're kind of focusing on this developmental age we're gonna own this space what's the plan with Say Kid the long term idea yeah so I would just say like I don't think it's going to I don't think it's possible to actually move the needle in education without reaching kids early and that means before kindergarten so if we think that we're just you know kids are gonna start learning at age like 6 and then they're gonna stop learning at you know 18 like that's just not how it works and um so no matter what you do in K through 12 if you're not actually helping kids develop early language early cognition executive function skills it's it's it's it's gonna be a tough sport and so I think it's almost imperative that we reach kids early um because once you if you do that well everything else becomes easier right so let's say a kid age 4 3 4 5 you know has really good attention skills and has the ability to focus and has the ability to sort of play nice with others right gosh I mean that's what that's what kindergarten teachers want and unfortunately that's like not happening today um last year actually we we interviewed 100 teachers try to figure out like why kids aren't proficient in math and by the way two thirds of the country kids in the U s are not proficient in math or or reading so like clearly there's a systemic issue but um one of the things on the math front that we Learned is that you know teachers teachers would say there's kind of two big issues one is that kids um kids are coming into kindergarten um with not only like sort of not having sort of basic foundation cognition but like inability to sort of pay attention for you know longer than X period of time at the same time they would say hey they don't have the time to sort of go to the lawn and say alright the lawn needs this skill and that skill um so to help Dylan learn math and so like there's it's just a broken system unless you actually reach kids early I just don't see it getting better unfortunately yeah yeah that makes total sense hmm uh so you have a background what's that oh go ahead no go ahead I was gonna say that that doesn't even get into sort of the the sort of model the business model of K through 12 yeah right so there's no incentives for schools to necessarily help kids learn right so schools make money based on kind of butts and seats so you know and I I saw something couple days ago about how like a kid graduated with 3.4 GPA but couldn't read uh I think I think people would be surprised at at what goes on in K through 12 education but I do think it gets back to the the system it's almost like a monopoly right like and it's really hard to sort of innovate when you have a monopoly and so that's why I'm just not unfortunately I'm not bullish on the system system's ability to sort of change itself yeah that makes sense so uh looking forward what's kind of what do you see is kind of the next step for sakeid and its evolution and and um as a product and as a you know as an offering to the public yeah so I mean we're always we're always learning so I I mentioned earlier that I spent uh most of the holidays literally in a mall um talking with strangers about education and technology and and kids um and the purpose of that was like to just make sure that we've sort of ironed out on boarding you know product just all those things that I think is hard to it's hard to know until you actually do it and so you know we had no returns a lot of great feedback and so now I think we're at the point where we're ready to scale I think a lot of organizations um make a mistake by trying to get too big uh too quickly uh and that's why we haven't raised venture capital we always wanted to sort of be able to make make decisions even if it meant us going a little slower that we're in the best interest of kids so done that and now we're now we're thinking about what are the channels that we can reach the families who we can serve so um literally exploring that um the products um we you know we're always building new games adding things making things better but I think our main thrust for 2025 is just um making sure that we're going um we're focusing on the right channels to reach parents where they're at hmm okay kind of a follow up question from that if if I may um I'd be curious to hear so if and maybe you don't have an answer to this this is just a question that that came to my idle mind um if you were to you know try to target cause right now you're targeting early childhood uh if you were to hypothetically target older children uh you know whether it's the middle school range high school range whatever um do you think that that's a you know a different form factor entirely or do you see that as simply you know an evolution on the current toy bot we add some more games you know we add a uh you know I had more ideas there but my brain train went off the rails there great great yeah it's a great question um so even though I I think like it's hard to move the needle education without reaching young kids like our sweet spot I would say is kind of 4 to ten to be honest with you we know if you're if you know if you're into board games which a lot of kids are you know like monopoly and clue and Candyland like those are all really really popular games that kids love and unfortunately like what's happened today is kids are playing those games for a few different reasons you know primarily screens but um but so but board games are so helpful for for early development and and so right like our sweet spot isn't just like you know 4 5 6 it's like 7 8 9 10 and and it just um I mean my kids right now are 13 11 and nine and my nine year old and 11 year old still I mean you can actually play football with a robot toy about touchdown it's my nine year old right now yeah it's crazy oh my god and I love that this is actually this is actually his toy about and usually he has a um an LA Rams Jersey on it last year yeah so that's awesome as I dress it up it's great but to answer your question Spencer um in terms of foreign factor I don't wanna get in the game of being a sort of a toy company because making physical products is just hard however I will say we have actually developed um I'm not sure that we'll release it but we have actually developed a different kind of um foreign factor geared towards older kids that's that's dope it's so cool so yeah that's awesome so maybe yeah maybe someday haha maybe yeah maybe we'll see I mean like we'll see we'll see we'll see but it's gotta be it's gotta be significantly better than this yeah and yeah we're finding that makes sense like kids care less about sort of what the product sort of um kind of looks like and more about how it feels whereas adults care more about what the product looks like and less about how it feels yeah yeah that's interesting so if we back up and maybe go big picture AI I'm curious from your perspective as you've been you've you've talked about this scale of artificial intelligence um are there really I mean there's Alexa there's Siri I assume is similar um and then as far as LLMs go how do those marry like how do those come together like how does that work what does that look like in your head yeah so I mean um you know I don't I don't have any sort of secret information about what Amazon or apple or others Google are doing but they're all obviously building um more conversational uh capabilities and um you know I I have two thoughts on this I mean a bunch of thoughts but two two sort of important ones one is I think the cost matters so let's say you make something more conversational if it costs 10 x to do that it may not be worth it so I think that's something to think about yeah um and and as you know from sort of you know training uh AI it can be really expensive um the the second thing that I think about is uh beyond beyond sort of cost is you really want to um do things that are gonna be developmentally appropriate so let me something I've been thinking about I don't have a firm answer on this but is it actually best to make a I as conversational as possible I'm not sure I'm not sure it is the more I think about it so so what we do just tell you how we do it is we actually train adults teachers kids to speak to it like a computer so if you notice when you play the magic trick I think I said you know open the magic card trick almost as if it was like an application and I don't know for sure but I think in five years I think this is gonna be the right approach because you know kids are kids are in a sort of a sensitive period of development and if you were to sort of blur that line between um sort of real person and sort of robot or something like yeah I kind of think it's it's not gonna end well and it hasn't been something that's thought about until recently but yeah I think that's the way that I think about it and so a lot of people just assume that it's better to make something smarter make something more conversational more human like more human like and I I have a feeling like we're gonna find that that's great for some use cases like adults but for kids I'm not sure that it is because of the downside that can come with it that we don't even know yet yeah this does present a very interesting I mean Elon Musk obviously is probably the most vocal about being careful with AI despite building huge AI but he very vocal about the potential long term impacts um but this is kind of an application of that that I haven't thought about really is like how does the rising generation think about how do they view AI is it a tool is it a friend is it a person like do they view it as a person I think it I think you might be onto something that it's important to draw a distinction there like it's not it's not a human it's it's not a human being cause developmentally that might be an important distinction for them I wonder if like what do you obviously you've probably thought about this more like what what consequences do you think could come about if we blur that line so I had a like an hour chat with someone earlier today about this specific issue um it was a researcher who um kind of characterized um I told about as a as a AI character um and I think that's that's fine um but I but I also explain my thoughts on AI and and why I don't like generative AI in the context of kids um or or I think it's like there's a lot of downside um and so I think like it's incumbent on people that make technology that that serve kids that we set a really high bar um I think there's a lot of well intentioned people that think you know cause it's not unreasonable to oh let's try to make a friend or let's but until you actually think through it where you realize oh wait a minute whoa no we want kids playing with other kids and we want parents playing with kids right so if we start to sort of frame this as a friend or a companion um that that's not gonna end well and like you know we've been um uh either included or not included on different market maps but like anytime someone includes us um and it says like companion I'll either say like either change the word or don't include us on the list like because I just don't it's not what it is it's not appropriate yeah um we don't you don't ever wanna have a sort of a machine trying to be friend or or be a companion with a kid cause kids aren't gonna be able to kind of process that yeah I think and you know eventually they will but like it's just not it's just not worth it it's also the question of like you know kids already get attached enough to their stuffed animals uh what happens when that stuffed animal has a personality and you know goes through the washer gets ripped up by you know the family dog right um oh my gosh yeah so this happened recently so there's another company that has since failed um and they made a really expensive kind of robot but a lot of people compared us um and I've never compared us cause I don't really focus on competition but um but anyways his dad took a video of his kid sort of really crying because the company went out of business so this like physical robot thing um like didn't work anymore and the dad was like you bad company why did you do this to my and I like everything about the video is bad like No. 1 it's terrible you should have you should have been more thoughtful like you know yeah two company you you shouldn't have sort of framed it as a friend like there's just everything about about it so much yeah um but you don't know that unless you spend some time really thinking through the implications of that and yeah so um yeah it's a it's been in several months now but I remember seeing in the news and maybe Dylan you saw this the woman who she was the first woman to marry an AI she legally married this AI character this persona and it this is just giving me like flashbacks to this article I read like if it becomes so blurred that's where that's where things like this just become normal and happen and all of sudden this woman at for it's funny because after several months she then came out and said like oh it's an open relationship me and this robot like cause I think all of sudden she realized like okay there's uh there's something missing like there's something missing this is not a human um yes he this robot will respond exactly as I want him to but do you need that is that what you want you know do you want this like parroting back to you exactly what you want to hear all the time like that stunts growth in a major major way at a certain point it just becomes annoying too like I get I get sick of Claude for that reason I find Claude is you know way too happy to give me really elaborate compliments and I'm just like shut up man let's just let's just write this stuff yeah exactly like I know what I wrote it was pyramid let's move on I mean unfortunately I think it's inevitable I really hate to say this but um and it does come back to the the for profit motive you know if there are people that struggle with some of the social skills that maybe they didn't learn earlier it's super easy to you know think of sort of some kind of relationship with a bot um and and there are companies obviously that will monetize that and it's not unlike what we see with the relationship that I think one of you mentioned already like people relationships with with ipads right if you were to try to take an iPad away from kid that like like watch out they may freak out um so I just think like we have an unhealthy relationship with technology in general and so um but at the same time it's not going away so I just think if you're gonna focus on on kids like we do you just have to have the highest bar and you have to make have to make every decision through the lens of what's best for a child uh uh so you said that we you think we have a unhealthy um relationship to technology I'm gonna push back slightly not because I completely disagree but also because mostly because I think it's an interesting question and will push the conversation in an interesting way um so in the past we have you know we have countless examples of people naming you know their their their swords their cooking implements whatever right and becoming very attached to these items and and kind of almost storing pieces of their personality in them you know so to speak um I'd be curious to hear how both from both of you um how you think the modern manifestation of that is is different from the past um and and in what ways you know what ways that matters yeah I I I off the top of my head I think cause I've thought about this a little bit too and I think a big difference is the interactivity of the modern technology where it's like what we've been talking about right is like for example building a relationship with technology where it really is a back and forth relationship especially now with generative a I and everything all this technology where you can have full conversations you can you can really be emotionally attached and invested in this object and feel like it is reciprocating whereas in the past these tools weapons whatever it was that people identified with or became part of them it was it really was just like an extension of themselves it was like I'm not complete without my sword right like I can't go into battle without this sword because I'm so comfortable with it and that's what it is but day to day it wasn't like a relationship it wasn't like they were sitting down at to go to bed at night and they're like I need my sword right like it it just like didn't it's just very different now with this technology where it is it's a back and forth relationship now I don't know if Dylan if you have thoughts on that too I think you said it well I mean I wouldn't I don't think I have anything to add to that I think it comes down to the emotional engagement with with the sort of object and if it's sort of pretending to um to have a feelings that's where I think it's um a problem so you know we don't um you know we we avoid things where in the robot be like I don't have feelings or um we we just there are just things we won't get into like I mentioned before about manners a long time ago we we tried this game and then uh I just I just got to the I I felt like it was not the best way to do this um and so we we punted on that and there's a lot of things that we'll try um and that we can do but we just decided not to release them because we just don't think it's good yeah I don't wanna know I'm I'm I'm now curious how does the robot respond like say you know a kid just starts shouting expletives at it uh how does the robot respond in that case cause you mentioned that you know the robot will mention that it doesn't have feelings um but how does it respond to something like that so um so the way it works is you play a game right let's let's take um the clean up game um so it'll say hey I'm a robot forgot how to clean up will you help me yes cool and then I'll say okay cool uh pick up everything that has a color from the rainbow on it and I'll give you like 20 seconds to do that so robot plays some music and and let's just say it'll say alright cool did you pick everything that has a color from the rainbow and a kid says some expletive right like um what's gonna happen is it's gonna say I didn't I didn't understand that and you know did you pick everything so it's almost like I see didn't happen I like that or if they're playing the um I don't know like the would you rather game and they were to sort of start talking about something else in that game it may not matter what the the child says cause we don't care we just want them to say something that gets that prompts the teacher to learn a little bit more about the child yeah that makes sense hmm so it sort of gracefully handles it is is the short answer yeah yeah I like that no that's that's great um the lawn I'm curious your thoughts on cause you have worked so hard to really build this framework and like set up these guard rails um have you heard of the movement Yack have you heard of this yeah maybe effective accelerationism effective accelerationism it's this big movement especially in the AI community um where Spencer correct me if I'm wrong but if I could boil it down it's it's innovate at all costs it's like accelerate at all costs what would you what would be your response to that mentality from what you've seen in the AI space so do you guys know the name Clay Christensen have you heard this name the innovators dilemma yeah definitely disruptive innovation so so he was my professor and I would say I've Learned everything that I've Learned about innovation from him and a lot of people would say he's the goat um the thing about disruption is that it's rarely the most sophisticated or complex or most expensive kind of technologies that actually disrupt systems it's almost always low cost technologies that are good enough and so if you were to say hey let's innovate it at all cost I think you're going to be missing um the bigger picture and I I think I think that happens a lot like there's a lot of technologies yeah you know think about like the Apple Vision Pro or think about um what's the what's the pin the AI pin that company just died humane pin that yeah humane pin humane pin that company just died yesterday um so all of those companies built a sort of um technologically advanced product right but what they found was sort of sort of market response and it's because they built something that's really cool but it was also really expensive and so the way I think about Alexa and and I've never heard anybody share this but I think of sort of the Alexa smart speaker as the lowest cost computer ever made and if you think about it that way it's not it's never it's never gonna be perfect but it's probably the lowest cost computer ever made that's in it that's in a home and if you think about it that way and you if you were sort of bend Alexa change the voice as you know it doesn't sound like Alexa it doesn't act like Alexa it doesn't look like Alexa but it the core is really the lowest cost cost computer that's actually what a disruption is yeah yeah yeah I like that I think yeah go ahead Spence oh I was gonna ask is there actually like an Alexa inside of each toy bot or is it like a you know a speaker and Raspberry Pie running Alexa software yeah so we've actually built it all kinds of different ways and what we found to be sort of the lowest cost way to do it is to actually make a kit for smart speakers and that's effectively what we've done gotcha okay way harder than it sounds but yeah yeah I'm sure there's a lot of moving parts there yeah for sure um you know as we're as we're wrapping up Dylan you have done a lot of leg work with this company and we have founders on from all industries and all different stages of company development what would be your advice to founders trying to to make it in the AI an AI related space um gosh cause I I mean AI is one of those generational technologies it's it's disrupting systems um the things that I that I think are it's disrupting are are like I said before where it can sort of it's not the best solution like you talked about both using it for coding right so that's it's a great it's a great sort of use case for that but in terms of like um I don't like to give advice but what I would say I've Learned is to um do things that don't scale I think it's really easy to have this kind of herd mentality where you sort of watch you know see the news or or hear from other people and you wanna sort of like try to get bigger and faster I actually recommend starting smaller and going slower and um it's kind of like everyone will say you don't talk to customers but like you know if you can focus on kind of a smaller market and really really do that well to me I found that to be the best sort of advice for me uh as we've grown and and like I'll be honest with you the coolest things that we've made weren't things that like came from a lot of people they came from like one child so for example I mentioned before that we make these playboards alright so this is like our version of Candyland that came from a kid at a children's hospital who was born without arms like that's oh my gosh so yeah like we didn't we didn't think about making board games like we hadn't even thought about it but we heard about a kid who um she was 9 born without arms and what you're gonna find out about kids who spend most of their life at a hospital is like what they care most about is just feeling like a kid that's it as you know we all play kind of board games as kids and unfortunately for different reasons these kids couldn't play board games so our mission was to like how could we make board games more accessible and what we discovered is that pretty much every kid now wants to play board game with robot so it gets the example of just building for solving one problem yeah and then scaling from there I want to quickly before we finish off I just want to quickly riff on kind of AI's and accessibility technology cause this is something that Jake and I like to highlight all the time cause I think I think that it's kind of like the killer uh the killer app so to speak in terms of um how AI can be best used like I love the example of I don't remember the the senator congresswoman's name um but you know she lost her voice due to some horrific sickness and now she has it back thanks to 11 Labs I think that that is incredible work you know you know things like this where this little girl is now able to actually play board games um yeah it's just it's it's really is the great way to use it like the great equalizer you know like it it enables everybody to do what they want to do like if you want to become a web developer you can become a web developer if you wanna you know figure out how to play board games you can do that like it's this amazing tool if used correctly to amplify our like amplify our abilities I I think it's a yeah yeah now go ahead I was just gonna say so Delon thank you for for uh pushing that cause forward cause that's what yeah that's what Jake and I started this podcast to do is like you know highlight people like you who are who are using this technology and in the right ways guiding humanity forward yeah all that stuff it's important otherwise who knows where we'll where we'll end up so you've got a good model thank you guys so much it's been a pleasure chatting with you guys thanks Alan we'll stay in contact oh real quick socials if people wanna follow you follow yeah say kid what's the best way for people to do that yeah our socials are at say kid play uh so we're on Instagram Twitter x Facebook all the good ones perfect okay we'll drop that in the in the comments thanks Dylan Cool thank you