What Kids Can Learn from Social Robots, with Paolo Pirjanian
This week Harry continues to explore advances in “digital therapeutics” in a conversation with Paolo Pirjanian, the founder and CEO of the robotics company Embodied. They’ve created an 8-pound, 16-inch-high robot called Moxie that’s intended as a kind of substitute therapist that can help kids with their social-emotional learning. Moxie draws on some of the same voice-recognition and voice-synthesis technologies found in digital assistants like Siri, Alexa, and Google Home, but it also has an expressive body and face designed to make it more engaging for kids. The device hit the market in 2020, and parents are already saying the robot helps kids learn how to talk themselves down when they’re feeling angry or frustrated, and how to be more confident in their conversations with adults or other kids. But Moxie isn’t inexpensive; it has a purchase price comparable to a high-end cell phone, and on top of that there’s a required monthly subscription that costs as much as some cellular plans. So it feels like there are some interesting questions to work out about who’s going to pay for this new wave of digital therapeutics, and whether they’ll be accessible to everyone who needs them. Pirjanian discussed that with Harry, along with a bunch of other topics, from the product design choices that went into Moxie to the company’s larger ambitions to build social robots for many other applications like entertainment or elder care.
Please rate and review The Harry Glorikian Show on Apple Podcasts! Here’s how to do that from an iPhone, iPad, or iPod touch:
1. Open the Podcasts app on your iPhone, iPad, or Mac.
2. Navigate to The Harry Glorikian Show podcast. You can find it by searching for it or selecting it from your library. Just note that you’ll have to go to the series page which shows all the episodes, not just the page for a single episode.
3. Scroll down to find the subhead titled “Ratings & Reviews.”
4. Under one of the highlighted reviews, select “Write a Review.”
5. Next, select a star rating at the top — you have the option of choosing between one and five stars.
6. Using the text box at the top, write a title for your review. Then, in the lower text box, write your review. Your review can be up to 300 words long.
7. Once you’ve finished, select “Send” or “Save” in the top-right corner.
8. If you’ve never left a podcast review before, enter a nickname. Your nickname will be displayed next to any reviews you leave from here on out.
9. After selecting a nickname, tap OK. Your review may not be immediately visible.
That’s it! Thanks so much.
Harry Glorikian: Hello. I’m Harry Glorikian, and this is The Harry Glorikian Show, where we explore how technology is changing everything we know about healthcare.
Two weeks ago, in our previous episode, I talked with Eddie Martucci, the CEO of a company called Akili Interactive that’s marketing the first FDA-approved prescription video game.
It’s called EndeavorRx, and it’s designed to help kids with ADHD improve their attention skills.
It’s one of the first examples of what some people are calling “digital therapeutics.”
And this week we continue on that topic—but with a conversation about robots rather than video games.
My guest Paolo Pirjanian is the founder and CEO of Embodied.
They’ve created an 8-pound, 16-inch-high robot called Moxie that’s intended as a kind of substitute therapist that can help kids with their social-emotional learning.
Moxie draws on some of the same voice-recognition and voice-synthesis technologies found in digital assistants like Siri, Alexa, and Google Home.
But it also has an expressive body and face designed to make it more engaging for kids.
Moxie Video Clip: Hi, I’m Moxie. I’m a robot from the GRL. That’s the Global Robotics Laboratory. This is my first time in the human world. It’s nice to be here. Oh, where is here, exactly? It’s a pretty big world for a little robot.
Harry Glorikian: Moxie hit the market in 2020, and parents are already saying the robot helps kids learn how to talk themselves down when they’re feeling angry or frustrated, and how to be more confident in their conversations with adults or other kids.
But just like EndeavorRx, Moxie isn’t inexpensive.
The robot has a purchase price comparable to a high-end cell phone, and on top of that there’s a required monthly subscription that costs as much as some cellular plans.
So, it feels like there are some interesting questions to work out about who’s going to pay for this new wave of digital therapeutics, and whether they’ll be accessible to everyone who needs them.
Paolo and I talked about that, as well as a bunch of other topics—from the product design choices that went into Moxie, to the company’s larger ambitions to build social robots for many other applications like entertainment or elder care.
So here’s my conversation with Paolo.
Harry Glorikian: Paolo, welcome to the show.
Paolo Pirjanian: Thank you. Hey, for having me on the show.
Harry Glorikian: Paolo, you’re the co-founder and CEO of a company called Embodied. And and you guys are in the field of, I’m going to call it educational robotics. But this is not your first robotics company, right? Can you can you start by filling in listeners about your history in the consumer robotics field?
Paolo Pirjanian: Absolutely. Yeah. So I actually got my education in Denmark. I got a PhD in A.I. and robotics and then moved to the US actually to work at NASA’s JPL. Which was a childhood dream job. Shortly thereafter, I got approached by Bill Gross of Idealab, who started one of the earliest incubators, who wanted to start a robotics company. So I joined that company as the CTO originally and then eventually became the CEO. We developed Visual Slam Technology, which is a vision based, camera based ability for a robot to build a map of the environment and know how to navigate it autonomously. That company in 2012 was acquired by iRobot. And we integrated that technology across Roomba and the other iRobot portfolio products to allow them to be aware of the environment and know how to navigate around the home, primarily for cleaning the floors. I was a CTO there for a couple of years and then decided to move on to do something that’s been a childhood dream, to really create AI friends that can help us through difficult times in our lives.
Harry Glorikian: But one of the projects you worked on, and correct me if I’m wrong, was the Sony’s Aibo Robot Dog, right? It’s not necessarily educational, but it was aimed at kids. So what sort of drew your focus on robotics for education and socialization, I want to say.
Paolo Pirjanian: Yes, correct. Sony Aibo, the robotic dog, my previous company, we developed a computer vision technology for it that enabled the robot to be able to see things and interact with things in the environment. And it was an amazing product, actually, the Sony Aibo. And I’ve always actually had interest in let’s call it mental health. And of course, my craft is AI and robotics. And so after my last company was acquired, I decided the timing is now to go pursue that childhood dream of creating robots that can actually help us with mental health. So we don’t categorize ourselves as education in the strict sense because we do not really focus on STEM education. We focus on for children. The first product is for children. It’s called Moxie, and it’s helping them with social emotional skills, learning, which in layman’s term you could describe as EQ, emotional intelligence skills versus IQ, which are more related to STEM type education.
Harry Glorikian: Yeah. And it’s it’s supposed to complement traditional therapy if I was reading everything correctly.
Paolo Pirjanian: Exactly. Exactly. We don’t believe in replacing humans in the loop. We want people to be treated by humans. But given the shortage and cost of mental health services, there’s always room for complementing that with AI and other technologies. And that’s what we are doing.
Harry Glorikian: So if I ask the question, is Moxie more like a toy that’s supposed to be fun, or is it a tool that’s supposed to be therapeutic or correct some help help a child that’s using it or is it both?
Paolo Pirjanian: It’s primarily a tool to help children with social emotional learning, things that you would go to a therapist for. The analogy that I use that may be helpful here is really Moxie is a tool to deliver therapy to children. But we we have to make it fun enough for the child to want to take that pill. So in a way, if you use pharmaceuticals as an analogy, a pill usually for children is sugar coated because you want them to take the pill to deliver the medicine to them. So the same way here, Moxie has a lot of fun activities and interesting things that attract a child to want to interact with Moxie. And then during those interactions, Moxie will find the opportunity to deliver techniques and therapies, for instance, to teach the child about mindfulness, teach them about emotion regulation, teach them social skills, to teach them about empathy and kindness, talking about your feelings and so on.
Harry Glorikian: I know many adults that may need Moxie for sure. With all those categories you mentioned. Right.
Paolo Pirjanian: I agree.
Harry Glorikian: But but let’s talk about the range of challenges, problems or issues that you’ve designed Moxie to help with. So can it help with relatively mild issues like shyness, or is it designed to help kids with more severe issues like, Autism Spectrum Disorder or all of the above?
Paolo Pirjanian: Yeah, no, it’s first of all, you’re talking about the audience that it’s appropriate for. Obviously, children that have been diagnosed with any neurodevelopmental challenges such as autism need to be trained on social emotional skills. But neurotypical children also can benefit from it. Actually in our customer base, we see a roughly 50-50 split between children that have mental, behavioral developmental disorders. And in the 50% are children that you would call neurotypical. But yet we know even within neurotypical children, they have to deal with things such as stress, anxiety, sometimes even depression. Covid obviously did not help it. It exacerbated a lot of mental health issues for every child, including adults, by the way, as you pointed out. And these techniques and tools that you use from therapy are really the same independent of the diagnosis. Now, some children may need more help with social skills. Let’s say if there is a child on the autism spectrum, they may not be very comfortable making eye contact, which is an important social skill to have. When you’re interacting with someone, you want to look them in their eyes and Moxie will help them, for instance, with that. And that’s maybe something that a neurotypical child doesn’t need. So Moxie will focus more on helping them with things such as coping skills, with coping with stress, coping with anxiety or managing anxiety, or even social skills. Like you can talk to Moxie about bullying and it will allow you to talk about it and understand how to navigate that and teach you skills about how to protect your own personal space. A lot of these foundational skills are are the type of skills that social emotional learning includes.
Harry Glorikian: So. Let’s talk a little bit more about the actual product. And because this is a podcast, I’m sort of like need to talk through some of the features, right? Because they everybody can’t see it. But so on the hardware side, you know, the arms, the waist, it bends, the rotating ears, the rotating base, the ears, the face, the speakers, the camera, you know, the program that animates the face and gives Moxie, a personality, the computer vision elements. Right. And then all the scripts of all the different interactions. Right, you know. Why was it important to give Moxie an LCD screen as a face rather than mechanical mouth or eyes.
Paolo Pirjanian: Yeah. Let me start maybe take a couple of steps back for the audience, as you said there are no visuals here. Think of Moxie as a AI character brought to real life. Right. So think of it as a, sorry, as a cartoon character brought to real life. So think of a cartoon character that has physical embodiment and it can talk to you. It can smile back at you. We can interact with you with body language and emotions and so on. To your question as to why it required a LCD display. We could potentially consider creating a mechanical face that can have enough expressivity, but that can add a lot of costs on one hand. On the other hand, if not done well enough, it can become uncanny and creepy. So we decided that the LCD display that, by the way, is very is curved because we did not want it to look like a monitor stuck in the head. But it was integral to the design. So it’s curved and looks like a face. And what you see on the face is an animated character, Moxie’s character, which is integrated very well with a hardware industrial design. So you can provide much more freedom of expression from facial expressions. And especially for children, you want to have a robot that has the ability to show facial expressions. By the way, the intonation of the voice will change as well, based on the type of conversation and the emotion we are trying to capture in the conversation.
Paolo Pirjanian: And then the other question, actually, a macro level question becomes embodiment, why did this need to be embodied? Why physical? Why not just a digital character on a screen? Well, so, evidence from neuroscience, from MRI, fMRI scans shows that when we interact with something that has physical embodiment and agency, it triggers our mirror neurons, our imitation neurons are triggered at a much higher level and much wider level than when you’re interacting with something just on a screen. And the implication of that is that things you can learn through interaction with the embodied agency have a deeper impact in terms of retention of the information, something that we may be able to anecdotally relate to during COVID. All education went online and the post mortem on that was that te quality of education that was delivered online doesn’t compare to what happens in the classrooms. And that’s, again, the same thing when it’s not embodied. You don’t feel that emotional connection. You don’t feel an obligation. Many children will just turn off the monitor and walk away, whereas with something that’s physically embodied, you feel you can’t do that. It has feelings, you feel it has a perspective. You can’t just turn it off. By the way, on Moxie, if you look at it closely, there are no buttons on Moxie. There is no input device on moxie like a keyboard or a touch screen or anything else. The way you interact with moxie is the way we interact with each other, using conversation, body language, intonation of voice, emotion, facial expressions and so on. There is one switch actually on the bottom of the robot that you don’t see. That’s for emergency situations in case something goes wrong. For certification reasons, we have to put that physical switch to turn it off if something goes wrong.
Harry Glorikian: So not having played with it does, and only watching the video online, Moxie’s voice synthesized like Siri or is it prerecorded? Like, how does it sound?
Harry Glorikian: It’s synthetic. Yes. So, yeah. So we cast the character of Moxie, decided what this character stands for, what are its values, what is the background story? And then based on that, decided the voice of Moxie, what it should be. And then the way you develop the synthetic voices that you take in neural network and train it based on a lot of samples that we captured from a voice actress in a studio recording hundreds and hundreds of hours of speech from a script. So we have this script and we know how it sounds based on the character’s voice recording, and that gets fed into a deep neural network that is trained over and over again until it models that voice. So that later I can just give a text and it will generate a synthetic voice that sounds exactly like that character.
Harry Glorikian: And then Moxie seems to emit a lot of sound effects and music. Does that element enhance the product somehow?
Paolo Pirjanian: Yeah. So we can underline mood and so on with sound effects or background music. For instance, one of the activities Moxie will suggest if the child is talking about things that are have to do with stress and so on, is a mindfulness journey. Where it will ask you to close your eyes. Imagine you are in a forest or other places as well. There’s a library of them. Let’s say you’re in a forest, listen to the wind and then it will start playing some sound effects in the background and calming music to get the child to imagine they’re in that space. For some children that have high sensitivity disorders to certain stimuli like sound, the parents can actually, through a parent app, provide that information which will adjust the settings. In that case, Moxie will actually not use sound effects or any jarring effects that may disturb that child.
Harry Glorikian: Interesting. So. Simple question, but is it battery operated? I mean, how long does it last on a single charge? Does it plug in?
Paolo Pirjanian: Yeah, it’s battery operated because the child usually likes to move it around. You carry the round almost like a baby on your arm. If you remember the days where we had young babies, it was literally ergonomically, it sits exactly right on your arm very nicely. And it has a battery that can run for hours of active usage. And then at night, usually like your cell phone, you plug it in any charges overnight.
Harry Glorikian: So, you know, this begs the question of where did the idea of Moxie really come from? Because you don’t decide on a whim to build a product this complex. You know, how did you persuade yourself and your investors that the technology is at a point where, you know, it could really make a difference with kids, you know, that have social emotional development issues?
Paolo Pirjanian: Yeah. I mean, the idea was sparked probably early in my early childhood, I would say. So, very briefly at a very early age due to a war, my world was turned upside down. And unfortunately, I had to flee my my homeland and seek refuge in another country where I looked different, sounded different and was different. Right? And and unfortunately, as such, you do get rejected by the society. You have a harder time in school. You get exposed to racism and rejection and all these things. So. I remember during that time I saw the first animated short by Pixar. Which was Luxo Jr., the two lamps, mama lamp and baby lamp playing with a ball. Which blew me away that a computer can generate millions of pixels on the screen that are moving to create, to induce or elicit such emotion in the audience. So that inspired me to actually seek education in computer science and robotics and A.I. because before that, as many immigrants you were taught that you were going to be a doctor, so that that’s.
Harry Glorikian: Or a lawyer.
Paolo Pirjanian: Lawyer comes second, but obviously doctor first. So so that inspired me actually to buy a computer and start coding by myself. And I started learning coding and then I decided I’m going to do well in high school so I can get into university and pursue my education. And I did. And to be honest with you, this has been something I’ve been wanting to do for since I can remember. My previous company, as I mentioned, Evolution Robotics, that was a Idealab company and I was the CTO then became the CEO. I wanted it to do it then, but that’s almost a decade ago, or maybe slightly more than a decade ago. We even tried. It was not possible. Absolutely not possible. I remember back then. Just to use an example that I think most people can relate to, voice recognition for even a single command was hard. All of us have had in-car navigation systems with a voice assistant that you would press a button, hold it down and say navigation, and would pull up navigation and say, Enter your address. It will enter the address. And you would have, to by the time you were done, enter the address because it would constantly misunderstand you and then give you options. Did you say A, B or C and no, no, no. I didn’t say that. By the time you were done entering the address, you were at the destination. So that was state of the art only a decade ago. Just for voice recognition. Same thing with computer vision.
Paolo Pirjanian: My specialty actually was computer vision. Computer vision. Also, we couldn’t recognize things very well. And the advancement that has happened in deep neural networks due to the increase in compute power, due to increase to labeled data sets that are available through many sources from YouTube and the Internet and so on. We have been able to solve age-old problems that for decades we were struggling with So it was not possible. The other piece that was probably not possible was that I was not ready as an entrepreneur probably to take on such a colossal challenge of building a product like this. So the stars aligned around 2015 when I decided to leave iRobot and said, You know what? The time is probably right now. And and fortunately, I was able to get some investors that believed in the vision of creating AI characters, AI friends that can help children with social emotional development. And obviously, this technology platform, we will in the future use it for also helping the elderly population with loneliness and Alzheimer’s and dementia and so on. We have just scratched the surface with our first products, right? And there is a lot more work to do. But today it’s possible. We have proven it. We have a product in the market. A five year old can will interact with it for months at a time without any human intervention. So yeah, so it was a series of events brewing over the last 30, 40 years for this to become possible today.
Harry Glorikian: Let’s pause the conversation for a minute to talk about one small but important thing you can do, to help keep the podcast going. And that’s leave a rating and a review for the show on Apple Podcasts.
All you have to do is open the Apple Podcasts app on your smartphone, search for The Harry Glorikian Show, and scroll down to the Ratings & Reviews section. Tap the stars to rate the show, and then tap the link that says Write a Review to leave your comments.
It’ll only take a minute, but you’ll be doing a lot to help other listeners discover the show.
And one more thing. If you like the interviews we do here on the show I know you’ll like my new book, The Future You: How Artificial Intelligence Can Help You Get Healthier, Stress Less, and Live Longer.
It’s a friendly and accessible tour of all the ways today’s information technologies are helping us diagnose diseases faster, treat them more precisely, and create personalized diet and exercise programs to prevent them in the first place.
The book is now available in print and ebook formats. Just go to Amazon or Barnes & Noble and search for The Future You by Harry Glorikian.
And now, back to the show.
Harry Glorikian: I mean, just looking at the system, there’s probably a lot of innovations that were required to put Moxie together. And so. I don’t know, maybe you can give us a few, you know, like “Oh, my God” moments that took place in this, right? I mean. I don’t know if it’s the physical movements. I don’t know if it’s the, you know, personality or the scripts. But, you know, give us the highlights of what you think was like the big breakthroughs that made this possible.
Paolo Pirjanian: Yeah. So there are many, many, many, many pieces of technology that we had to invent or partner for to make this happen. So what I mentioned, deep neural networks, generally speaking, in the field of AI have advanced to the point where we can have very reliable speech recognition technology, for instance, right? Where you have an accent or not, you’re speaking loud or soft and so on, you have background noise and so on, it will be able to transcribe what you’re saying pretty accurately. There are still errors, but it’s pretty accurate. It’s accurate enough, let’s put it that way. The next stage of the conversation pipeline is actually understanding. Now you have a transcript of what was said. Now I need to understand the semantics of what was meant, what was the intent behind this, this string of characters, and that’s natural language understanding. In that area, Embodied has made huge advancements because we have to be able to understand what the child is saying. And the state of the art when we started is defined by Siri and Alexa and Google Home, where it’s very command and response. “Alexa, play music for me. Alexa, how is the weather? Alexa, tell me a joke. Alexa, read a story or read the news for me.” And so on. So short utterances and and direct mapping to a function that the device can do. Whereas in our case it’s not about this transactional command and response, it’s about relation and social interaction. So the child, Moxie will actually ask and encourage the child. It says, “So how was your day to day?” There is no way any human being can script all the possible answers that you could expect to hear because you could basically say anything possible to that question.
Paolo Pirjanian: So we had to develop natural language understanding that can understand what was said no matter what was said, and provide a relevant response. Because if you don’t, if the robot says something that’s absolutely not related to what the child wanted to talk about, then children get disappointed. They say, well, this thing is a dumb robot. It doesn’t doesn’t understand me. And they will dismiss it, right? The illusion of intelligence breaks away very quickly as soon as you you misunderstand or say something off script, let’s say. So we had to develop a combination of systems to be able to address that. Another major challenge, and this was actually much bigger than I thought, we spent a lot of time on this challenge to solve. Again, it has to do with interaction using Alexa as an example also, and Siri as well as Google. They all have this notion of a wake word, Hey, Google, hey Siri or Alexa. When you say this keyword known as a wake word, the device is actually at the, when it’s on standby, it’s putting all of its attention to look for that keyword before it does anything else. So as soon as you say it, a couple of things happen. It’s almost like turning on a switch to say, I’m going to speak, right? So number one, you’re telling it, I’m going to say something now. Number two, as soon as you have said that phrase, these things have multiple microphones on them. And the mic array allows you to be able to be informed and focus your attention on the location from which you heard this phrase. With doing that, you can also filter out anything that’s in the background. So you focus the attention of the device on that location of the user that said Alexa. And then you say a phrase and then it processes and executes the action. In our case, in social interaction, it will not be appropriate if you had to say Moxie in every volley of the conversation. Every time you want to say a sentence to me, you would start by saying Paolo and I and I would look at you, and then you would say something, and then I would stop listening. And then you say, Paolo, for every sentence, right. That would that would be a very awkward social interaction. So we had to solve that problem. It’s a tough problem to solve. And we use a combination of cameras to know where the child is, the voice, where it’s coming from, and what was being said to focus the attention of Moxie on the person that’s engaged with it so that Moxie doesn’t respond to the TV or mom and dad maybe having a conversation on the phone over there and it filters all of that automatically, without the need for having a wake word phrase. And I can go down the list. There is many, many more. But this is just examples of the type of things we have to solve.
Harry Glorikian: So, you know, I think some people might make the argument that kids should really be learning their social and emotional skills from other human beings. From a parent, from a teacher, from their peers, maybe their therapist if they have one. You know, how can a robot fit into that picture in a healthy, productive way? You know, how would you respond to the potential criticism, which I’m sure you’ve heard before. When a parent who buys Moxie for their kid, are they offloading their parental responsibilities?
Paolo Pirjanian: That’s an absolutely valid concern and a good question to ask. And obviously, even before inception of the company, I personally myself was thinking about this because there is a there’s a contradiction in saying that a child that is not very good at social interaction, let’s put them in front of a robot, they’ll get better at it. There’s a contradictory element to that potentially. Right. So let’s put it this way. In the extreme case, what if the child does not have the ability to have interaction with their peers? Right. So they do not get the opportunity to interact with other peers from which they’re actually learning to hone in their social skills. Well, that happened during the pandemic. There’s a huge mental health crisis happening in the US now that will take years for us to to address. That was because children were locked in their home without the ability to socialize with other children because of worries about being getting COVID, right. So now pandemics are rare events that hopefully don’t happen that often. But now let’s put ourselves in the shoes of children that are, for various reasons, are not successful in providing social interactions. An extreme case is a child on the autism spectrum. That does not have the right skills to have social interactions nor interpret social cues in a conversation. Let’s say if you’re annoyed at someone on the spectrum, it’s likely that they may not even understand that you’re annoyed at them and they may keep saying the same thing or doing the same thing. That’s going to make you more and more agitated or the other end of the spectrum, which is not as severe.
Paolo Pirjanian: My example when I was a child. And I lived in a foreign country where I was different. I had an accent. I looked different. I came from a different cultural background and other kids didn’t want to play with me. And there’s everything in between. Right? So then. What do we do? Well, you can have therapies and that’s what we do. There’s a massive shortage of therapists. If you have a child, usually the way this works is that your school teacher will come and say, we think your your child may be on the spectrum or your child may have ADHD or your child have some other neurodevelopmental challenge. You should get your child diagnosed. Okay. Hopefully no one has to try this. The waiting list for getting diagnosed is minimum six months, minimum six months. And that’s if you have connections and good providers and all these things. While imagine for six months your mind as a parent, you’re like, dying. What the hell is going on with my child? I’ve got to figure this out quickly. Once your child is diagnosed and you spend 6000, 7000 hours on that, then you’ve got to find providers. There’s a huge shortage of providers, and even when you get to the provider, there is a massive cost associated with it. So typically children on the spectrum, as an example, get diagnosed at the age of three or so. Ideally, actually, because the sooner you can intervene, the better the outcomes. And when they’re diagnosed, they will be recommended to seek 20 to 40 hours of therapy per week. 20 to 40 hours of therapy per week. Yeah.
Harry Glorikian: They’re not doing anything else.
Paolo Pirjanian: No. And many times, many times schools are supposed to provide it. But you have one or two special needs teachers that are to deal with the whole population of kids on the spectrum in their school as an example. So they’re not going to get 20, 40 hours per week. The cost of therapy is super expensive. Insurance also has to pay for it. Nowadays, they’re mandated to, but the cost still adds up. On average, a family will spend $27,000 out of pocket per year, even despite insurance coverage. So not everyone has access. And also if you live in rural areas and so on, you don’t have access. So. Why not have an automated system that can do this, at least filling the gap? Right. We think of Moxie as a springboard to the real world. So we want to use Moxie as an opportunity to for the child to open up to Moxie, use that as an option, teach them a number of techniques for how they can be more successful in social interactions, and then Moxie will actually encourage them to go in the real world and experience these things and come and tell it about what what, how it went. So we use Moxie as a springboard to the real world. There is another phenomena that happens, and I don’t know how to describe this. You may actually have more insights in neuroscience than I do. Children, especially children that have neurodevelopmental challenges, open up to a robot like Moxie better than they do to humans.
Paolo Pirjanian: Let’s take autism as an example again. I remember the very first experiment we did with our first prototype. We took that prototype to a family’s home. They had a ten year old son on the spectrum, and we put Moxie down. At the time we did not have the AI yet. It was the robot remotely controlled by one of our therapists. On an iPad they were typing what the robot should do and say. The child immediately opened up and start talking to Moxie. And if you look at that child, you say. And you know, as a matter of fact, I asked Mom: “I don’t see anything wrong with your child. Why do you think he’s on the spectrum?” And he says, well, you have to see him how he treats his peers. He doesn’t open up to them. He doesn’t want to talk to them. When he comes home from school it takes me, mom, a couple of hours to “find,” quote unquote, my child. Tuning into the channel. So they shut down. And there’s a few reasons for for sort of, I think, anecdotal or maybe rational reasons to why that is. One is that children that are on the spectrum, they completely understand feelings and emotions and so on. They are not very good at expressing themselves or or showing their feelings, but they understand if they are being rejected or teased out in a conversation and so on. So they shut down. A robot is non-judgmental, right? They understand that it’s a safe, non-judgmental space.
Paolo Pirjanian: The other part is that when someone like me who comes with a warmer blood and too many gestures and intonation, voice and expressive, it’s too much there’s too many signals going on. And that’s overwhelming to a lot of children on the spectrum. And they shut down. It’s too much. I cannot deal with this. Right. And so hence, a robot is finding social doing social exercises and experiences on training wheels. And helping them develop those muscles and get better at how to handle different situations when they go in the real world to interact with their peers or other people in their circle, social circle, to be successful. And that success will hopefully breeds more success. So ideally we are successful when people actually stop using our product. And as a matter of fact, we have parents reaching out to us and say, my child could not stand up in front of their classroom to say a word. Now she stands up and gives a whole presentation and we have stopped using Moxie. Thank you so much for the help that that’s what what it is. It’s like it’s stepping stone. It’s training wheels for social emotional learning so that they can have a chance of being successful, because otherwise they do not have the chance to to have these exercises to learn. We learn a lot by interacting with each other.
Harry Glorikian: So the company describes Moxie as just the first iteration of a larger platform that I think you call SocialX. So what is SocialX and what other kinds of products do you envision coming out of it?
Paolo Pirjanian: Yes. SocialX is our technology platform, which which allows a machine to interact with us using real conversation, eye contact, body language, gestures, intonation of voice and and for the machine to do that as well as understand you on all those channels as well. That’s what social platform is. The name SocialX is a juxtaposition to user experience, UX with an emphasis on the social experience. Right? We are creating a social experience. We are not just creating a user experience where you can push buttons or say a command, play music. Tell me the weather, what’s the stock market like? But rather social interaction which involves social skills, emotion, skills, empathy and so on. And this is our first iteration. It’s going to get exponentially more advanced. With every single user we add to our customer base, it allows us to improve SocialX because the data and the interactions that we can experience allows us to keep improving our algorithms to get better and better and better. So we decided to start with children because they are the most vulnerable in our society and we thought that’s where we can have the most impact. The other end of the spectrum, where we become vulnerable again is when we are aging, right? And mental health is extremely important for aging people. And loneliness leads to a lot of mental health challenges that lead to a lot of physical challenges.
Paolo Pirjanian: We know this. The surgeon general of U.S. said a couple of years ago that loneliness for elderly is equivalent to smoking a pack of cigarettes in terms of the health implications it has. And it’s true. Again, during COVID, a lot of elderly that were alone suffered massively because they were high risk for COVID. Even my mom, who lives 5 minutes away from me, I didn’t visit her for a few months until we sort of figured out that we think we know how to handle COVID so it was safe to to meet meet each other. It’s extremely difficult. So that’s the other end of the spectrum that we intend to address. And then in between every age group, in between that, from your teens to your aging, every person in their lifetime deals with mental health challenges. As a matter of fact, the US population, 17 percent of the population at any given time deals with mental health challenges stress, depression, suicidal thoughts and so on. And having a life coach that can help you through these difficult times, we believe can have a huge impact. So eventually with those three pillars, we will be able to help the entire population. You can go beyond mental health, which is what we are focused on, because that’s where we think we can have the biggest impact you could imagine.
Paolo Pirjanian: You go to Disney Park and you could have an interactive character coming up to you that’s not a person inside a suit, but it’s actually an animated character that’s walking around and talking to you and entertaining you. You can imagine going to a hotel lobby where your intake to the lobby will be serviced by an interactive character, AI character. By the way, we are also working with hospitals and schools. Right now for hospitals we work with University of Rochester Medical Center. We are currently doing a pilot of using Moxie to help children, diabetic children, to educate them about how to treat themselves and how to adhere to their treatment plan. And then there is a number of other use cases that we are going to expand into, including intake to the hospital, dealing, sort of holding their hands and making sure they are not stressed out, coming to the hospital for the first time, pre-op and then post-op. Also a lot of complications you want to avoid by making sure there is someone to remind you about your care plan and so on. So to be honest with you, the sky is the limit. But the three areas we are focused on is children, elderly and then everyone in between that suffers from mental health or loneliness type of challenges.
Harry Glorikian: Yeah, there are so many other applications that I can think of that I would, you know that I could use my self. So hopefully, that will come into play because this would be something interesting for me even to interact with, depending on, you know – Don’t forget to work out or, you know, there’s something that you interact with regularly. Right. But so let’s go to sort of the crux of the some of the issues. Right. It’s it’s not an inexpensive device. I mean, it does a lot. So you can’t expect that it’s going to be inexpensive. Right. It’s it’s $999 to purchase plus a separate monthly subscription of about, what is it, $39 per month for a minimum of 12 months. And so how how do you get this out to a larger group of people that really need it. Is it subsidized purchases? Is it insurance? What are you guys thinking of from a business model perspective?
Paolo Pirjanian: Yes. So we actually launched the product in the second half of last year for the first time and we sold out. But I agree with you that it would be much better if it was more affordable, because we don’t want this to only be a product available for high income families, for rich kids to use a derogatory term maybe. We want it to be available to every every child. And for that to happen, there is a couple of different strategies we are pursuing. One is that once we get to a scale of efficacy studies that are convincing enough that we can get insurance, potentially insurance coverage to cover it or at least subsidize part of it to make it more affordable. The other approach is that we are working with bigger institutions such as hospitals and schools and libraries, by the way, which can buy it and make it available to their population. As an example, this library actually came to us, which is a very interesting business model that addresses the reach to the society that may not be high income. The library bought a fleet of Moxies from us, and they’re lending them out to their society, to their members as a book. So a child gets to take Moxie home for a month and then bring it back, which is awesome because we have, by the way, we have done efficacy studies and it shows that even within a month you can see significant improvement on a lot of these social emotional skills.
Paolo Pirjanian: But ultimately, that’s that’s how it goes. And also, just to put it in perspective to two examples. One is that robots of this nature….By the way, there is nothing like Moxie because the technology has not existed today, but people have tried, actually, SoftBank has a subsidiary called SoftBank Robotics that have spent hundreds of millions of dollars developing this robot called Pepper that costs $14,000 to buy and $2,000 a month to subscribe to it. Yeah. So we are orders of magnitude better than that. And that was part of the design principle that we said we want to be on par with an iPhone ownership of a cell phone. Buy it for roughly about $1,000. And you pay roughly about $50 a month in subscription. So we met that goal, which was a major accomplishment, very hard to do, but we are not satisfied with that because as I said, this has to be available. The other part of the other example is that if you have a child that needs therapy and if this cuts your therapy by a handful of therapy sessions, it pays for itself. Right? Again, ideally, we will have insurance pay for it. And so that will take some time. As you know, sort of navigating the medical fields and insurance organizations and so on will take some time, but we will get there eventually.
Harry Glorikian: Yeah, I mean, I recently interviewed the CEO of Akili Interactive, Eddie Martucci, and they are the first group to get an FDA approved prescribed video game for children between eight and 12 years old with certain type of ADHD. And so, you know, they’re using the prescription route as a way to have somebody pay for the clinical trials and everything else and the product itself. So I know that this business of robotics is not for the faint of heart. I mean, there’s there’s many different companies out there like Jibo, which was out here. Or I think there was a company in in San Francisco called Anki that, you know. You didn’t pick an easy one, that’s for sure, Paolo.
Paolo Pirjanian: Definitely not. Definitely not.
Harry Glorikian: But but, you know, I you know, I wish you incredible luck. I mean, this this thing sounds so exciting. I mean, it brings out, like, the Star Trekkie guy in me and wants to interact with it and have it do certain things or say certain things or or maybe even like interact with my wearable and be able to see something and then make a comment to me as I’m using it. So I can only wish you incredible luck and success.
Paolo Pirjanian: Thank you. I need it and I appreciate it.
Harry Glorikian: Excellent. We’ll talk soon.
Paolo Pirjanian: Talk soon, thank you so much for having me.
Harry Glorikian: That’s it for this week’s episode.
You can find a full transcript of this episode as well as the full archive of episodes of The Harry Glorikian Show and MoneyBall Medicine at our website. Just go to glorikian.com and click on the tab Podcasts.
I’d like to thank our listeners for boosting The Harry Glorikian Show into the top three percent of global podcasts.
If you want to be sure to get every new episode of the show automatically, be sure to open Apple Podcasts or your favorite podcast player and hit follow or subscribe.
Don’t forget to leave us a rating and review on Apple Podcasts.
And we always love to hear from listeners on Twitter, where you can find me at hglorikian.
Thanks for listening, stay healthy, and be sure to tune in two weeks from now for our next interview.