Under-Representation in Robotics and AI is Solvable

instead with anger, fear, and apprehension. And this is the reception that has met artificial intelligence, that potential negative of being reliant on this software, on AI, on robotics, that the negatives are less than the post as it is. But it's not just the fear of reliance on these technologies. It's also the fear of the technology magnifying and multiplying our worst two intendencies. There have been troubling instances where bias and robotics and AI have been identified as a

potential disqualifying factor for their wide deployment. These systems have some aspect of bias, time and time and time again, they're still better than the human biases. Doctor Ayanna Howard is the Dean of the College of engineering at the Ohio State University and one of the few non white male roboticists in the field. So I don't think that these systems can ever get to zero bias, because there will always be a group that the system has not

interacted with. It might be that it's perfect, as perfect is perfect, and then there's an unknown community in South Wales somewhere that had never interacted right, and now it doesn't work with them. Doctor Howard founded a robotics company called zy Robotics. They develop mobile therapies and educational products for children with special needs. She firmly believes that the proper use of AI and robotics is an invaluable asset

to humanity. Underrepresentation in robotics and AI is a solvable problem. So when I was a kid, me and my sister got left at home a lot during the summers, which meant that we watched a lot of reruns on television, and one of the things that we watched all the time was The Biotic Woman, and I heard that that

was a favorite of yours as well. It was um although I will tell you the Beyondic Woman when I was young was current, so I'm sure you must have saw the reruns, say but yeah, that was that was my favorite, and I think what you know, I was always into anything that was science fiction, everything from you know, Battlestar Galactica, Star Wars. Do you mean the yeah, no, yeah, the original, not the ones that they tried to remake over and over and over again. Back then it was

all of it was like new right, it was like fascinating. Yeah. So The Beondic Woman was this show where this this woman was horribly mangled in like this this horrific car accident, right like accident, it was a skydiving she would have died, right, and instead the doctors took her and rebuilt her basically added the Beiondic parts and she would go around saving the world, which was was it was awesome, but she was human, like she had a personality. That was the

nice thing about that. Yes, like she wasn't exactly a robot, but she was like she was a robot. It was actually if you think about it, there was only two sci fi shows that showed women in like a positive superhero light. One was wonder Woman and one was The Roonic Woman. So like one of the earliest forms of a superhero. I saw was the biotic was So I appreciate that. But tell me a little bit, how about how that watching that show kind of pique your interest

in STEM. Well, so when I was watching all of these things, it also coincided with the time in middle school where you have to define the rest of your life. So they still do this, right, you have to write these essays about what do you want to do when you grow up? And so I wanted to build a broonic women. So that's what I wanted to do. Now. Of course, as my teacher said, well, building a bionic woman is not actually a career, so you have to

think about, like what label do you want? And so originally I thought I wanted to be a doctor because that was who put her together. So what changed? I took biology and we were dissecting frogs and I remember we had to learn how to kill the frogs and open them up, right, I mean this was this was the days because you had to do this. Yeah, and I hated it. I mean I absolutely And I was always good at math and science, and I was like,

I do not like this course. And then I thought, man, if I don't like biology, how am I going to go to med? School, How am I gonna actually start this whole process of being a doctor. But I had a teacher who said, hey, why not think about engineering? So what is it that you do now? I am a roboticist, which means my research and my practice is about building, designing, programming robots with the focus on the human, improving the human quality of life, with an even more

focused initiative effort on children, so pediatrics. And you started off designing rovers for NASA. So my very first job as a roboticist, I was a robotics researcher and my task was thinking about future rovers and how do we enable them to navigate long range traversals on Mars. So how did you go from rovers to the types of robots that you're making today? You're really thinking about them

in a way of being assistive towards humans. Yeah, so I think one of the things about my perception on robotics systems is really the reason why I do it is to assist us and improve our quality of life. And you know, mind you, when I used to talk about robotics, the first thing that people would say would be like, oh, you're You're the one that is taking over the jobs you're the one that took you know,

a lot of people in manufacturing. I had people say, yeah, my grandfather got fired because of your robots, like that was a common thing. And I understood that from very very early on, and those were not the robots that I wanted to build in design. But I think because I always understood that I am part of an ecosystem, I am part of a community, and my responsibility is

to become a contributing member of the community. And anything I did, because my talent was my mind, anything I did was about a overall positive net result versus a negative you know. I was the kid that had to go out into the neighborhood and paint over the graffiti and pick up the like that was our Saturday. Like. I didn't get to go out and play like the other kids until I had done my chores, which was

you know, all of these community kind of things. And so I think I just grew up realizing that it was about community. But also my talent was my mind. It was how I designed a thought about robotics. When it comes to the types of robots you designed to help people who have disabilities, is there anything personal that happened to you that made you connect with them specifically.

I had always done these stems, so science, alogy, engineering, and math camps primarily to engage girls and I'll say underrepresented minority, since that's what we were called back in the day. I had ran this one camp where there was this young lady who had a visual impairment. She was bright, bright, like smart, but the system did not work for her because it wasn't accessible. And this was the first time I'd heard this, were like, what is

accessible and what is assessively technology? Because I hadn't no idea, And what it was is like I saw a problem. I was like this, like makes no sense. Why is our technology not able to be used by everyone? As an engineer, you know, I see these kinds of problems as basically a challenge to design a solution. It's just the way I'm geared and think about it. And so that started me down this rabbit hole of looking at this target demographic understanding that we as engineers aren't really

addressing the needs of all the different populations. And I was always a proponent of diversity, and I didn't really understand until that moment that diversity includes raise includes ethnicity, includes gender, but it also includes ability and disability. So you have a number of apps, games, and toys all under their company, zy Robotics. Is there a particular product that you're like, you're especially proud of. I actually enjoy all of them. I will say the one that I

especially like is one called Tommy the Turtle. Learn to Code. Help your kids learn the basic of coding with Tommy the Turtle. Your little one will absolutely love interacting with Tommy and his colorful friends as they gain valuable skills in programming. I can get a four year old to learn how to code and love it. And you basically have to do things like, you know, Tommy wants to play with cat, but cats too far away. Can Tommy

move towards cats so they can play right? And so there's a always these this you know, goofy rhyming and stuff, and so then it'll go through like okay, in order to do this, you know, you have to put in this little button which is move one space right, and so it walks you through this and at the end it says, guess what you've done your first coding program.

Now let's try it with dog. And so it basically builds up this sequential understanding sequences, understanding logic with the subjective of again having friends, playing around and things like that. And the Tommy to try to learn to code is also accessible, so if you have a children within motor disability, it actually has accessibility functions so you can use things like switch devices. If you have slight visual impairments, there

are things you can do with the print. So I'm most proud of that because I've actually seen like four year olds like excited about coding, which is amazing and fascinating. What types of benefits do you think children with disabilities received from working with robots that you design? So one of the things that we know is that with anything repetition, repetition,

repetition is good. The problem is is that the amount of and I'll just call it exercise, which is repetition, The amount of exercise you have to do has to be done consistently, has to be done in a repeated fashion, and typically you need someone to kind of guide you because you know, they're kids, and we just don't have

enough resources. So most parents do not have are not as privilege enough to bring in a human therapist into the home every day to work with their child, And so what we focused on was designing these robotic systems that are adaptive so they have artificial intelligence in them so that they can work with the kids basically every day. So they are augmenting the services of a human therapist.

But in the home environment. Did you see an uptick in people being interested in this type of technology during COVID when you wasn't so easy to have like a human aid in the house to help with children with disabilities? Yeah, so I will say yes, and that goes across the board in terms of AI, so artificial intelligence software as well as robots. There was an uptick in the use since March of twenty twenty, and it's primarily because of the fact that you know, humans were dangerous, right, like

we know this. This was the whole thing is like you're shut in. You're locked in because other humans who are outside of your community, outside of your home are dangerous, whereas robots were not. And so there was this whole change and shift on the perception of robots, like robots were more of like, oh, this actually enables me to

have a little bit of semblance of livelihood. And we saw an uptick in companion robots because again, you couldn't have you know, your young grandkids, for example, coming in, or therapists necessarily come in, but robots could because they were safer than people. There's folks that feel like our addiction and technology is a problem, and building technology that assists humans in whatever capacity kind of builds on that

addiction a little bit, or it feeds into that addiction. Rather, what's your response to that to people that would would say that maybe we need less technology and not more. Yeah, So I actually have two answers to that. One is, yes, it does feed on these addictions. Like my robots would not work if I didn't model them based on humans propensity to interact with these systems, right, it wouldn't work.

So we know that as a fact. But I would say that that potential negative of being reliant on the software on AI on robotis that the negatives are less than the positives, and the positives are. It improves so so many things. It makes access much more accessible in terms of you know, equal opportunity about jobs, education, food sources, and things like that. So the positives I think outweigh the negatives over and over and over and over again.

When you see this type of interaction and you see, you know, children interacted with the technology you've created and the tools that you've created. How do you feel like, how does that make you feel to see them interacting with it successfully? My very first moment, I remember it to this day. We were this is at Georgia Tech. It was one of our very first studies. We had a gamified therapy protocol called Superpop. We had to pop these bubbles and it was linked to some movement therapy.

So I remember we go into the home and there was this young child, child with cerebral palsy who was in a wheelchair and severe sebesticity, so very limited movements. And I remember the game because the game adapts to the abilities of the child, and so you know, the very first one, it started adapting and started adapting. And I remember at one instance, this child touched the bubble because it's popping bubbles, right, and I just I remember like there was this smile. It was like the sun

like shone in this room. It was like all of the wats that are out there, it just shone in his face, pure joy, and it was like you you couldn't have sold it. You couldn't have bottled it. You couldn't have like bitcoined it right. It was like that pure joy that only a child could show. And at that point that was like my first moment going like, oh my gosh, I feel really really really really good.

I like this feeling, I really do. It makes me realize that I'm doing the right thing, impacting directly someone's life. You're now working with children with disabilities, specifically in pediatrics, and you know you said you got there because you first wanted to work with with women, and being a woman in STEM, I know, is not the easiest path for most Do you find that you had any personal experiences that kind of led you to say that I want to kind of kick the ladder down for more

women to come into this field. Yes, when I started at NASA and things like that, I was always passionate about working in with young kids. Actually, even in undergrad I tutored at the local high school and things like that. So I'd always done that just because I remember growing up, you know, engineers would come to my school and I was inspired, and so I just felt that that was

a responsibility as an engineer. I didn't really feel that I had a responsibility to be quote unquote, I guess a role model until I was much older, and it was because I suddenly realized that I was tired of not seeing myself and therefore, like, I'm old, So it's not like I can produce myself right now, that's my age age, but I sure could start reaching and making sure that the generation behind me saw themselves and didn't

have to go through the same things. And so it then became more of a mission than you know, things like oh I'm doing out right now here. No no, no no, no, it's a mission because it sucks when you don't see yourself. Are you speaking specifically about women? Are also specifically about black women? So I actually have three. It's women marginalized community, so black, Latin X, Indigenous, and it's also black women.

So it was actually three because those three there's not a lot of us in any stages at the upper echelon, it's not a lot. How do you think enfranchising those groups that you're talking about, how do you think that's solved? Do you think it's just solved with them just seeing you in the room? Ward? What's the extra step you feel that needs to be taken in order to draw

more of those people into the room. Yeah, so being able to see yourself is only the first step, but it is an important step because if you don't even see yourself, you're not even going to think that it's something for you. Right. So one is about, you know, getting this motivation. The other is about ensuring that there's opportunities.

And so you know, if I see myself, but I don't have access, I don't have the curriculum in the schools, I don't know, you know, where am I supposed to go in order to learn how to do artificial intelligence or robotics? Then I might see myself, but I don't have a path for it, right, And so it's about seeing yourself, believing in that, and then having the opportunities to excel. Irrespective of your environment or how you're growing up,

are those individuals who are around you. So there's a lot of people that are afraid of a future with robots at it. What would you say? How would you talk to about and and you know what, I was one of those people. I watched I robot and I was like, no, no, not like that. That. What it's easy to anything these days? What do you think the future looks like with robots in it. As a roboticist,

what do you think that future? And actually, and I'm sorry to take another step back, but those those Boston Dynamics videos are really scary people when they see the robots though in the backflips. And of course we've all seen the Matrix. I'm now referencing all the movies. But what would you say to people who have a fear of the future with robots in it. The fear that people have, honestly is the fear of the uncertainty, which

you're going to have irrespective of what it is. You know, I think about you know, horse and carriage and the cars came. I'm sure there were people like, oh, automobiles, you know they kill people like people die in automobiles. Why would you ever get it? Right? Like every single instance where something has transformed our society for good, there's always been this fear always always. We can go historically

when this has happened. And so robots in AI, artificial intelligence is now that technology that is causing fear because of the uncertainty. And this is happening and whether you like it or not, whether you're afraid or not, it is happening, and it will accelerate period guaranteed. And so what I say is figure out how to become part

of the solution. If you're you know, fearful of the I robots are feel full of, you know, bossing dynamics kind of thing, well, become a computer scientist and go work for those companies so that you don't have those kind of robots being developed. Right, Like, this is like part of the solution and also allows you to conquer

your fear but also take charge of your future. I don't know why the idea of me taking years and years of trigonometry just so I could go into box of dynamics and be like, hey, what's going on in here? What do you guys do these robots just sounded very, very funny to me. But I love that. I think that's a great solution. Well someone will hopefully someone's listening. It's like, yeah, I love that idea. I'm doing it now. Can we expand on the bias and racism forbid it?

Those stories do come to life, and I feel like there has been some truth to some of them in the research that I've done. Can you talk a little bit about maybe dispel some of the some of the fear that folks are having when it comes to bias and racism from artificial intelligence and front robots, Like, what exactly are we seeing there? Yeah, so there is there is truth. So a lot of the systems that are

deployed that are out there have aspects of bias. And what the definition of bias is that there are differences in the results in their behaviors toward different types of people based on gender, based on racial, ethnic identity, sexual orientation, religion, like basically any type of attribute you can think about. These systems have some aspect of bias. So that's the negative. The positive, though, is that time and time and time again, they're still better than the human biases. I think our

systems can be better. I honestly think we can design and build robots that can make us better humans, that can make us less biased humans. I totally believe that. But unless we really focus on that, we're just going to keep doing the same thing and propagate and just continuing the biases that we have of the past. That's what I'm afraid of. What do you say to the activists who just heard this and heard you say that

robots are biased but less biased. How do you call dib down to say, to get because I'm sure what people are going to want to hear you say is like they're not biased at all. Like we get there, we get to a place where we're using these machines and the data we're feeding into them and the way that they're interacting with the world makes them get down to zero bias. Is it even possible for that to happen when they live in a world with biased humans

creating robots? Yeah? No. So I don't think that these systems can ever get to zero bias because there will always be a group that the system has not interacted with. It might be that it's perfect, as perfect is perfect, and then there's an unknown community in South Wales somewhere that had never interacted, right, and now it doesn't work with them, right. So I don't think that we can ever get because as humans, we're unique and we're different,

Like there's an attribute that we're going to miss. So I don't believe in zero bias. What I do believe, though, is that because these systems are based on data and they're not based on our lived experiences, it means that you can basically quiz them. You can look at them, right, So you can basically say, look, I have you know, this person who is from a rural community that is identifies as black. That's female, and I have the exact same person. But now we change from female to male.

You know what's the outcome? Right, Like, you can ask that and be like, oh wait, the differences are here, Okay, we have some bias. Let's go fixes. Now you ask human the same question, it'd be kind of hard to figure out if they're biased or not, just in general. And so that's what I'm saying is it's like you can quiz the AI, and because it's based on data and an algorithm, it can give you the answer you ask a human and if they know they're biased, they're

gonna lie. And if they don't know they're biased, they're just going to make up excuses because they don't realize that they are putting their own lived experience in their decision. So you would say the difference between robot bias and human bias is the ability to be I guess, radically transparent in the fact that we can open them up, look at their code and see what the answer is. Correct.

Although we don't do that now. And that's why a lot of this stuff is going on about AI and biases because companies that are putting these out are not opening it up. They're not doing the assessment, they're not doing the analysis. It's other researchers like myself that are you know, third party looking at and be like, hey, there's something going on here, right it shouldn't be our

role and our responsibile. But that's what we're doing for the community because the companies are not doing it for themselves. What do you think our listeners can do to kind of support this mission and improve the marketplace and kind of help more with learn more about AI and assistive technologies, but kind of also engage with it more. One thing that I think everyone has to figure out is how

to code and how to program. Honestly, the jobs are changing, and even the jobs you think are, you know, well, what about law, what about policing? What about Guess what? If you don't know how to code, there's going to be a whole next generation that's going to come take your job. It's going to be humans that take it.

But it's because they know how to do the new types of jobs that are going to be out there, So that's one thing everyone needs to learn how to program and the code, not be a computer scientist, but learn that as a skill set. Second is is I think people need to do a little bit more investigation of the things that are out there instead of just listening to media, because a lot of things that are

out are the horror stories. So what types of publications or media sources do you think that listeners can go to in order to get accurate information about STEM, about robotics, about science and technology generally. So I actually personally like Wired. It's a technology magazine, but it is really accessible to like a general audience, and it shows both sides. So it has stories that you talk about the negatives of AI, but it also in the same article we'll talk about

the positive. So I think in terms of being accessible, that's one of my favorite Doctor Howard, thank you so much for being with us. Thank you. Doctor Ayana Howard is a roboticist and dean of the College of Engineering at the Ohio State University. Be sure to check out our show notes to find out ways you can learn more about robotics and how to get involved. Next week, we're talking about vaccine passes and how they might help solve the problem of safe reentry and to some of

our busiest cities. Solvable Senior Producer is Jocelyn Frank, Research by David Jah, Booking by Lisa Dunn. Our managing producer is Sasha Matthias, and our executive producer is Mio Lobel. Special thanks to Heather Fame, John Schnars, Carl Migliori, Christina Sullivan, Eric Sandler, Maggie Taylor, Emily Rostick, Maya Kanik, and Kadijah Holland. Solvable is a production of Pushkin Industries. If you like

the show, please remember to share, rate, and review. It helps us find our way to the ears of new listeners. You can find Pushkin podcasts wherever you listen, including on the iHeartRadio app and Apple Podcasts. I'm Ronald Young Junior. Thanks for listening. M