Preparing the next generation of AI scientists, with Taniya Mishra

0:00Starting a business comes with its share of ups and downs, which is why staying true to your vision is essential. a non-negotiable for Romeo and Milka Bregali, Capital One business customers and co-owners of Ra's plant-based restaurant in New York. Romeo and Milka took a leap of faith when starting their own restaurant, gutting an empty space and building it from the ground up. Every pipe, every wall, every detail. But building from scratch came with a heavy financial burden, which is when they turned to their Capital One business card. With the flexibility of the card's no preset spending limit, they were able to spend more and earn more rewards while bringing their vision to life.

0:36Today, Raz's success is proof that with passion and the right support, it's possible to make your dreams a reality. Learn more at CapitalOne.com slash business cards. If you have kids, you probably know that schools are struggling to keep up with the pace of AI. I get it. There are a lot of factors to consider, including how to teach AI and how to integrate it into the classroom. Which is why I was so excited to talk to Dr. Tanya Mishra, who is the founder and CEO of SureStart, a company focused on preparing young people from middle school all the way to college for an AI-driven world. And in particular, they focus on training a more diverse cohort of leaders who can shape our AI future.

1:23Just listen to some of what their students are up to. I'm currently working with a couple of peers on a startup called Connects. which integrates AI through physical therapy platform, having student athletes actually able to perform their exercises at home at their own convenience. If I don't use AI, I will most likely fall behind. So I'm not going to do that. I'm going to use AI. I'm going to use it to enhance my abilities, not replace my abilities. AI is actually going to be foundational to my career. I study computer science and data science. I'm literally going into this working in the intersection between cybersecurity and machine learning as a job.

2:01I'm also using it for my startup, where we're actually using computer vision and speech analysis to detect signs of stroke on onset and get people to the hospital. You just heard from Anika Satchdeva, Lukash Pampati and Ashmeetha Kumar. Impressive, right? Today, Tanya and I are talking about why we need AI education now, the importance of a more diverse tech workforce, and why staying the course is a winning strategy. I'm Rana El-Khalyubi, and this is Pioneers of AI, a podcast taking you behind the scenes of the AI revolution.

2:47Hi, Tanya. Welcome to Pioneers of AI. It's so good to see you. It's been a while. It has, and thank you for having me on your brilliant podcast. So today we're going to talk about how we prepare the next generation of AI scientists and entrepreneurs. But before we dig into that, you and I have worked together for many years at my company, Affectiva. You are our director of research. And there's this one moment in our time together that, you know, has stuck with me all these years. It was one of Affectiva's annual family gatherings. And you and I, it was like a sunny day in August in Boston. And you and I were sitting on a bench and we're just chatting.

3:31And you said, you know, one day I want to be an entrepreneur and a founder like you. And your eyes just lit up. I literally like still remember this to this day. There was a sparkle in your eye. And I sat there and I was like, I got to help this amazing person start her company and start her entrepreneurial journey. So I am curious. Do you remember that moment? I remember even what preceded that because that was actually our annual summer picnic. But it was the last day I would see the summer interns because I used to live in New York City and work in Boston. and our intern program was a bit of my baby.

4:10So I had gone up in the morning, taken a three and a half hour Acela ride and I was gonna come home the same day. And as I was riding up, I was starting to form what would in the future become Sure Start, the sort of the seedling of the ideas of that around this idea of building our future AI leaders. and then you and I had that conversation about it. And I still remember, I said to you, what if we grew this into a really large part of Affectiva? And you said, Affectiva is great, but think bigger than Affectiva. Like, let's do some blue sky thinking. And that's when I said to you, I would love to, you know, have a startup, a company that was focused on building up the next generation of AI talent.

5:04And yeah, I remember that conversation. Our listeners can't see this, but I literally have goosebumps right now. It's such a powerful moment. It actually, it was powerful for you, but it was equally powerful for me because it made me realize that I just love unlocking potential in people and just making people believe, right? And that's sometimes all you need. So let's rewind though. So you and I came to the US as adults. and you were kind of just alluding to it, mentors have played a key part in your journey. So just tell us, you know, what brought you here? How has your journey unfolded? So I came to the U.S.

5:45as a sophomore college student. I had never seen a modern computer ever before. And I saw it like the second day I was in the United States. I was staying with a relative who took me to her workplace because, you know, she couldn't leave me home myself. I probably didn't know how to use any modern appliances. So that was a good thing. But she literally sat me down at a desktop next to her and brought up, if I remember, this old web browser called Netscape. And she said, you can type in anything you want and you'll get, you know, you can get information. And so first I started testing it. Like I'm from India.

6:27So I was like, where is the Taj Mahal, right? The canonical question. And I was like asking it state capitals. But my mind was blown that I could get information. And then I started, of course, asking it questions about physics and chemistry and like things that I was interested in. And what used to take me, frankly, you know, conversations with my adults or like my parents, my teachers, or if they didn't know, a hour and a half long journey to the bigger city close to where I lived in India and the bigger library there. And then I could get that information. I was just like access to information.

7:11A week later, go to university. and they set me up with my first email. I am now in love. I'm like, what? I can send a letter, I used to call emails letters, now to my parents and they can get it now. Nevermind the fact that they didn't have a computer for a very long time after that, but the possibility of access to information and connectivity with loved ones. It's like, this is magic. Whatever this is, I need to learn more about it. And that was when I met my first mentor, Professor Ann Smith. She was the head of the computer science department at my university, St. Mary's University of Winona, Minnesota.

7:56It was a small liberal arts college. And I must have talked her ear off about how amazing computers are. And she said, Tanya, why don't you take CS101? And I remember it being the moment of truth, feeling like I am the least prepared person at this university to take CS 101 because I didn't even know how to type. And I said that to her. I said, but I don't know how to type. And she said, that's OK. It's going to be hard, but you can learn how to type and you can take computer science 101 and learn about logic and algorithms. and coding. And that was, you said unlocking potential. And I think she did that for me that day because she took not a chance on my skills because I don't think I had any computing skills.

8:50I didn't know any computer science or even how to type, but she took a chance on my goals and dreams and excitement and frankly gave me the vocabulary for what my career would be made up of, right? Yeah, it's amazing because then you then went on to do a PhD in computer science. So tell us about that and how that led you to Affectiva. Yes, so I had a PhD in computer science with a focus on using machine learning to build educational tools for children on the autism spectrum so that they could learn how to have stronger social interactions, because that's often the most challenging part of, you know, being on the autism spectrum for young people.

9:44So that was really my PhD thesis. And I did it at Oregon Health Sciences University in Portland, Oregon. But that was where I also first started employing affective computing, because my PhD thesis was learning about how people express emotions through their voice, through their intonation, through their prosody, and then building models so that text-to-speech systems can be more expressive and more emotional because ultimately that is what we as children are able to learn and are able to pick up in our social interaction. So our goal was to build systems like that for children on the autism spectrum.

10:30We always kind of joke that we had like similar career paths because I did a lot of that but using computer vision and facial expressions and you were doing that work but analyzing voice. Actually that was how I found out about Affectiva, funnily, I went back and looked, you are actually cited in my PhD thesis. So while I didn't know you as well as I know you now, I knew about your work. Like I would follow your work and then Affectivas. And so when the time came for me to, you know, look for a new opportunity, do something different, I actually reached out to you over LinkedIn. I remember that.

11:09I had been following your work and I sent you an email and I still remember that email because it was such a moonshot. I didn't know you at all. And I think you responded in about like 10 minutes, which having known you and how busy you are, I think was a miracle. It was like meant to be. And then I like to say that the rest is history. We met in New York when you were visiting New York. And a few months later, I was working at Affectiva and really, you know, together we were bringing to fruition sort of the beginnings of our conversation around making affective computing more multimodal and applying it in many, many ways, all the way from impacting children on the autism spectrum, all the way to addressing well-being and safety in semi-autonomous.

12:02and autonomous vehicles. Yeah. So when you were at Affectiva, one of the things you and I were both passionate about at the time, and obviously you still are, I am too, we wanted to start this internship program. And we actually got a lot of pushback from our team because, you know, we had this vision of like bringing in high school interns and, you know, they would take a lot of hand-holding. But you persisted, we both did, and you championed the Empath program. One of the things that I'm really proud of was our team commitment to the diversity of the individuals we brought to that program. And if I recall, like 50 % were female.

12:42And perhaps you can share like maybe quickly the highlights of that program and some of our successes. I mean, the empath program at Affectiva, I see that as sort of the beta of what SureStart went on to be. It was 50 % female. We had, I believe, students not just from across the U.S., but we had students all the way from Egypt. And we had expanded the program to not just be for college students and for grad students, but there were high schoolers in it. And I think that was sort of, for me, the realization, the impact we can make on the AI ecosystem in making it more equitable and inclusive. It was by going younger, by creating space for really smart, engaged high school age students.

13:41and now at short start, even middle school age students, to learn about AI, to build with AI, to connect to others who are already working in AI, and then be able to hold the space for them to build what matters to them. That's where it goes from a technical concept to a real tool for making change in the world. So what does this real world change actually look like? After a short break, we'll hear about some of those success stories coming out of SureStart. Stay with us.

14:36If you've spent any time building AI products or leading technical teams, you know this. Transformation doesn't fail because of ideas. It fails because teams can't move together. Enter Atlassian's Teamwork Collection. It has planning in JIRA, documentation in Confluence, video updates in Loom, and now AI agents in Rovo, which connects the dots across your work so nothing gets lost. It's one AI-powered teamwork platform designed for how modern teams actually build. Learn more at atlassian.com slash teamchanger. That's A-T-L-A-S-S-I-A-N dot com slash teamchanger.

15:25I wanted to hear some real examples of SureStart students making an impact on the world. And there was someone who quickly came to mind for Tanya. So I want to share the story of Erica Corral. When I met Erica in 2021, they were in the first summer program that we collaborate with MIT Rays on called MIT Future Makers. Erica, by their own admission, was a quiet, reserved young person. They had just started college. And now, after having done that one program, but also built those networks, they are now pursuing a PhD in computer science at the University of Denver. They are a mentor in our program, mentoring dozens of kids.

16:16They are also a part of the SureStart team helping develop content around AI that is inquiry-led, project-based, but most importantly, co-created with young students themselves. And, I mean, that's such a success story. It's like creating a virtuous cycle because with Sure Start, my hope is to reach, you know, thousands, tens of thousands, maybe one day millions of kids. But in order to be able to do that, we use a decentralized model. When we train up a young person or really any person, hopefully teachers as well, in AI skills, our hope is that they're taking it back to their communities. They are teaching other people with it.

17:03And AI is a disruptive technology, but its dissemination has also got to be disruptive. And it has to be community-based. It has to be collaborative. I love it. What are some examples of the AI skills that SureStart kind of trains these people on? Yes. So we have AI courses for middle school students, like starting in sixth grade. What does that look like? Oh, my goodness. Middle school students, it's been so much fun to create these programs. So we have courses that are just 25 hours long. We call it AI Explorations. And that's a really fun one where we leverage students' natural creativity in art, in music, in storytelling, even journaling.

17:53I know you are a big journaler. In poetry and say, OK, we are going to use this natural creativity to introduce you to a tool that can amplify your creativity. It's not going to replace your creativity. It's going to create efficiency. It's going to become a co-creator with you, advise you. But through that, we're teaching them what is generative AI because, you know, everyone's talking about it, but also what is discriminative AI? And additionally, we start to build their ethical AI lens. I love that. What are the important questions around ethics you need to ask? So in a creativity course, the first question that kids naturally ask is, so who is the author of this piece that I wrote?

18:45Am I the author? Is AI the author? Or is the author all of those amazing authors on whose work this AI has been trained? I mean, these are questions that none of us actually have answers to, but how amazing that 12-year-olds can ask this question. We also have AI foundation courses. We have some partner schools in Asia, specifically Singapore, who are very tech forward. Now, you were just there, right? You just had a trip there. Tell us more about that. Singapore is, they're very tech forward as a community, but also the Singaporean government is very tech forward. As soon as the sort of the chat GPT led revolution happened, within a year of that, we were hearing that, you know, schools are quite interested in introducing AI to their students.

19:41And so we got connected to a school called Excel World Academy in Singapore. And they have a powerhouse CEO and also head of school, Tammy Murphy. So we have a one-year, two-semester course. These kids are learning about AI every single day, five days a week, first hour of their day. So again, what are we teaching them? We're really teaching them what is AI? What is discriminative AI? But also, how do you build with AI? How do you build with AI in a way that impacts from your local community all the way to your global community? So what we do is we have them align their projects to UN sustainable development goals.

20:32and there through it we also teach them interesting skills like not just problem solving but problem identification. Once they've identified those problems then we give them an introduction to entrepreneurship because entrepreneurship is a very real career path for AI or really any disruptive technology and what is coming out the projects that come out at the end of it are mind-blowing. An example one was an app to help students with disabilities navigate, you know, cities worldwide. They're starting with New York City. So what if there was a map that told you where all the accessible subways were, where all of the accessible restrooms were, the buildings.

21:24And it came directly from our students with 12 and 13-year-olds who are thinking about it. And so you can easily imagine the projects that our high school students are doing or our college students, because these projects are so powerful that they are not staying projects. They are then becoming startups. Yeah, so cool. So Stanford just released their AI index report for this year, and they reported 81 % of K-12 computer science teachers say that AI should be part of the foundational kind of program at the school, but only half feel equipped to teach it. You and I both have younger kids. I am curious what you're seeing in the U.S.

22:10Like, are people leaning into embracing AI in the classroom? I can, you know, I sit on the board of our school and we're definitely trying to navigate that. Like we're trying to lean in, but there's pushback, obviously. Yeah. What are you seeing? So I'm seeing, it feels like a mixed bag, right? So teachers are rightfully concerned that how will knowledge acquisition happen if then there is a tool that gives you all the answers, right? So I hear the tension and then And I think we need to address it. And I think AI is absolutely able to address it. But AI cannot just be built by technologists. It needs to be built with domain experts playing as much of a role, if not more of a role, in the development of these technologies.

23:02So I think I'm seeing certain schools that are very bullish on technology. and they are like, yes, like our partner schools in Singapore. So we are seeing some of that here in the US as well. But we are also seeing schools that are very interested, but they are almost like we don't know where to start. So where, you know, more teacher training needs to be done, not just around workflow, but the other kind of training, either how to teach AI, because it's a whole new curricula, but also how to teach with AI. So if you're an English teacher, because we get these questions all the time, how do I use AI in my classroom that is able to bridge that tension between learning how to write well and also use AI as a co-creator, right?

23:59So I think we are seeing that. But then we are also seeing something else, which is kind of probably more prevalent, which is currently curriculum in the classroom already is quite wide. So then we get this question. They're like, we're interested in bringing AI to the classroom, but what will I need to drop in order to pick up AI? Because nobody, like no governmental standards or state standards require AI. So we're seeing the full spectrum And this goes back to the thing that, you know, you have and I have talked throughout this conversation, which is it requires collaboration from technologists, educators, but also, you know, we need legislation and the government involved.

24:50All three stakeholders need to come together in order to say, how do we begin the education of young minds starting today so that our future leaders have skills, have confidence, have self-efficacy and can make great decisions with tools like AI. SureStart is intentional about their outreach. They particularly want to bring in students across the diversity spectrum. We get to why after a short break.

25:40Meet Nicole Nicholas, Capital One business customer and co-owner of Ansett Uncles, a plant-based restaurant and community space in Brooklyn, New York, that got its start from a need for unity. The inspiration, it was born from the desire to create a space that felt like home, where we can connect community culture, good food, and come together with family and friends. That's how we birthed aunts and uncles. Nicole and her husband, Mike, were fulfilling their dream of bringing people together out of their home kitchen. But they soon learned that the demand for community was greater than they knew.

26:11It became overwhelming and we were like, we need home, but not in our actual home. We realized that there was also a need in our community for something bigger in our neighborhood. So we had to find a place. Moving from a home operation into a storefront was a huge next step, but Nicole and Mike were able to take it on with the help of Capital One Business. It's not for the weak as a small business. Finding resources is super important because that's the way you'll be able to manage and scale. We would have never done that without having Capital One to be able to help us along the way. The cashback rewards are very helpful.

26:47You know, it just gave us that runway to be able to breathe a little bit. then you get to focus on the cooking of the food and making the experience great. To learn more, go to CapitalOne.com slash business cards.

27:04Okay, so let's talk about diversity. It's something that you and I both are very passionate about. So Sure Start is very intentional in supporting students with diverse backgrounds in every sense of the word, right? Like diversity of gender, age, ethnicity, race, lived experiences, socioeconomic too. But the current AI landscape, as you and I know, is not that diverse. So I'll just put some numbers behind it. According to the U.S. Bureau of Labor Statistics, women make up only 26 % of the computer science and math-focused workforce in 2024. Only 9.6 % of that workforce is Black or African-American, and 8.3 % are Hispanic or Latino.

27:45know. I mean, these numbers don't represent what the U.S. actually looks like. And I think it's kind of concerning and depressing for a whole host of reasons. So let's unpack that. One is that if we, you know, AI is going to change the way we work, the way we, you know, it's going to change every aspect of our lives. And if it's designed by a very small subset of human beings, it's not going to work for the rest of us. What's your take on why it matters to ensure that AI is equitable and inclusive? I think all of us are asking, irrespective of how it's being applied, people are asking, how do we make sure that AI is not biased?

28:24How do we make sure that the AI landscape is more equitable? And I think the work that we are doing at SureStart is actually almost foundational. We have to bring more people to those decision-making tables where AI is being brainstormed on, designed, developed, used, deployed, right? And we need to start early. We need to start by bringing AI knowledge, AI literacy, AI informedness much, much earlier to kids. And again, it needs to be a public-private partnership. You know, the way we implement Sure Starts programs are great. We have amazing tech forward partners. But really what we need to do is bring public private partnerships so that these programs like ours are going into every public school that is out there.

29:24We are actually supporting them. We are scaffolding the teachers. We are pre-skilling teachers to upskill their students. We have to get this education into the hands of every kid from every background, every lived experience. If we put the work in today, then in a decade, the landscape is going to look really different. And, you know, how we can ensure that our AI is less biased? If we ensure that the people building it represent the richness and the diversity of the world we live in. So you guys mentioned DEI several times on your website. And in this moment in time, it's kind of almost like a statement, right?

30:06There's a national rollback on DEI efforts. Companies like Google have said they will no longer set hiring targets for people from underrepresented backgrounds. How has this moment affected your company? And has there been a ripple effect on what you do? I teach my kids to raise their hand. I say, leaders, don't fly below the radar. and I want to live. I want to be that representation for my students. So no, we continue to do the work that we do. Inclusion is good for everybody. I think we are lucky in that we work with people who are visionary and people who are not swayed by, I would say, momentary trends, interests, or even fears, we stay the course.