AI and the Future of Learning with Padma Bhamidipati

Kate Bevan: Hello and welcome to this episode of the Infosys Knowledge Institute's podcast on enterprise AI, the AI Interrogator. I'm Kate Bevan of the Knowledge Institute, and my guest today is Padma Bhamidipati. She's an Associate Vice President for learning at Infosys. Padma, thank you so much for joining us. It's really great to have you.

Padma Bhamidipati: Hi, Kate. What an absolute privilege and an honor to be here with you in conversation.

Kate Bevan: The honor is all ours. We had a really lovely chat the other day, and I really wanted to ask you how you came to be in the role you are now, and then we can get onto how you're using AI in that.

Padma Bhamidipati: Typically, when you think of daughters and dads, one would think that, in dads, you have the greatest champion. Well, in my case, my dad was really a person who doubted just about everything that I had to offer. And my mom, God bless her, she passed away last year, but she was the only one who believed in me. So really, I never had anything going as per the norm. And I think that's something that should have told me nothing what I do is going to fall into the traditional buckets. Which then leads me to me and Infosys. And the reason why I start with my dad, because when we got the email saying I was invited for an interview, my father, with this very deadpan expression said, "It's spam. I think they got something wrong." And here my mother was rooting and I think she just ran around telling everyone in her school and neighbors, "Oh, she's got it," and all of that.

I wasn't sure which parent to believe, but then eventually both of them did their best. And that's how I think, accidentally, I landed into Infosys. But on hindsight, Kate, I think it was all an interesting journey just as in Alice in Wonderland and her wonderful stories because I think I fell into this constructive rabbit hole, if you will. The rabbit hole or this entire ecosystem of Infosys was so kind, that it said, "Hey, guess what? Stay here. Stay here because you are somebody who's going to make it a little bit more interesting for us." And they had more faith in me than, for sure, my dad had. And then that's where I am 23 and a half years and every day expressing a whole lot of gratitude to this organization.

Kate Bevan: And here you are also very much leading on how we use AI in learning, in training, not just in training each other, but also in how we can use AI in education more generally. How did that come about?

Padma Bhamidipati: I come from a non-technology background. Though I don't think I can say that anymore, but I love to say it because that makes a great conversation starter. You see, when I walked into Infosys, my role was, true to what my profession at that time was, which is marketing, branding, and engaging with people. And that's really what I continue to do. I'm engaging with the learner community. And over a period of time, I've had the opportunity to take our digital learning ecosystem which was Infosys Wingspan, both to Infosys as Lex to our clients as Infosys Wingspan and to the communities as Infosys Springboard. And for someone who was not sure, when our computers folks said, "Launch the application." And I stared at him and I said, "How can you launch something without a marketing plan?" He clearly thought I smoked something and walked inside the campus, which was not permitted.

So these 23 and a half years have gone in understanding the lingo, the entire ecosystem, and appreciating what technology is. And here I am today appreciating how both AI and GenAI is really doing a lot to learning. And particularly, my role here is one where I've had the opportunity to work with Rafi's team in the AI first live enterprise effort to look at different tracks who want to engage in change management and how do I take the Lex and its AI features to our own employees. So that's been a interesting journey for a little over a year and a half, particularly in the GenAI space.

Kate Bevan: How does generative AI and actually AI more generally mean algorithms machine learning? How does that work in learning, in a learning platform?

Padma Bhamidipati: Well, let me possibly step back a bit. When we, as Infosys, look at learning, we've come through different stages. The very first one, we love our campuses. We can't stop talking and reading about it. So when everyone was congregating in campuses, we brought to life Mysore JEC, 15,000, 18,000 concurrent training, et cetera, a multiplex and what you have. From there, we moved to the next phase. When everyone was going digital, we brought in Lex. We were amongst the first app to be a cloud first, mobile first. And when you look at Lex, Lex is our digital learning ecosystem, which is a combination of a learning experience platform and a learning management system as well.

We perfected this in three months with 13 people led by, of course, the vision of Nandan, but the leadership of Thirumala Arohi and Shyam Prasad Geyar, the chief architect. And we first went live with the CSG or the sales community, the client servicing community, because Lex, if it was supposed to be or what is also known as Infosys Wingspan, is supposed to be an anytime anywhere learning platform. We needed to test it with those who were not strapped to laptops. And they really were our north star in more than one ways, and we needed to make sure that they gave us the inputs. And if this worked for them, this would work for anyone else. So that's really how the entire learning ecosystem came together.

It has evolved over a period of time. So when we were in campuses, we had our lovely campus classrooms, etc. Then came in this digital learning ecosystem. And then came the phase where everyone went berserk over Metaverse. And coincidentally, for us, the unfortunate pandemic really disrupt any of our learning efforts because we were already two and a half years into really having a hybrid way of learning, if you will, where there was digital learning, self-learning, instructor-led learning.

And then Metaverse came in at a time where we were able to give an immersive experience. And now we are literally in what we call as learning 4.0, where it's a human plus AI experience because GenAI seems to be literally that magic silver bullet, but personally I know it isn't. And that's really where the human elements are so far nuanced and so much critical now, which I know is also going to be the theme of our conversation. So that's how learning has evolved. We stayed true to technology, but I think what we've always been staying true to is what the learner wants. The learner is at the heart of this entire journey for us. And we really want to meet them where they prefer to really learn and meet us back.

Kate Bevan: How does that work for taking it out, to say, to clients, to the wider community? Where's the role for AI in that?

Padma Bhamidipati: Like I said, we are in this phase of human plus AI and we at ETA believe that this is something that's going to help us in hyper-personalization of learning at scale. Now, at Infosys, our numbers are in a few hundred thousand. That's normal. We look at tens of thousands for pilot or POC purposes, but an impact is always at scale. Now, when I look at what GenAI has been able to do, I think it has been able to bring in a fair bit of that personalization, which would really make it so important for an individual to relate with the piece of content.

Kate Bevan: What do you mean by personalization? Tell me the details.

Padma Bhamidipati: Kate, possibly you would be somebody who would love a Socrative way of really engaging with. And I'm not an astrologer. Your personality really comes across in that manner. Now I, true to my Zodiac sign as well, I'm a dreamer. I have nothing to do with reality and throw stories at me. I love that. So if I had to attack a topic, possibly Java, let's also remember, we come from a technology community.

So if I had to learn Java, I would prefer and I would stay engaged if somebody started off by saying, "Once upon a time there was this Java and it's like coffee and everything," and you'll have me engaged. Now, I would not be able to expect this kind of a personalization. This is one of the many. In terms of the pedagogical style without having a technology intervention or this layer below what I'm actually engaging with. But then it doesn't stop with that.

Having received a lot more information and as I gain confidence, for example, I can switch my pedagogical method. Now in a classroom, this is extremely intuitive because you have an educator, they're looking at you, you raise your hand, or if you're sleeping, they just wake you up and then get you on board with the rest of the class. Now, when you are doing this in a self-learning manner, you need to have the choices. It's like you're setting up your classroom for yourself. You've brought in the educator, you've brought in the virtual tutor, you're now good to go. And if you are stuck anywhere in between, then you have multiple choices.

And the reason I say stuck is sometimes even repeated explanations up, you really don't do the trick for you. Now, this is where GenAI comes in with language options or translation options. Now, the reason why I love that aspect is having my personal background is I am a native where I speak one language as my mother tongue at home. I've grown up in another part of this country, so I know that native language as well. My husband, by nature, speaks another. We are a conglomeration of at least five different Indian languages. So if I needed to understand something, which personally for me, English was not doing the trick, I would rely on that native language of mine to help me. Now, that is true for everyone across board. Any learner would need to have that comfort option. In India and a South Indian would always say, curd rice, that comfort food and this translation is possibly the same.

So when we go in and I'm not able to understand a word or there's a phrase, I have the option of using this translation. And with an emphasis, we've seen a lot of European languages, Hindi, even Arabic coming in because that allows an individual. Why English is a corporate language, learning is personal. And these kind of personalization elements really help us without disrupting the overall, if I could say, learning economy that is there. So this just augments the learning experience for an individual. And staying on personalization, it also allows us to track telemetry. We at Infosys, in the context of learning, have something called a heartbeat telemetry. We don't mean to pry into somebody's learning life, but having said that, if somebody is stuck or has abandoned a piece of learning or is really not able to scale through a certain technology playground, that really is a nudge and a signal for us that possibly we can reach out to them, with all the data privacy in place, and help them really crossover that little bit of a tricky situation that they're in.

So GenAI helps in terms of pedagogical style translation. In fact, even translating documents. And not to mention for those who are creating content, we've seen this at Infosys, where you would have spent months waiting to create a course, you now are able to leverage this technology to piece that entire framework, stitch the whole literally the body together. And then the SME or the educator comes in to breathe life into it by making the appropriate changes, bringing in the relevant case studies. And what used to take four months, possibly in less than four days, you have a brand new course available. So it's something that is really making things a lot easier, provided we know how to tame this brand new shrew.

Kate Bevan: The one question I want to ask about that, if I'm a learner and I'm maybe trying to learn Python or Java, does it adjust itself on the fly to my learning style, or do I have to give it lots of prompts? How does that work?

Padma Bhamidipati: So you have a virtual tutor, which is now rolled out to everyone in the Infosys ecosystem. So every employee has access to this virtual tutor. Now, the best thing about this is it has deep, deep loads of patience. It really never runs out of patience of course. And that is important because you are not afraid. You are not being judged by anyone. So you start off, number one, by looking at, if I would say, we've spoken about translation, we've spoken about the pedagogical style summary. Sometimes just getting a quick summary of what this page is all about helps. So we have a feature which will allow you to summarize and get access to it.

The other thing that also comes into play over here is the equivalent of flashcards or flashcards themselves. Something I know a lot of people would really think, that's how kindergarten education is done, but hey, what works for learning works across age groups. And in addition to that glossary, what we've been trying to do is really get inside the shoes of multiple learners because I've always believed that no two learners are the same. And you need to throw the choice at them so that they don't feel stuck. Now in all of this, I can keep asking the virtual tutor. It will give me a response, it'll give me a summary, it'll give me ways by which, and in all of this, I can even keep quizzing myself. So that allows me to test my own knowledge whether I've been able to grasp. And when everything fails and that's where the human plus AI comes in, I can always reach out to two sets or possibly three sets of individuals.

Kate Bevan: Do you think things like the amazing multimodal models that are coming out, you know the GPT models where you can do music, you can do spoken language, you can do video, is that going to be part of how we do AI learning going forward?

Padma Bhamidipati: Possibly because, Kate, you mentioned music, I couldn't help but take a bit of a digression into. AI and the music industry has been doing such amazing stuff, and more recently, and those who are tuning in from India and are fans of the Tamil music industry, and particularly one music director, Anirudh. More recently, and this has happened, I've seen a couple of the very notable music directors leverage AI to bring to life the vocals of a very well-known singer who passed away 27 years ago.

But when due rights with due permissions from all their stakeholders, they've been able to recreate the voice. And trust me, Kate, that's a banger of a music any which ways. And for those who are keen, it's a Rajnikanth. It's filmed on him, so it's already topping the charts. But the point I wanted to make is, we are seeing a lot of things which otherwise seemed lost to the cause or lost forever come to life. And I think therefore it becomes a lot easy for us to not really limit ourselves. And GenAI really helps us in that kind of a thinking. It's possibly removing all the shackles and instead of saying, "No, we can't," it's saying, "What if."

I think that's just the way it has really made the change. And for example, we at Infosys ETA have created this very interesting piece of Rover, which we've called it as Aero Zoiee because Zoiee is the name of our virtual tutor as well. Now what this little thing does is it's actually navigating, it's moving. And it's actually, can I say, configured to help you right from helping you to find directions. Two, if you were really stuck with a question that you couldn't and you possibly didn't even have an educator in front of you, but you were on campus, you would ask it a couple of questions to say, "Can you please help solve this for me in Linux?" It is powered into give you those responses. So really, I don't think there is even a sky as a limit these days, and that's what GenAI is really challenging us to do.

Kate Bevan: You are incredibly optimistic about this, and I actually love your enthusiasm for it. I'm going to ask you my final question now, which I don't want to be a downer, but do you think AI is going to kill us?

Padma Bhamidipati: AI will kill us if we don't know how to tame it. We all know that we need to tread with caution. So I think possibly, now I would call upon having a middle path. Again, invoke Buddha over here, if you will, and say, "Let's not get too depressed. Let's not get too excited." But if there is a way that we can have a very objective, my preferred word for life nowadays is stoic. If you have a stoic view towards it, I think you will be able to really push back, have a neutral objective, and then make it work for you. And also get yourself to see how you can partner and do well with it.

Kate Bevan: That's a great place to leave it, that optimistic but realistic viewpoint. And Padma, thank you very much. It's been a real pleasure having you today.

Padma Bhamidipati: Thank you, Kate.

Kate Bevan: The AI Interrogator is an Infosys Knowledge Institute production in collaboration with Infosys Topaz. Be sure to follow us wherever you get your podcasts. And visit us on infosys.com/iki. The podcast was produced by Yulia De Bari and Christine Calhoun. Dode Bigley is our audio engineer. I'm Kate Bevan of the Infosys Knowledge Institute. Keep learning, keep sharing.