Between boarding a flight to Dubai earlier this month to
deliver a talk and landing in Orlando in the United States a few days later for
the Orlando Keynotes, Tanmay Bakshi cheerily agreed to a transcontinental
interview. Tanmay is 14. And is no mean teen, pun unintended. For he is a
cognitive developer, honorary IBM cloud adviser, IBM Champion, author, neural
network researcher, appdeveloper, inspirational TEDx speaker, and has spoken at
IBM Watson summits around the world.
To cut back a little in time, which isn’t much in the first place, Tanmay first fell in love with computers at age 5 while observing his father Puneet Bakshi working on codes. At 7, Bakshi set up a YouTube channel, posting tutorials on coding and web development. At 9, he released his first iPhone app, which teaches multiplication. At 13, he began working with IBM on artificial intelligence alongside his mentor IBM Watson’s Chief Technology Officer Rob High.
Are we pushing at the stratosphere of cognition and computing?
So we began with the Second Question, not the First.
You are 14 and you began your fascinating journey 9 years ago. Besides your natural gift — prodigious at that, given that there is something extraordinary about your own DNA ‘coding’ if I may — what comes through in your talks and interviews is the idea of ‘passion’, not to mention ‘fun’. What would you valorise: ‘passion’ or ‘genius’?
I definitely say that I’m mostly passionate about technology. As you mentioned, I originally started coding when I was five years old, and at five, no one really knows the concept of a “job”. I just treated coding as another toy: not something that I’m paid to do, but something to play with when I have spare time. Because of this, coding grew on me as a passion and a hobby. This is why, at seven years old, I started using the Internet and books as learning resources. I learned Visual Basic, and created my very first real Windows app in the same year. From there, I learnt Objective-C and iOS development when I was 8. I had my very first iOS app, tTables, accepted into the Apple App Store at nine. However, from there, I started to lose my passion for technology. I started to think technology was something very repetitive: right as you code something in, it immediately starts becoming obsolete. But when I was 11, that all changed — I stumbled upon IBM Watson, more specifically, a documentary on how it played and won the Jeopardy game show in 2011. As you can imagine, for an 11-yearold, that would be very very fascinating. I was immediately hooked to the power of Artificial Intelligence (AI), and ever since I’ve been using the power of cognitive computing in my applications.
Today you straddle the world stage as a top coder, app developer, cognitive and neural network researcher, IBM cloud adviser, and inspirational speaker. And yes, you write books as well, having written Hello Swift. That’s at once formidable and intimidating. Tanmay, how does your mind function? It isn’t meant as a rhetorical question. How do you grasp, ingest, process, interpret, and isolate such disparate realms and networks of data and knowledge?
Well, I like to approach things from the bottom up. I like to start from the very very base of things, and build up from there. One example of a technology that I think wasn’t thought out completely, is robotics. A lot of people are very very interested in robotics — and I understand the excitement behind it. However, if you think about it, we’ve got a lot of robotics hardware — like, for example, the pepper and Nao robots. What we don’t have is the software that goes behind them. We don’t have the great artificial intelligence that goes behind the hardware, and allows the hardware to do wonderful things that we all need. But if you ask how I grasp, ingest, process, interpret, it really depends on what I’m doing. For example, if I’m working on a project that uses neural networks for facial recognition, I’ll start by taking a look at the work that others have done in the field. Because again, there’s no point to reinventing the wheel or rediscovering fire. If someone’s already done some work, I should take that work, and build upon it, to build a better system in general. And I also really believe that the only way to truly learn is via practice. If you’re trying to learn about neural networks, you can’t just read a book or take a course, and expect to be an expert! You need to have hands-on experience of actually using and training the neural networks and actually working in their backend.
While referring to Microsoft’s HoloLens in an interview you say: “Seeing the technology immersing with our real lives is really fascinating.” As I understand, your work, especially in neural networks and cognitive computing, you seek out that ‘immersion’ or intersection of technology and human lives. Can you, Tanmay, tell us a little more about how you visualise yourself pushing the frontiers in this area of intersection?
Exactly! This is the field I’m mainly interested in. Trying to bridge the collaboration gap between technology and humans, so we can bring out the best of both worlds. We all know that computers are immensely powerful in terms of mathematical power; we also know that humans are amazing and immensely powerful in terms of understanding abstract concepts and data, like emotion and natural language. Through technologies like cognitive computing, virtual reality and augmented reality, we are able to bring out the best in both worlds. For example, in the field of healthcare, we can almost completely eliminate the chance of human error, while at the same, using humans to eliminate almost all AI error. Of course, there is a small gap where mistakes will be made; it’s inevitable. But if you think about it, all the interactions we have with technology today are basically workarounds; if we want the computer do something, we use a button. If we want another human to do something, we speak in natural language. Through the power of cognitive computing, we can bring that human ability to computers, so we humans can utilise those computers to our full capability.
In view of your rather extraordinary preoccupations, you are being homeschooled. Allow me to interrogate you at some length in this area. What do you learn at home? And who teaches you, shall we say the Social Sciences, History, Mathematics, and so on?
My family (mom, dad, sister) home school me, I do study myself and research as well. This method of learning has numerous advantages, including the fact that while I still learn all regular school subjects (that’s still important), I’m still able to take whatever I learn and align it towards technology. This works especially well with math and science, because as you can well imagine, artificial intelligence is basically the science of biological neural networks, put into the math of artificial neural networks, put on very efficient algorithms running on technology.
Does the Social Sciences and History interest you?
All subjects really fascinate me, mainly math and science, and while social studies and history are definitely very interesting and fascinating, they aren’t my favourite subjects!
What does your father, Mr. Puneet Bakshi, teach you? He was your first Coding Guru.
Yes, you’re right! My dad was originally my “coding guru”. He has been working as a tutor, and since I’m now homeschooled, that becomes very helpful. We have resources like books, course material, curricula and my dad enjoys teaching.
What does Mother teach you? Obviously, a lot, we assume! For, learning is often more to do with what we learn outside our classrooms…
Mainly, she teaches me language and life skills, how to deal with people and situations, since what I do puts me in contact with people from all different demographics — and of course, manages my busy and hectic schedule, full of fun!
Creativity is of the essence in your work and research. Is there a method in the madness at the heart of creativity? How and where do you chance upon your line of enquiry?
As you mentioned, creativity has to be a huge part of the work that I do. See, neural networks and AI are more like an art than a science at this point. There is no “documentation” as to what to do when you want to build an AI for a specific use case. You have to, yourself, experiment with all the possibilities, which can take up to weeks. Over time, you do gather experience though; this adds a little bit of a method to the madness. It’s not very easy to distil all of that experience into a single answer without getting too technical, but I can say that it’s a skill more learnt than taught.
In your work you obviously go beyond theory or theoretical assumptions, say for example your considered view on AI. You have on numerous occasions allayed with confidence that AI can never ever take over human functions, or shall we say humanity itself. Where does science slip into fiction and vice-versa?
As I mentioned, since AI is currently more of an art than a science, it’s hard to definitively comment as to what the future of AI truly is. However, what I can say is that due to the fundamental way AI works and functions, some of the basic fears that people have with AI, like it’s going to replace us overpower us, are simply impossible to come true. There are some possibilities, like for example, HUMANS themselves may train artificial intelligence to do something negative; however, the AI will NOT, itself, learn how to do something negative, something that it wasn’t trained how to do.
The ‘Cognitive Story’ is a genuinely human story…
“The Cognitive Story” is a project that is aimed at augmenting people’s lives through the power of AI and Cognitive Computing. The first chapter in this project is about helping those with special needs and disabilities, and our first goal is to help Boo; she’s a quadriplegic girl who lives just north of Toronto, and suffers from Rett Syndrome. As mentioned, she’s quadriplegic, meaning she cannot communicate in any way. In order to help her, I’m trying to give her the ability to communicate artificially, by having my custom architected Neural Network system analyse and decipher her EEG (electroencephalogram) brainwaves.
Finally, to the First Question: Who is Tanmay Bakshi, beyond that prodigious global tech influencer?
It would be hard to describe who “Tanmay Bakshi” is, APART from the fact that I love technology and intelligent algorithms... What I can say, however, is that Tanmay Bakshi is just like any other 14 year old. My favourite sports would be biking, table tennis, and cricket; my favourite TV Show is SpongeBob Square pants; and my favourite movie series is Home Alone! I also like watching comedy movies.