AI for Business
Mr Gopi Kallayil, Chief Business Strategist-AI, Google, sheds light on the impact of AI and how its growth is impacted by the triumvirate of consumers, communities, and companies. Mr K Mahalingam, President, MMA, led the conversation.
Mr Mahalingam: We both joined REC, Trichy (Now NIT) in 1980. Can you share a little bit about your days, prior to joining REC?
Mr Kallayil: I come from a fairly modest background. Three generations ago, my grandparents were rice farmers in a small Palakkad village called Chittalancheri. We were four children. Even though, my parents had no prior college education background, we went on to get ten advanced degrees between the four of us, including two US Ivy League MBAs. Getting into REC was a breakthrough moment for me. I was the first in our extended family to get into any kind of professional degree. My parents could not even guide me. They didn’t know the difference between mechanical and electrical engineering. We had very little aspirations, not because we lacked ambition, but we didn’t have the environment or exposure. REC Trichy gave me access and exposure to classmates who had more exposure, more knowledge about the world and more sophistication than I had. It opened up the world through the aperture of my classmates. It also gave me a level of confidence that I did not have, graduating from a high school in Kozhikode.
Mr Mahalingam: Considering that half of our batch went to the US immediately after college, you went to Uptron. From there, you joined IIM, Calcutta. How did that happen?
Mr Kallayil: After I got into engineering, I realised there was one pathway where I could go for higher studies, possibly in the US. I had zero awareness of it and no one in my family, or anyone in my network had ever done it. I learned the process from my classmates. I just copied and mirrored them. One of the key emotional skills that I used was that if I didn’t know something, I would ask people who knew about it and learn from them. I did it constantly, like a sponge. I asked a lot of my senior college mates and they told me about the process.
I set that as a goal and applied for a Master’s in Computer Engineering to ten US universities. I got accepted in all ten. But there was a financial problem. My family couldn’t afford the tuition fee or even the airfare. So, I took up a job at Uptron. After joining there, I thought I could pursue MBA in India, applied to B schools and joined IIM Calcutta and when I joined, you were already there, once again!
Mr Mahalingam: From IIM, you went to the Far East for some time and then to the US?
Mr Kallayil: At IIM, my classmates had an extraordinary amount of intelligence and sophistication that I was not used to. In fact, I had a tremendous impostor-syndrome. But again, I learned from them, copied their ways and asked them for advice. We would also occasionally meet people who had graduated from IIM, five or ten years ago and who were doing extraordinary things in corporate India. I looked up to them and told myself that if they could do it, I could also reset my aspirations and come out higher. When I joined MBA, my extended family members questioned my parents as to why they let me study after quitting a job. Luckily, my parents encouraged me and gave me the freedom.
After nine years of doing my IIM MBA, I built my financial capability and applied for my second MBA at the Wharton School of Business, University of Pennsylvania. Once again, there was resistance from the family circle. But I didn’t bother. After Wharton, I joined McKinsey to do strategy consulting and moved to their Silicon Valley office, because I wanted to work for technology clients. I stayed focused and consistent on one thing—to have technology clients. Soon, I quit and did a couple of startups.
Mr Mahalingam: How long did your stint in startups last?
Mr Kallayil: I did two startups in six years. Both the startup companies were bought or acquired. Then I took some time off to travel, as I went through a life crisis. Many things in my life fell apart suddenly in an otherwise well-functioning life. I had let go of my last startup job, where I was VP of a Tech company, reporting to the CEO and it was very well funded. Nobody’s life is linear and it always goes up and down. I took six months off and travelled around the world with one simple rule that I would not stay in a hotel and that I would only stay with someone. I went to Iceland, India, Dubai, Kenya, Tanzania and Zambia. I went to Mount Kilimanjaro. I came back to India and spent some time in a lot of ashrams and spiritual settings, where accommodation was cheap or free and the food was not great.
During this period, sitting in a retreat centre in Bangalore, I had a vision of what I wanted to do next. One lesson I learned was, “If you’re making a career change, don’t come up with a list of 100 companies. Come up with a list of one, two, or three.” I put down a list of three companies. Google was at the top of the list. I was intensely focused.
Mr Mahalingam: You’ve had some very interesting roles as chief brand evangelist and now Chief of Business for AI, etc. How did that play out?
Mr Kallayil: When I joined Google 18 years ago, it was a very small company. It had just two products—the search engine and the little text ads that ran next to it. There was no YouTube, no Chrome, no Android and no self-driving car. Now it’s a much different company. But I had this intuition based on my research that this company would change the world. The founders defined their mission in just 10 words: “Organize the world’s information, make it universally accessible and useful.” It was simple and profound. The company had a great culture. We were experimenting, launching products, trying things and failing. There was something going on and I was determined to work there, no matter what my job was.
Even the founders would not have known the direction the company would take. Part of the reason is that the computing power did not even exist then. Prior to 2006, nobody walked around with mobile devices in their pockets. If somebody had told me or people at Google, that one day humans would walk around with the map of the entire world on their phone, we would have said, “No way. That’s not going to work.”
Mr Mahalingam: You’re such a multifaceted personality. You’re a yoga guru at Google. You spoke at world speaking championships five times. You are a Kirtan singer and you were considered for a nomination to the Grammy twice.
Mr Kallayil: Just to put that in context, when I was at NIT, I went and lived in Shivananda Ashram and became a yoga teacher. Back then, yoga was not so popular and cool. My classmates ridiculed me. Now people around the world and celebrities are doing it. When I got to Google, I established a yoga program for Googlers, called Yoglers. It has become the largest yoga program in the world in a corporate setting. We have 150 classes a week.
I have written two books, given five TED talks, competed in the World Championship of Public Speaking and ended up five times in the semi-finals. This year also, I was in the competition in Bahamas and finished 15th in the world. Along the way, music happened. I fell in love with the sound of Namasankeerthan. They started recording it and one thing led to another. In the Grammy, they’ve created a new category called New Age Chant and Ambient music. My work was considered in that category. The music is actually very old age but interestingly, they put it in the new age category!
As I mentioned, in 2005 when I had a big crisis, I was thinking about things a lot. I was reading a lot. One of the books I read was about ‘Saying Yes to Life,’ in which, the author talks about how she experimented for one year with just saying yes to whatever showed up, trusting it and moving towards it without self-doubting. I took a similar approach. From that point onwards, I decided to say yes to a lot of things. That year, I went to the Burning Man festival. It’s a counterculture festival in the Nevada desert, where people practice radical self-reliance, radical self-expression, radical inclusiveness and so on. Now I go there every year and this is my 17th year.
There’s a core principle of Burning Man that I really liked and I made that a part of my life. This is my summarisation of life. “All of our life… is one giant experiment.” In our life, we conduct lots of experiments—in our jobs, with our money, with our marriage, with the way we raise kids, with the way we practice our faith… We try to find out what sticks and what works. Experiments, by definition, lead to unexpected results. But we will never find out the results till we try. So, one decision I made was that I would experiment and experiment very big. I know I will fail most of the time, but I will learn something. That’s what led to experience all these.
Mr Mahalingam: The kind of things that you do is so engrossing. Gopi is also a triathlete. He is a yoga guru, a kirtan singer, TED talker, Toastmasters’ semi-finalist, a triathlete; and by the way, he is also the Chief Business Strategist -AI for Google!
Mr Kallayil: We have witnessed in our lifetime, three big platform shifts. We saw computers showing up on our desktops and then in our homes. Till 90s, these computers did not talk to each other, even though internet had been around. The invention of browser is the first big shift that allowed ordinary people to get on to the internet, find information, book tickets, buy milk or talk to friends, etc. At the University of Illinois Urbana Champaign, Marc Andreessen and Eric Bina created the Mosaic browser, which became the Netscape. Then Microsoft created the Internet Explorer, then came the Firefox, Apple Safari, and many years later Google Chrome. Sundar Pichai led that project, which led to him being recognized and creating enormous success for the company.
Then came the second wave. In 2007, Apple created the iPhone and a year later, Android was launched. Then the mobile phone became the most deployed piece of technology in human history. We now have more mobile devices than we have humans on this planet. It has changed the way we deal with commerce, communication and entertainment. We and our businesses adapted the mobile.
Now we are watching the third biggest shift, which is AI. Starting 2022, what we witness is ordinary consumers engaging with AI. We believe AI is the biggest technological shift humans have ever seen. I would put AI on par with the kind of innovations humans come up with, once in 100 years, like 1462 when Joseph Gutenberg created the printing press. That changed the way we disseminate and consume information. In the same way, in 1752, when electricity was discovered or in the 1890s or so, when the automobiles were invented, we saw massive change happening.
Tweak the Question
When I meet CEOs of large corporations, a question that I often get is: What should my AI strategy be? I tell them, “Strike off that question. Tell me what is your strategy. Then ask, how can I enable AI with my strategy?” Another way of approaching this question is asking, “What is the biggest opportunity for my business in 2024 that AI can enable?” There’s also fear and apprehension about AI and it is very common. The economist ran a cover story in May of this year on AI. It carried a picture with a lovely design. There was a ring of halo above ‘A’ and devil’s horn and tail with the letter ‘I.’ This, in essence captures what most people feel about AI. Let me demystify some of these fears.
Natural Intelligence vs. AI
To understand AI, we must first understand what natural intelligence is. There are so many human concepts that came from our natural intelligence that have gone into various products and services. Everything we see around us is the sum total of enormous amount of natural intelligence. Generations of people have worked on it -from metal to lighting to electricity to computers and smartphones. But this capability with natural intelligence is limited. If 15 people are on a leadership team of a company and they take a break or a vacation, the work will suffer.
So then, how do we amplify natural intelligence? This is what the computer science fraternity has been doing with artificial intelligence. At Google DeepMind, we define artificial intelligence as any computer system that captures natural intelligence and mimics it, and once you mimic it, then you can amplify it. If you look at human locomotion, we first learn to crawl, then to walk and run. There’s a magic moment when someone teaches you to ride a bicycle- the simple system with two wheels, a gear system, and the handlebar. That allows you to go further and faster than you’ve ever gone before. You use the legs that you use for running, but differently. It has amplified your locomotion. That is the same concept that we are trying to bring with artificial intelligence.
Let me give you two examples. First is a point solution, which is translation. We have AI based systems to translate languages, like Google Translate. We use machine learning to create the capability where you can speak in one language and get it translated to hundreds of languages. Another example is complex systems like Med-PaLM, which is used by healthcare professionals. It can amplify a doctor’s skill. It is not going to replace the doctor. On the contrary, it will allow us to bring health care to populations that would not have access to it.
Around the world, we need a single ophthalmologist for 20,000 people. In some parts of India, there might be only one ophthalmologist for 200,000 people. Therefore, many people don’t get proper eye care. In the future, because of systems like Med-PaLM, someone sitting in a rural place could take a picture of the eye on a mobile and an AI system will detect and say, ‘You are susceptible to diabetic retinopathy and you need to quickly see a specialist in your area.’ This is how we can scale and deliver health care to a lot of the world’s population that doesn’t have access to specialists.
Right Brain and Left Brain
As AI mimics the human brain, it mimics both sides of a brain. Right-brain thinkers, we say are good at creative thinking; and left-brain thinkers at being logical and analytical. Similarly, AI system that mimics the right side of a brain is called generative AI and the left side of a brain as analytical AI. Sometimes, they merge and come together.
Four Capabilities for Business
With AI, we now have four capabilities available to us in the world of business.
1) It can synthesize vast amount of data and analyse them.
2) It can drive predictions based on that model.
3) It can create new content, art, music, etc.
4) It has conversational capabilities, like ChatGPT, about which, the world has got excited.
Large Language Models
Let’s start with traditional programming. If we want to teach someone to drive, we teach him step by step how to drive, how to look out for and interpret the signals, how to follow the rules, etc. To teach a computer system how to drive a car, for every possible outcome, you must write a line or many lines of code. From there, we evolved into a concept called neural networks. Then came the concept of large language model. The computer is fed with lots of information. It synthesizes everything that is fed and it can even come up with a new theory that humans did not even know about. This is how large language models work, like Gemini from Google, or ChatGPT from open AI. It can draw a picture based on dictation of a scenario. It can summarise YouTube videos and generate lists from that video. It can create music. It can create a movie. There are endless possibilities.
You can use these capabilities in your business. You can use AI to run a marketing campaign. AI can generate numerous options and you can select from those options. For instance, you can come up with 1000 variations of a face cream display, shortlist six and get them tested to see which of the six will work best, and then amplify that. AI here is not replacing the marketing team. It amplifies the marketing team’s capability. Instead of one version of an ad, you can have 1000 versions, each of which can be highly localised and personalised.
Everything we discussed so far deals with physical things that can be digitized- like words, images, music, video, etc. We live in a physical world. Can we build a model that understands our physical environment, and change the way we interact with this physical environment? Of course, we had robots in the past, but robots work like traditional programming. We are working on a futuristic model called Visual Large Language Model.
3 Audiences, 4 Pillars
At Google AI, we are focused on three audiences: consumers, communities, and companies. The question we ask is- how can we use AI to help these three audiences reach their highest potential? It means four pillars to our strategy. First, we help consumers. Consumer behavior will shift because of AI empowerment in their pockets or the desktops. As a result, businesses will adapt or will have to adapt to that.
The second, we look at big societal problems (like healthcare), and try to solve for them. The third area that we look at is how we can help companies market in a much better way to the consumers. The fourth one is besides marketing, every organization does a lot of other things – for example, a bank does a risk assessment; a shoe company does a product design; a shipping company does logistics. How can we make these processes more efficient? These are the four pillars of how AI will reach more consumers, more societies and more companies.
