With the growth of AI and technology, data science as a field might shrink: Nitin Seth
The co-founder of Incedo Inc, a US-based data transformation company, talks about his new book, the digital age, and the synergistic relationship between data and AI
Before co-founding Incedo, Seth has previously worked as a consultant and later director at McKinsey & Company; MD and country head at Fidelity International; and as the COO at Flipkart.
In his latest book, which became a bestseller on Amazon within a week of its release, Seth explores the intricacies and possibilities of data in the digital age. He addresses key questions about the role of data-driven AI in fostering innovation and generating value, while also providing strategies for managing the complexities of data effectively. In a conversation with Forbes India, Seth sheds light on data, AI, and the digital age. Edited excerpts:
Q. What inspired you to write Mastering the Data Paradox and what is its central thesis?
The central thesis of the book is ‘Water water everywhere, and not a drop to drink’. That's the paradox from the poem, ‘The Rime of the Ancient Mariner’ by Samuel Taylor Coleridge. That's the kind of story with data too. There is a data deluge happening, but at the same time, there's also drought because it's not very easy to get things out of data. We are in the AI age. We can argue about the good or bad of it, but I think we can all agree that it is inevitable. And, if something is inevitable, then you have to deal with it. And if you look at the AI age, the most central piece in that is data. AI is built on three things: Algorithms, compute, and data. Of them, both the hero and the villain is data. That's also the paradox.
Also read: 91 percent Indian corporate leaders believe their companies need to adopt AI to stay competitive: Report
I've seen it in my work with clients, having been dealing with data for 25 years now in different forms. During these years, what I have come to realise is the fact that the biggest factor in digital enterprises either doing very well or failing is data. And then I began to understand the reason for it. Why is it so difficult to deal with data? And I figured it’s something I call the ‘seven blind men problem’, as per which there is an elephant in the room and there are seven blind men who are trying to deal with it. One person finds the trunk, the other person finds something else, and what each person finds is the reality for them. And that's the thing with data. Somebody sees it as analytics, somebody sees it as data infrastructure, somebody sees it from a business perspective. And it has to be brought together. My experiences—starting in McKinsey, as a consultant and setting up analytics within the company, running Fidelity, then Flipkart which is a very data-driven business, and then running my company Incedo which is more focused on the technology aspects—provided me a very unique way of bringing all of these different threads together. So, one, I saw it as a very big issue for the times we are in, and second, it is not well understood. And I felt that given my set of cross cutting experiences, I had a somewhat unique perspective to offer on this, and that was the inspiration for the book.
What has fascinated me and stood out in my research for the book is the fact that 90 percent of the data of the world has happened over the last two to three years. There are thousands of years of recorded human history, and yet, the last two-three years account for more than 90 percent of the data. In the last 20-25 years, data has grown by 100,000 to 150,000x, which is a remarkably large number. There is nothing in human history which has grown that fast. When I wrote my first book Winning in the Digital Age, a lot of it was from experiences, but when I wrote this book, the process had to be very different. This has been a very, very research-driven effort, because this is a very dynamic topic and which is evolving almost every day.
Q. How do you see the role of data evolving in the next decade with the advancements in AI?
Data and AI have a very synergistic relationship. AI is dependent on data. And, lately, data is becoming so huge that I believe the only way you can deal with it is through AI. It's kind of both a cause and effect of each other, which is fascinating. In fact, I would add a third aspect to this, which is digital. I call it the Holy Trinity—Digital, Data and AI have a very recursive relationship. Now, I think data will continue to explode over the next 10 years. I think there'll be a lot more teaming of data that will happen. Right now, it's like a wild horse that has not been broken in. That’s what I have tried to bring about in the book. I think it's one of the first books of this type in the world, which is kind of bringing data together. But it's not that this is a unique issue. Every enterprise I know is dealing with these issues. I think right now the methods and the best practices to manage this data are not that well established. But I think over the next couple of years, there will be a lot more of that understanding and maturity that will happen. I am hopeful my book contributes to that. I think there's a lot more taming that will happen, and I think AI will play a big role in managing data, especially around data quality and data security.Also watch: Why you should worry about Generative AI
Q. Where on the one hand AI is helpful across activities in various industries, it can also be said to have a fearful element to it. Are there any common misconceptions about AI that you have come across and have addressed in the book?
I acknowledge that AI is both good and bad. I don't think we can say that AI is only good. There is a big unknown with AI. The way AI is different from other technologies is that it is self-learning. So far, technology has been about making human life simpler, easier in some fashion. Now, for the first time, we have technology, where it's not just about that, it's actually more intelligent than humans in some ways. So there is a lot of unknown with AI, which I think we have to deal with. But the way to deal with it is not by being like an ostrich and kind of putting your head in the sand. That would be the worst way of dealing with it.
There are a lot of misconceptions I've tried to break around data. The traditional approach of dealing with data is more technology driven—first getting all the data together in one place. That seems most logical. My contention is that it seems logical, that is what companies do, but it never works because data is growing at a much faster pace. Companies start with the good intentions of bringing data together, but it just keeps on becoming bigger and bigger and bigger.
Another misconception is of ‘customer 360’, which means bringing together a full customer view. I believe that you can never really get all the variables about a customer today. These are some of the key truisms which are there in the world of data, for which I have a very strong point of view in debunking some of these traditional ways of how we have been doing things so far.
Q. There is one school of thought that believes that AI is going to take away people’s jobs. What is your take on that?
It surely will. I think it's not very visible right now. But I think AI will impact most jobs over the next two to three years. If I just look at the clients that I work with, most of them have targets which are about 25-30 percent productivity improvements over the next 12-18 months. Now, I think if this much productivity improvement will take place, then some jobs are bound to go. I have no doubt that jobs will go, but new jobs will get created. Today we have data scientists and data engineers, maybe in future we will have more categories of engineers managing data. Today, organisations are spending more on AI than the benefits they're getting from it. I think the key question is not if AI will take jobs, I think it is how each job is going to get redefined, and how we can be proactive about it. How do we upskill and reskill people towards that, as opposed to waiting and kind of hoping.
Q. In your opinion, what are the most significant ethical considerations when dealing with large datasets and AI?
One of the biggest issues is around privacy. Who owns the data? The moment you engage with anything which is digital or AI, you have to share your own information. For example, if you go to Amazon or Netflix, you get a personalised service. How do you get something which is personalised? You get it because you're sharing your own data. Once you share the data, you don't have full control over how that data is going to be used. That I think is one of the biggest ethical considerations. And, I think for that we need to have global data privacy standards, which I have even mentioned in the book.
Also read: AWS launches generative AI service Amazon Bedrock in APAC region
The other ethical consideration is about bias. Especially when you have large language models, if you introduce a bias in one thing, it can have very long-term ramifications. And that also gets down to data. A bias comes because you're introducing some wrong data. When you think about ethics, I think there are many considerations, but I think privacy and bias are probably the top.
Q. What do you believe are the most exciting developments in AI currently?
The technology itself has grown much more than how we are applying it. From a business perspective, one of the most exciting things for me is hyper personalisation, in which we are talking about a segment of one. Traditionally, marketing has been driven by segments, but those were always based on approximation because data was not available. Now that so much data is available on each individual, you don't need to proximate. Hyper personalisation is used in social media. It is used in shopping. It is used in entertainment. It can further be used in education, healthcare, and many other industries. To me, this is one of the most exciting advancements.
Q. Where do you see the field of AI and data science heading in the next five to ten years?
Data science is a field that was created in the last 20 years, and now with the growth in technology, I feel data science is actually shrinking. And the reason I say that is that a lot of what a data scientist was doing historically is now embedded in the platforms. That job is going and is getting transferred into either the data engineer who has to bring this humongous amount of data together, or it is getting into the more classic business analyst job, which is becoming a lot more data savvy. The data scientists jobs that will remain will become highly specialised and technical. I feel that with the growth of AI, technology, and more cloud-based platforms, data science as a field might shrink. It will get more specialised. It's a somewhat controversial view. I talk about it in my book as well.
