To understand how AI will impact human lives, we interacted with Dr Prashant Pradhan, CTO of IBM India and South Asia. He leads a team of deep technology experts to drive IBM’s Platform conversations around Cloud and Cognitive with Enterprises, ISVs, developers, and startups. In his earlier role, he has also led IBM’s Watson business for India and South Asia. In this interaction he shares his views on AI enabled functionality, why it is important, how humans may lack if they do not adopt AI, and more.

Analytics India Magazine: What are some of the practical implementations of AI according to you that have revolutionised the way humans are functioning today?

Dr Prashant Pradhan: AI has been around for a while now, especially in areas like financial services such as credit risk and fraud analysis, which has seen interesting use cases in the past. But recently, we have been witnessing more interesting concepts across multiple domains that are adopting AI to solve complex problems. For instance, in healthcare, IBM is doing work across cancer and genomics with a flavour of AI. It is being used for instance to understand medical literature, spotting the right kind of treatment options available, look for stronger predictor for success of a particular type of treatment, and so forth. Apart from healthcare, AI has seen interesting use cases in areas like financial applications, education, and others.

AIM:  Please tell us about Watson. Who are some of the other clients apart from Manipal using Watson in India?

PP: We continue to grow in the healthcare domain and Apollo has recently signed up with us. Healthcare is a vertical that we are exploring in India. We are working with leading banks in the country who are adopting Watson for their insurance and investment-related engagements. We are also working with the telecom industry on how machine learning can be used to enhance performance in terms of reducing network call drop and others. We also have partners where they are picking up the technology to enhance their new offerings. For instance, a lot of companies in the skilling space and education space are trying to introduce personalised learning to enable candidates with the best learning pathway. In retail, we have announced facility in fashion metrics, which is the use of AI in terms of inventory and stock management, assortment, selection, and other areas.

AIM:  There is always a fear that AI would take away jobs. What are your thoughts about it?

PP: Practitioners always have a different view on this (chuckles). I think it’s very simple. Think about a doctor who is using AI to augment their capability of handling more complex cases or more number of cases. If they don’t use AI, they would otherwise take a lot of time to go through case history, literature and other background work. Humans have a limited capacity, even the best ones. If they have to view 100 clinical trials or new study, it would be time-consuming with no effective result. Also if the ratio of doctor to patient or teacher to student is not working in your favour, then there is a need to augment human capability and make them closer to human performance to improve the scenario.

The second aspect around jobs, let’s say doctors who have access to all the AI. What are the things that could happen? In the age of AI-powered or AI augmented world, if we become complacent, we are going to be lagging behind. An interesting way to summarise this is, if not explicitly, the only professionals those who are actually going to lose jobs are those who would not be leveraging AI and additional abilities that are available to live. Think about it, would a doctor be able to compete with another doctor who is using AI?

AIM:  Will AI take away the creative thinking and downgrade the intellectual quotient of humans?

PP: I don’t think so. AI is assisting professionals significantly in tasks such as understanding data, interpreting it or in decision support. If you assign these tasks to AI, the next natural response of each professional is going to be to look at the next creative thing. We actually are betting on the fact that this will lead to exponential growth in many of the fields. It might bring better treatment options or offer personalised medicines. All of these people in the field are actually waiting for the opportunity to spend less time on churning data and look at that as a way to be accomplishing a task in seconds instead of months or years. This ability of AI is going to boost our ability to do more creative work such as analysing data, learning from it and others.

AIM:  What are some of the other challenges that come on the way while adopting AI?

PP: If we talk about applications like chatbots, voice assistant, and others, they have a certain type of challenges such as better understanding the nuances of human communication. This can be dealt with by using a lot of publically available data to develop a good AI. For more serious applications, where decision-making is involved, it is a challenge to carefully develop the AI. For instance, when AI is used to deny or accept credit for somebody, or while making a person to be on the watch-list for committing a crime, it should go through rigorous check in terms of applicability and deployment. If you misclassify cat as a dog, it wouldn’t impact much, but if you make the wrong recommendation it might have some serious implications.

All this conversation largely gets centred firstly around data, so the availability of high-quality data is the key. The bar gets higher when you are dealing with high-end AI applications. The second thing that becomes equally important is the whole notion of transparency and avoiding bias.

AIM:  What are the changes in terms of policies and infrastructure that you would like to see for a better adoption of AI?

PP: Data is the king when it comes to AI. If we look at the cases where AI has been called out, there is a basic assumption that it is digitised and on top of that, you have to make sure that it is high-quality data. Most thinking is required in enabling an environment where we can thrive with AI applications. It has to do with good data policies and quality so that innovation can happen on top of that data, and this innovation is accessed and applied by professionals across the globe. Then there comes the cover around how the data is shared privately and governed so that there is no unintended use of data. Data policy is the single most important element that governs all of these elements.

AIM: How do you see the future of AI in Indian scenario?

PP: Like many other technologies, AI is at the peak of the hype cycle in the current scenario. India is no different. But the readiness probably may be looked at much more carefully in India than in other parts of the world. This goes back to my point of having an aspiration in terms of game-changing applications of AI in various fields and having relevant resources. In terms of larger game-changer such as agriculture, education, healthcare, and others, the sooner we get to address the challenge of collecting data, the sooner it will grow. The faster we can get data ready, the sooner the adoption will be in India.