Episode Summary: We usually discuss the impact of artificial intelligence on a business’s bottom line, but, in addition to law enforcement, governments and NGOs are also considering AI as a mechanism for improving society.
We spoke to Anandan Padmanabhan, CEO of the Wadhwani Institute for Artificial Intelligence in India, as part of our research report on AI in India. This week on the AI in Industry podcast, Padmanabhan speaks to us about where and how the public sector should consider leveraging AI.
Padmanabhan discusses the challenges that the Indian government faces in providing education and healthcare to its citizens. Although AI might help overcome these challenges, those who need these services most may not have access to the technologies necessary to work with it.
Subscribe to our AI in Industry Podcast with your favorite podcast service:
Guest: Anandan Padmanabhan, CEO of the Wadhwani Institute for Artificial Intelligence
Expertise: AI, computer vision
Brief Recognition: Padmanabhan is a distinguished researcher in computer vision and artificial intelligence. His career spans over 30 years in academia and industry in the US and India. He was the Founder and Managing Director of Microsoft Research India, Vice President for Research at Adobe, Group Manager at Sarnoff Corporation, and Assistant Professor of Computer Science at Yale University. He is an alumnus of the Indian Institute of Technology (IIT) Madras and the University of Massachusetts Amherst, and he is a member of the Board of Governors of IIT Madras.
Interview Highlights
(03:40) How will the next 3 billion people be reached by AI, and how will the technology tangibly benefit their lives?
Anand Padmanabhan: There is a growing belief that AI technology can be extremely transformative and can make our lives better. But if you look at the concentration of this technology and its benefits, it is primarily limited to the haves: the middle class, upper middle class, and the rich who are part of the mainstream market economy. In India, there are about 300,000 or 400,000 in this segment. That means the remaining 800 million do not have access to AI.
The next 3 billion are those whose needs are not served well, whose economic capabilities are not good enough to participate in the AI process. They are farmers, merchants, or even those that have their own business.
When it comes to education, healthcare, or guidance in their own profession, they rely heavily on government and charitable non-government organizations. How do they get taken care of? Firstly, government programs usually have offices in every district. They function somewhat, but there is a lack of efficiency and a lack of knowledge. Secondly, there are frontline workers, either volunteers or paid workers who need to translate these processes. And thirdly, other services that can be delayed.
I see AI entering government through these three means. Government officers can improve their efficiency and be more effective in focusing on problems. AI has the power to correlate real information in the data and outcome better than any other technology. It looks at the data and finds what is relevant to the outcome, finds similarities in the data, and highlights it.
Even using AI in the simplest things, such as speech-to-speech translation for dialects, can be very useful.
(07:47) Can we have some examples of AI for social good? What are some of the initiatives and how are they working?
AP: The Prime Minister of India came to inaugurate our institute because he realizes that this technology needs to be harnessed for social good. He wants to be part of setting the agenda and shows support at the national level. We have also received a lot of support from the government of Maharashtra.
Why? For example, public health initiated hospitals and clinics. In northern India, there about 150,000 subcenters in each village, a few thousand primary healthcare centers, then district-level hospitals. People would like to go to the subcenters to receive proper healthcare. However, their number of trained medical professionals at these centers are inadequate. Frontline workers try to help and cope, but they also are not trained medical professionals.
Here is where the power of AI can fill the gap to provide automated guidance: by understanding how a medical or health process is done in other places to reduce the dependence on the few qualified medical professionals.
The challenge is not purely technological. It is understanding the user experience and use-case you want to drive. You can’t just throw in the technology without being capable of using it.
You want to work with the partners who have invested time, government partners, and NGOs who understand how to work with this audience and design your technology towards that direction. We are working in close partnership with non-government organizations. Ultimately, the government has the capacity to scale efforts at the national level, so it is important to keep them involved from the beginning.
(10:55) How do we get these health centers the equipment? How do we ensure health professionals know how to use it? How do we put this process into the normal loop? Can we change their behavior?
AP: One challenge startups have is market development. How do you develop your market? The problem is more acute in social good because it is not a single, uniform market that you can directly reach. You have to go through partners and get them on board from the beginning.
So doing innovation and R&D in social good has not been attempted because of the market challenges. We are excited and have a good start. We believe that as we do it, we will learn what some of the bigger challenges are and the ways we can get around them.
(13:15) What others ideas are you exploring? What are the other exciting areas for government collaboration?
AP: Agriculture is another area. If you look at the Indian economy, we are very agrarian. One of the challenges the country faces is that farmers are constantly in debt. Their investments in small lots are not always fruitful due to climate conditions, soil problems, diseases, and market dynamics. For them to succeed, they need to have a unique positioning in the market. However, they are not equipped to manage and control market forces.
But there is data available related to climate or soil, companies that offer sensors that measure these factors. There is technology that looks at crop diseases and people who understand how markets play. AI can combine these things to potentially guide and advice farmers or government agriculture offices. If we can reduce the farmers’ loan burden and effectively and efficiently use resources, that will be a good thing.
(16:30) From a high level, what will it take for India to turn machine learning and AI revolution into a net win?
AP: The National Institution for Transforming India recently released a white paper that addresses India’s AI strategy.
There are similar papers released by the Ministry of Information Technology and previously Ministry of Commerce and Industry. If you look at these papers, there is a fair amount of convergence and, that is, India’s leadership position in AI for all. It is a unique opportunity for India to have inclusive AI for all.
Subscribe to our AI in Industry Podcast with your favorite podcast service:
Header Image Credit: Matera Basilicata 2019
