After her Masters, Chandrima studied for a PhD in Applied Mathematics and Theoretical Physics with the Jawaharlal Nehru Memorial Scholarship at Cambridge. After a Postdoctoral Fellowship with the Lindemann Foundation at Dartmouth College and a Junior Research Fellowship in Wolfson College, Cambridge, she changed careers and started working in AI Ethics and Responsible AI.

She is currently the Technical Lead for the Responsible Gen AI Team at the Centre of Excellence for Data Science in Lloyds Banking Group. She has also been nominated to the global 100 Brilliant Women In AI Ethics list for 2024.

We spoke with Chandrima to know more about her work, interests and aspirations.

--

Could you describe what you do as part of your work in AI Ethics, and why?

I am currently the Tech Lead for the Responsible Gen AI Team. As a scientist, my passion has always been to understand how to translate the needs of end users of technology, regulatory requirements, and societal impact into technical requirements towards which to build solutions.

My team builds the technical safeguards to ensure that Gen AI applications, when deployed by Lloyds Banking Group, remain safe for end users — especially those of the Bank’s customers who are the most marginalised in society.

AI, and how it is to be used, is an evolving, high-stakes conversation in a new territory. What are your thoughts on this?

The use of AI, and the terms under which it is used, are crucial to clarify as this technology is used more broadly. My strong stance here is that if this use and terms of the same are not determined democratically, then collectively we are not building, deploying or using this responsibly.

If AI is the revolutionary technology it is supposed to be, then it is our duty to find ways in which it can be used to solve or alleviate challenges faced by the most marginalised in society. This should and must take precedence over finding use cases which can simply add convenience to the lives of those of us who are more privileged. This must take precedence over finding ways to enhance the technology towards making it ‘generally intelligent’ — which comes at great cost to the environment, and makes it inaccessible to people who are less resourced.

My aim as I enter the next phase of my career is to do exactly this. To find versions of AI models that are cheap to run, easy to use, and train them to solve problems relating to meeting the material needs of the most marginalised people in our country.

Yours has been an interesting journey professionally. What are some of the lessons and challenges you’ve experienced along the way?

One of the biggest lessons I have learned is the value in flexibility. When I first started on my journey with Inlaks, I was convinced that theoretical physics was the field that I wanted to pursue for life. I was surprised to learn that as I grew in the field (and in life) that a just world, and my role in creating one, began to become the most important thing. I began to see that while we may feel powerless to bring this about individually, collectively we have immense power. I realised that everyone has skills and knowledge that can be used to bring this to fruition. And it was at this moment that my eyes were also opened to how narrowly I had defined my own skillset.

Setting myself free from restrictive definitions of my own abilities and skills is perhaps the biggest lesson I have learnt to date.

I am committed to using the skills I have obtained over the years to learn more, grow more, and ensure that I can facilitate others to feel empowered, and to also work for a just future for all people, rather than a few.

What are your plans for the future? How would you like to contribute to the field?

My biggest lesson about myself is that I find the greatest fulfilment in working to bring about a just, more equal future for more of us. Being away from home for over a decade has been difficult. I have found great friends, and learnt a great deal over this time. When I was awarded the Inlaks Scholarship in 2013, I knew that it was an opportunity to learn and grow — but I also knew that one day I would want to reinvest that knowledge and growth into India.

I feel that this time has come and I am writing this at a turning point in my career. After years of working in Responsible AI, I feel it is time for me also to use AI responsibly, which, in my definition, is to find use cases for AI that are democratic, that are inclusive.

I believe that the transformative nature of this technology, when applied correctly, can alleviate some of the most pressing problems faced by marginalised rural communities in India. I believe one of the most pressing of these to be the lack of consistent access to clean drinking water.

To ensure that AI algorithms used can be accessible by more people, my aim is to use smaller, lightweight versions of AI models that can be run offline on mobile phones, or on a 3G/4G network for example.

The problem I wish to tackle is as follows: how can we accurately and continuously monitor the quality of clean drinking water in a cheap, scalable and resource-effective way?

My aim is to implement, test, deliver and proliferate a solution that can continuously and effectively monitor water quality, using AI models, rather than the current manual, less accurate solution.

This solution would have transformative consequences - it would prevent disease outbreaks, and if implemented in an operating and distribution model correctly, could drastically improve the quality of life of millions of people.