BuddhiMed at Cypher 2019 — merging medicine with computer science

A short talk from Cypher 2019, India's analytics and data-science summit, given while Dr Bakshi was leading BuddhiMed Technologies. He is asked where the opportunity in healthcare AI sits, and answers by putting the doctor, not the algorithm, at the centre of it. What follows is an edited transcript.

What we call AI today is the product of more than forty or fifty years of work, going back to Alan Turing in the 1940s. But the real excitement has come in the last ten years, and the reason is simple enough: there is far more data available, there is far more computing power available, and the mathematical foundations have been laid that let us point the tool at genuinely complex problems.

You asked about the opportunity. Healthcare is a fundamentally complex industry, because it deals with the lives and the misfortunes of a great many people. All of us fall ill. All of us eventually die. Our understanding of disease, and of the technologies that treat disease, has improved exponentially over the last forty or fifty years. But it has now reached a stage where it has gone beyond the ability of any single doctor to understand all of it.

So at a very basic level, what we feel at BuddhiMed Technologies is that AI is a useful tool for the doctor. Almost a humble assistant. The AI goes and collects a great deal of information, processes a great deal of information, and lets the doctor make his or her decisions better. That is a powerful opportunity.

The doctor is not expected to learn AI. I think what comes next is technologists who work with data and bring the power of that data to the doctor's fingertips. Doctors have already learned a great deal. They do not have the time to learn new disciplines, and I do not think it is required of them either. What is required is that doctors frame the problems they face. I was a doctor, and I am still a doctor, so I can articulate a few of the problems my colleagues from computer science and from data analytics then try to solve. Together we deliver the answer to the patient.

Which is why I think this notion of AI replacing humans is very far-fetched. What will actually happen is that AI helps people make better decisions, and understand complex situations that are otherwise very hard to understand.

We are excited about what is next. We are ingesting and processing a very large set of Indian healthcare data, and building some good use cases on top of it. For doctors, as I have said, but also for others working in the system, and, more interestingly, for the patient. That will come out over the next few months.

On the conference itself: I appreciate that India is becoming a global centre for artificial intelligence. Go around the world and there is the Bay Area, there is New York, there is London, and then there is Bangalore. That is a privilege to have. We are still in the early days, but conferences like this one, which bring people into contact and let them collaborate, are how the ecosystem gets built.