In early June, the government think tank Niti Aayog unveiled its discussion paper on national strategy on Artifical Intelligence (AI) which aims to guide research and development in new and emerging technologies. While the paper per se focuses on five sectors for India — healthcare, agriculture, education, smart cities and infrastructure and transportation, this post will focus on healthcare alone.
In the paper, the think tank says India is not leading in healthcare because of issues such as shortage of qualified healthcare professionals and services like doctors, nurses, technicians and infrastructure. There are 0.76 doctors and 2.09 nurses for every 1000 people (as compared to WHO recommendations of 1 doctor and 2.5 nurses per 1,000 people). Hospital infrastructure is creaky and bursting at its seams with 1.3 hospital beds per 1,000 people as compared to the WHO recommendation of 3.5 hospital beds per 1,000 people.
Amazingly, the government believes that AI, and not fixing the inherent problems that are rotting the sector from the inside, can play a major role here. In the backdrop of India’s stretched and limited basic healthcare infrastructure, at one level the strategy for AI is indeed a bit like putting the cart before the horse.
New tech such as AI can often be spoken of to solve problems, but one forgets that tech is only an enabler and is not a substitute. It is only as good as its supportive infrastructure. The “intelligence” in AI basically comes from massive data repositories that the machine gleans information off. Not sure if India has much of that to boast of.
Even in the West where data is meticulously collected, stored securely and analysed, massive deployment of AI has still not happened and is still being tried out. In India which has 66% population in rural areas and 67% of doctors in urban areas, the requirements are still fundamental such as the need for physical hospitals, healthcare professionals and last mile connectivity for drugs and pharmaceuticals.
In a country where people in remote villages are just becoming aware of the magical effects of electricity; where clean drinking water and two meals a day are aspirational, talks of broadband, the internet and AI seem quite out of place.
Like most things that come from govt, these are noble intentions which pave the road to hell.
Over the long term, technology upgrades in the healthcare delivery model can create efficiency especially at the point of care level (in PHCs, clinics, hospitals etc). There are definite areas where the existing burden of care on doctors can be eased. But this has many pre-requisites – from as basic as having physical hospitals and doctors, medical supplies, electricity, and water to advanced things such as cloud-based tech, data privacy and security, trained personnel etc. The starting point would be to collate and build out massive data repositories, do away with silos and get different data collection points to inter-connect.
In a country as large and diverse as ours with a population as immense, this is a colossal project – one that can take years. Paradoxically, we Indians are great at managing scale (as we’ve seen with General Elections, the Kumbh mela and Aadhar enrolments) but are terrible at relatively basic tasks (simple queues, traffic rules and building roads & ensuring health and sanitation). One worries that we might succeed at putting up the technology but forget basic training and infrastructure creation.
If we get all that right, the potential is unlimited in areas such as screening and diagnosis, home healthcare, DIY healthcare for primary health (tend to simple cuts and wounds, treat common colds, aches and pains etc) and making health care participatory (arming patients with enough information for them to understand and participate in treatment).
Lessons can be drawn from tele-medicine (eg: The Sky program launched by World Health Partners in Bihar which didn’t do too well due to lack of participation from both providers as well as the targeted social groups). This of course, doesn’t mean that nothing new will work. It just means that a lot of things have to be in place for this to work.
One shouldn’t fall prey to the fallacy that technology is the panacea to all problems.
A good way to foretell the future is to study the past. And the past in this case doesn’t hold much promise. We have had “digital India” and “make in India” and now we have “solve for India”. The intention is noble, and the premise is strong. However, it doesn’t account for the fact that a very wide and capable ecosystem is necessary for innovators and entrepreneurs to survive. Especially those of the social nature who will dedicate their work to helping out strata of population that will not necessarily have the ability to pay. This will require major govt funding and backing which we haven’t really seen before. Plus, the fact that it takes years for such an ecosystem to thrive and investors have to be patient.
One must but look at China and how Govt invested millions of dollars into building such an ecosystem that is on the verge of catapulting China ahead of the US in terms of innovation. A recent MIT report that compared China and India in the race for dominance in AI, didn’t have very kind words for India.
There is a need to reskill many people in a short span of time. Insufficient research support, poor data quality, and the lack of expertise in the field will be stumbling blocks for India. China has had an early start with a robust digital infrastructure and their unmatched ability to produce PhD-level mathematicians and stem engineers.
The AI game is on. India is playing catch-up with China and the West in terms of technology, research prowess, investment, and most importantly, data, which is the lifeblood of AI.
How India grapples with issues such as water scarcity and poverty alleviation, while daring to believe that AI can change the world will be fascinating to watch. However, it’s important to remember that AI is just a tool, it is not an end.