2023: The Year of AI
AI is an ally and not an adversary. When employed for our benefit and with care, this tech has the potential to level the extremely uneven playing field in India.
By Nandan Nilekani
2023 is the year when AI will begin to forever change the world — and India — as we know it. We are living through a seminal moment in history. But this is not the first seminal moment in India’s digital journey.
The recent advances in artificial intelligence and the revolutionary promise they carry for our society take me back to another pivotal time not very long ago. The year 2016.
By April 4 of that year, Aadhaar had reached 1 billion Indians. That instance of providing widespread access to identification actually laid the foundation for doing things at scale in India. Then on April 11, we launched UPI, India’s national mobile payment system. India now tops the world in real-time digital payments. Soon after Jio was introduced in September that year, and mobile connectivity shot up, data costs dropped dramatically, and data usage went from half a gigabyte a month to half a gigabyte a day. Demonetization followed in November, and in December, the Prime Minister launched the BHIM application, which contributed to our comfort with electronic payments.
When you connect the dots to each of these separate events, the picture which forms is that of India’s enviable and singular digital public infrastructure. The experience of building such population-scale technology has put us ahead of many other nations.
But there is more work to be done. As the home to more than 1.4 billion people, India is the world’s most populous country. The country’s human capital is its most important capital, and it cannot be just a small set of elites who gain access to high-quality services. Any new technology must be employed towards benefiting every section of its large population.
I have no doubt that well-considered AI interventions hold the key to closing the gap between India’s potential and India’s reality.
AI as an ally
In 2023, there will be a plethora of AI innovations reaching critical mass, creating another pivotal period of time. We have the government programme Bhashini, the AI for Bharat initiative at IIT Madras, our work on AI at EkStep, in addition to the advancement of generative AI and LLMs such as Bard, ChatGPT, and LLaMA.
Connect the dots to these and we have an opportunity to create a more equitable society.
As we work towards adapting and applying AI tools for use in India, it helps to clarify what is not going to be the end goal for this technology. We are not trying to build autonomous cars so that we can get rid of drivers. We are not trying to build health platforms that no longer require doctors. We are not trying to build a justice system with no need for judges.
We want AI to be an amplifier. It will be an ally for humans rather than an adversary, and a collaborator rather than competition. As we build our AI infrastructure, we can repeat the inclusive and expansive approach we took while building our digital infrastructure. To my mind, this technology can contribute towards solving some of our most critical concerns:
Education
In a country with a short supply of teachers, we need more high-quality education to reach millions of children.
India has opened a remarkable number of schools, yet finding enough qualified educators for them remains a struggle. As a 2021 UNESCO report established, we have about 120,000 single-teacher schools, from which 9 out of 10 are in rural areas, compounding the challenges of an already underserved population. No wonder then that the NGO Pratham consistently observes only marginal improvements or drops in children’s reading and numerical abilities in its Annual Status of Education Report.
The DIKSHA (Digital Infrastructure for Knowledge Sharing) has touched millions of lives but while we have managed to push learning and enrolments, there is a lag in learning outcomes. Through conversational and generative AI, acting as a teaching tool or teacher’s assistant, we can significantly ease this problem.
Agriculture
In an economy where agriculture is crucial but agricultural productivity is still poor, we need to enrich farming practices with more efficiency.
By combining data on farmers and agricultural methods, we can cut wastage and draw more yield out of each unit of land in the country. Precision farming will let us make the most out of limited resources and investments.
The Agristack Project is already a major step in gathering this information.
Healthcare
In a society with a shortfall of doctors and nurses, we need more high-quality healthcare to reach the people.
The ratio of doctors to India’s population is 1:1445, which falls well below 100 doctors per one lakh people, the standard prescribed by the World Health Organization.
AI tools have the potential to revolutionise the diagnosis, treatment, and monitoring of patients.
Law & Justice
In a scenario where there is a formidable backlog of judicial cases, we need to maximise resources and reduce pendency.
Because courts across India are understaffed, they were dealing with at least 47 million pending cases, based on a count as of March 2022. About 70,000 of those cases were pending in the Supreme Court, and 40% of them had been pending for more than five years. In the 25 Indian high courts, there were 5.9 million pending cases.
Amplification with AI will improve productivity and ensure that justice is neither delayed nor denied.
Addressing the challenges
India is no stranger to adopting technology at scale and achieving consensus over its use. We know we can bring people on the same page and work through problems together. We know this from having seen the playbook before.
With Aadhaar, we hit 200 million authentications a day, and 10 million KYCs a day. For UPI, the current run rate is 300 million transactions a day by 300 million users at 50 million merchants across the country. More than 150 million people store 5 billion documents on Digilocker, so that they can walk into the airport by simply flashing their driver’s licence or Aadhaar over the smartphone.
These digital initiatives have also brought important legislation onto the fast track. In a period of 10 years, we went from being a country where nobody had heard of privacy to one where every stakeholder — judges, lawyers, plaintiffs, defendants — understood privacy and established a solid framework for it. The benefit of that touched more than just Aadhaar, leading to a marquee judgement on the fundamental right to privacy.
As AI grows, the discussion will bring up challenges, such as copyright, explainability, bias, and hallucinations. The way we navigated through questions of privacy and security earlier tells me that we can harness the full power of artificial intelligence while also putting in the necessary guardrails.
A sovereign and self-reliant approach
It is fortunate that the AI world so far has by and large published the technology as free and open source. In fact, the large language model underlying all the tools we have today was based on a Transformer paper published by Google. I hope that that continues and that the AI arms race does not make vendors become far more closed in the way they approach things.
One of the tenets at the core of all the things we are building in India is that we want India to be in charge of its own destiny. India should have its own technology and, at the same time, leverage global developments. We have to be very clear that we will study and abstract the best available AI tools for our design architecture, and have an AI strategy where India will ultimately be in charge of its own destiny.
Already, I feel encouraged by the difference in the nature of the debate over AI in India versus the rest of the world. Across the world, a lot of the focus is on how to make money from AI or how to help businesses use AI. But the challenge in India is how do we use this to create a better society? How do we use this to create a more equitable society? How do we use this to give every Indian the opportunity that he or she deserves? And, how do we use this to ignite economic growth?
There are few other countries where everybody — the political leadership, the bureaucracy, the business leaders, the technocrats — realises that digital transformation is the only way we can fix large problems. If we can bring in exponential change through digital transformation, we can make a difference in everybody’s life.
Where we are standing, AI is not a castle in the air. It is in use and is making a difference on the ground. It is already a part of Aadhaar authentication. Today, because AI is used to monitor tax compliance and leakages in the digital systems of GST and income tax, the country’s tax-to-GDP ratio is at historic highs.
This gives me the confidence that the next five to 10 years are going to define the future of India. A future where children are better educated, where people are healthier, where they can use technology in any of India’s 120+ languages, and where everybody can get every imaginable service they want. A future where nobody is left behind.