The national AI portal of India, INDIAai, has completed one year of being the most trusted content powerhouse for artificial intelligence in the country. To mark the momentous occasion, we brought together some of the most authoritative voices to lay down the future course for India as it paves its way to becoming a global leader in AI. The dignitaries included: Ms Debjani Ghosh, President, NASSCOM as the moderator; Mr Nandan Nilekani, Co-founder and Non-executive chairman, Infosys; and Mr Ajay Sawhney, Secretary, MeitY.

In a fireside side, our prolific speakers took stock of the past year of India's AI journey and chartered the path for another stellar year.

Against the backdrop of the massive digitisation carried out in the country, Nilekani believes that India is on the cusp of some major AI innovation. “I think the time is right to apply AI across the board to India's most challenging issues, whether it's healthcare, education or judicial pendency, and so on," he says. Hailing the National Language Translation Mission of the Ministry of Electronics and IT (MeitY) as a key disruptive opportunity for India in AI, he believes that the nation's culture of multilingualism makes it well poised to create the best capabilities in language technology.

Sawhney, who is the key architect India's AI mission, believes that the primary focus of the country will be on the application of AI to real-world problems, such as for improving service delivery or decision making. While the core AI research is happening under the NM-ICPS mission of the government, its application, he believes, will happen on top of the national public digital platforms that are also starting to take shape. Citing Aadhar and UPI as its shining examples, he makes a convincing case for platformisation. "These platforms organise the data in a systematic form across the entire country. And as soon as you have a way of pulling up data and systematically making it available to users, researchers and startups, it is magic." Therefore, we will see the emergence of similar open national public digital platforms in healthcare, education, urban developments, logistics, and language technologies. "We must recognise that our primary strength in India is India's scale, and technology plays on top of the scale." Notably, it is scale which generates data and demand to create a powerful combination. "When we have data, we have the demand, we have talent, we have a supply of services – tremendous amount of innovation can happen on top of that," adds Sawhney.

Stressing further the importance of application, impact and scale, Ghosh says, "My biggest takeaway from the pandemic is that leadership is not going to be defined by talk or papers. Leadership is going to be defined by actual impact: how big a problem are you able to solve?"

Nilekani gives some pertinent examples of population-scale implementation of AI in India. "In the Aadhar example, they have this ability called face authentication, where they compare your photograph to the photograph in the Aadhar database, and then authenticate just like a fingerprint or OTP; that's entirely built with AI. It's a national model, a model for a billion people, and it's trained on the server, and they do AI on the edge by doing liveness testing on your smartphone using AI so that there's no fraud." He further cites the use of neural machine translation by the Supreme Court, use of analytics for tax compliance in GST, use of AI by NPCI for fraud detection, and OCR use by Gujrat's education sector. "We obviously have the population scale demographics, we have the demand, and we also have data because over the last several years we put in so many transactional systems, we now have the data to feed." He credits tech companies for enabling easy availability of open-source AI models, citing Google's BERT for machine language translation and Facebook's Wav2vec for speech recognition.

Taking about the need to collate datasets and make them available to all stakeholders, such as direct service providers, researchers, startups, and companies, Sawhney says, "It's not just provision of data, but actually rewarding what you can do with data after innovation and after applying the new technologies and AI." These public datasets, however, must only be made available for use in a responsible manner, as has been done in the case of Aadhar, GSTN, and UPI. Even though the ethical use of data and AI must be self-regulatory, Sawhney believes that it should be guided by auditable and certifiable dos and don'ts that are accepted across the industry.

Nilekani believes that the Indian industry is already at the cutting edge of applying AI to global issues. Comparing India's AI trajectory to that of China and US, he says, "We have a chance to have a third way of AI, which is in a much more open, collaborative way, where the government and private sector come forward and build joint architectures which allow everyone to contribute but nobody to dominate."

Ghosh adds, "Collaboration, partnership and an open model can be our biggest differentiators, because that's really what the world needs."

Sawhney gives his valuable advice to the INDIAai team. First, we must keep track of the latest AI innovations and developments. Second, we must be able to benchmark what we are doing vis-à-vis what is happening across the globe. Third, we must propagate and encourage the bridging of gaps between the laboratory research and actual application of AI. Fourth, we must raise awareness about the responsible use of AI. And finally, we must celebrate success – the success of startups, of departments and of unsung heroes – by bringing out stories of the people who are behind those successes. These words shall be our guiding light as we embark upon another year of bringing unparalleled AI coverage for all our patrons.

Once again, team INDIAai thanks you for your continued support and trust which gives us the opportunity to look onwards and upwards!