India's AI Adaptation: Insights from ET Women's Leadership Forum

'Decoding the AI Disruption' panel discussion at ET WLF: Take AI and adapt it to India's needs

India holds immense potential for personalised AI platforms leveraging extensive local data sets, said Nigel Vaz, CEO of Publicis Sapient. The panel discussed AI applications in various sectors including healthcare, education, and traffic management, while flagging infrastructure challenges. Microsoft's Physicswallah and GitHub's Copilot were highlighted as key AI-driven innovations.India offers great potential for personalised AI platforms that leverage the vast and abundant data sets locally available, Nigel Vaz, chief executive officer (CEO), Publicis Sapient, said during the 'Decoding the AI Disruption' panel discussion. "India's got tremendous access to diverse data sets, and you can start to build very specific solutions for India by India, because India's also got an extraordinarily tech-literate workforce," Vaz explained. "So, you can start to think about things like city governments or state governments carrying traffic data and traffic patterns with the fastest routes to the hospital for people. You can start to imagine so many of these possibilities where the personalisation happens at multiple tiers. It happens at the tier of the country, it happens at the tier of the state or the city. It happens at the tier of the hospital, and then ultimately, at the tier of the patient." And, he said that for every one of those, the more interactions one has, the more one will start to create data sets to train models to create more specific solutions, like for instance, traffic data and health data prioritising access on the streets to allow somebody to get from where they are to a hospital at the earliest. The panel also explored AI's applications in healthcare, enhancing medical imaging and diagnostics, and in education with a particular emphasis on providing inclusive STEM education. Ahmed Mazhari, president, Microsoft Asia, brought up the example of Noida-headquartered Physicswallah which focuses on STEM education. STEM is an acronym for science, technology, engineering, and mathematics. "They have 40 million students on their platform and one of the programs they run is for ₹4,000 a year," Mazhari said. "That can only be enabled by using data sets and AI technology at scale, because ₹4,000 a year creates an ability to empower kids from the most rural communities in India and provide them with an equal platform, where they can go to an IIT, or they can go to an NIT. So I'm excited about what we can do and how we can be more inclusive." Thomas Dohmke, CEO, GitHub, too echoed said that one of the most fascinating use cases of Copilot was when children used it to learn about the world or learn about coding. And he said an important aspect that hadn't been highlighted was that it can engage with users, not just in English, but in other languages like Hindi, or Spanish, Portuguese and many other major languages. He also spoke of the democratization of AI through tools like GitHub and Copilot. "So a six year old that is fascinated by Minecraft or by Super Mario and wants to build their own little game, they don't have to learn English first to figure out how that all works, and can start asking Copilot questions of how I do that," he said. "In the past, you always had to have parents with a technical background or you had to know somebody to ask all the questions that you have. Copilot does that for you. Does it for the parents. It has infinite patience and it can answer any questions without any bias. And so kids can really explore coding." Dohmke also made a strong case for the open source model and said it was "better than any other collaboration on Earth," as it has no bounds, borders or boundaries. He also said that it will help to create a better world, because developers around the world will keep innovating in open source with the help of AI, data, open baits and open source large language models. While the panelists discussed the vast potential that the space provides, Nivruti Rai, CEO & MD, Invest India, flagged important infrastructural issues that will be crucial for India to navigate in order to make the most of the opportunity. "300,000 kilowatt hours- that is the electricity 300 people will use for an entire year to run one ChatGPT compute, leveraging 1 million GPUs. The world today uses 26 petawatt of electricity and data centres use half a percent. The US-3%, China-4% and India-0.05%. So in the next few years, we will get to half a percent, which needs so much energy consumption," she said.

'Decoding the AI Disruption' panel discussion at ET WLF: Take AI and adapt it to India's needs

India holds immense potential for personalised AI platforms leveraging extensive local data sets, said Nigel Vaz, CEO of Publicis Sapient. The panel discussed AI applications in various sectors including healthcare, education, and traffic management, while flagging infrastructure challenges. Microsoft's Physicswallah and GitHub's Copilot were highlighted as key AI-driven innovations.