Nandan Nilekani, the co-founder and chairman of Infosys, has suggested that integrating AI with businesses will require more effort over a longer period of time as compared to the deployment of AI models such as chatbots for consumers. In an interview with CNBC on Tuesday, September 17, Nilekani said, “I do believe that enterprise AI, which is what companies have to do, is a longer cycle than consumer AI, because it requires firms to reinvent themselves internally. It’s a longer haul but it’s a huge thing that is happening right now.” He further said that most companies are likely to have structured or unstructured data from disparate sources that exist in silos and firms have to organise their data to be consumable by AI. However, appearing as a guest on Zerodha co-founder Nikhil Kamath’s podcast, Nilekani suggested that AI adoption far outweighs its potential costs. Referencing global companies that spend as much as $50 billion a year on AI infrastructure, he said, “They can’t afford not to do it (invest in AI), so they rather spend $50 billion a year than not spend…if there is a big platform shift and they don’t have it, then they are out of the game permanently.” The 69-year-old, who was recently included in TIME’s 100 Most Influential People in AI 2024 list, is optimistic about the impact of AI on jobs. “While the productivity of software development will improve considerably, in reality, we are in the early days of a ten or twenty-year cycle of reinventing companies with AI technologies. So, while the productivity will come, the new business that will come up in enterprise AI transformation will more than make up for it,” he said. “There’s going to be enough work for firms to do the AI transformation which will make up for the fact that the productivity is also going up,” he clarified to CNBC. When asked about the pace of AI development in India, Nilekani said, “There are many companies coming up that are building India-specific LLM solutions for Indian languages. We have a company called Sarvam AI and Krutrim. Large firms like Infosys are also investing heavily in AI. I think what is going to happen is that since AI is dependent on the data it is trained on, every part of the world which has unique data will be required to do something about training their models for that data.” “Ultimately, the AI models will become commodified and the value will switch to the application layer of the whole stack,” the Aadhaar architect added.
