India’s AI ambitions are often measured by GPUs, sovereign AI models, and billions poured into data centres, but Tech Mahindra’s Chief Innovation Officer and Global Head of AI and Emerging Technologies, Nikhil Malhotra, believes India is measuring the wrong things. He argues that India’s AI future depends on producing original research, solving uniquely Indian problems, and creating breakthroughs the world follows, not on building another LLM. Malhotra notes that there has been no Indian contribution to the transformer architecture powering today’s LLMs, and he advocates for more research-based startups. While he supports building sovereign foundation models for technological independence, he says sovereignty alone cannot fuel innovation. He admits the ecosystem, including Tech Mahindra, has contributed to the problem, stating, “I am as guilty as anyone else for not putting papers on arXiv,” and criticizes a focus on patentability over research. Malhotra believes India lacks a culture that rewards research, not talent, and calls for organisations built around long-term scientific work similar to Google DeepMind. He also highlights the “diffusion of AI,” saying India has barely scratched the surface in applying AI to real-world problems, such as public transport, where he questions the need for a ticket to enter the metro.