India Should Prioritize Practical AI Over Expensive LLMs, Says Mohandas Pai
Mohandas Pai, former Chief Financial Officer of Infosys, has advised India against pursuing expensive Large Language Models (LLMs) and instead urged a focus on practical Artificial Intelligence (AI) applications. Pai made these remarks on Wednesday during the launch event of eKosha, a new voice-first merchant banking platform developed by the fintech firm ToneTag. He suggested that India's resources and efforts would be better directed towards developing AI solutions that address immediate needs and offer tangible benefits. The emphasis should be on creating accessible and useful AI tools rather than competing in the high-cost race for advanced LLMs, which require substantial investment and may not yield immediate widespread utility for the country. This strategic shift, according to Pai, would allow India to leverage AI more effectively for economic growth and societal development. The launch of eKosha, a platform designed to simplify merchant banking services through voice technology, exemplifies the kind of practical AI innovation Pai advocates for. ToneTag's new offering aims to enhance financial inclusion and efficiency for merchants. Pai's comments highlight a potential divergence in AI development strategies, advocating for a pragmatic approach tailored to India's specific context and developmental goals.
The discourse around AI development often presents a dichotomy between cutting-edge, resource-intensive models and practical, accessible applications. Mohandas Pai's suggestion for India to focus on the latter reflects a strategic consideration of national priorities, resource allocation, and potential return on investment. While LLMs represent a frontier of AI research and development, their high computational costs and specialized applications may not align with the immediate needs of a developing economy seeking broad-based technological adoption. Prioritizing practical AI could foster wider digital inclusion and address specific sectoral challenges more effectively, potentially creating a more robust domestic AI ecosystem. This approach may also mitigate risks associated with dependence on foreign-developed, high-cost AI infrastructure, encouraging indigenous innovation tailored to local market demands and user bases. The long-term implications involve balancing the pursuit of global AI leadership with the imperative of leveraging technology for immediate socio-economic upliftment.
AI-generated to prompt reflection — not editorial opinion, not advice, not a statement of fact. How this works.