AI sovereignty will happen, and India will achieve frugal AI, says IITM director Kamakoti
What Happened
Prof V Kamakoti believes India is charting its own course to achieve authority in the field of artificial intelligence, and will come up with domain specific language models, which can be trained with fewer resources compared to LLMs.
Fordel's Take
This sounds good on paper, but let's be real. Achieving 'frugal AI' doesn't just mean using smaller models; it means mastering the domain-specific language models that don't require massive, wasteful LLMs. It's about making models work for specific industries without drowning us in GPU costs.
It's a strategic pivot, not a tech sprint. We need to stop chasing the biggest parameter count and start focusing on quality, domain-specific accuracy. Otherwise, we're just exporting the same expensive, bloated systems we already rely on.
Actionable: Prioritize research into domain-specific, resource-efficient models over raw parameter scaling.
What To Do
Prioritize research into domain-specific, resource-efficient models.
Cited By
React
Get the weekly AI digest
The stories that matter, with a builder's perspective. Every Thursday.