Energy and chips: Ashwini Vaishnaw outlines India's five-layer AI mission
Davos, January 22
Union Minister Ashwini Vaishnaw on Wednesday described the five distinct elements that form the backbone of India's artificial intelligence mission. During the ongoing global summit, the Minister pointed out that the whole world, and especially the AI-related industry, is appreciating that India is working methodically across all five layers.
On the industry's reaction to India's progress, Vaishnaw commented, " The whole world today, and especially the AI-related industry, is appreciating the fact that India is working methodically on all five layers."
Breaking down the technical framework, the Minister said, "If we look at what AI is, AI has five elements. The first element is the application layer, that is, how we use it. The second is the model layer, the models that are created. The third is the chip layer, the semiconductor layer. The fourth is the infrastructure layer, the data centres. The fifth layer is energy."
He further added, "In the world of AI, which is the fifth industrial revolution, energy is going to be a very big factor. In this kind of situation, from energy to applications, India's methodical work has been highly appreciated by the world, and especially by the AI-related industry."
Meanwhile, Ashwini Vaishnaw has detailed India's comprehensive strategy to dominate the global artificial intelligence landscape, emphasising a shift from big-tech-controlled resources to a public-private partnership model.
Speaking at a World Economic Forum panel on the "Role of AI in Economic Growth and Global Influence," the Minister revealed that India has successfully established a public-private partnership with 38,000 GPUs as a common compute facility, accessible to students, researchers, and startups at roughly one-third the global cost, unlike many countries where big tech controls GPU access.
Addressing the critical issue of regulation, Vaishnaw advocated for a "techno-legal" approach rather than relying solely on standalone legislation. He argued that the complexities of modern technology require robust technical tools to address risks such as bias and deepfakes, including detection systems accurate enough to stand judicial scrutiny. He added that India is developing technologies to mitigate bias, enable reliable deepfake detection, and ensure proper "unlearning" before AI models are deployed.
The Minister also highlighted a strategic shift in the economics of the Fifth Industrial Revolution, suggesting that the massive ROI of the future will come from cost-effective, scalable solutions rather than just "brute-force" computing. He debunked the myth that all AI progress requires expensive hardware, noting that "nearly 95 per cent of AI work can be done using the 20-50 billion parameter models."
— ANI
Reader Comments
Finally, a plan that goes beyond just apps! The emphasis on the semiconductor layer and energy is crucial. We can't build AI sovereignty without our own chips and reliable power. Hope the execution matches the vision.
The 'techno-legal' approach for regulation sounds smart. Just making laws won't stop deepfakes; we need the tech to detect them. But my question is about data privacy - how are we ensuring the data used to train these models is ethically sourced?
As someone working in tech, the point about 95% of AI work not needing massive models is so true. It's about smart, efficient solutions, not just throwing money at compute. If India can master cost-effective scalability, it will be a global leader.
Appreciate the holistic framework, but the proof will be in implementation. We have great plans on paper. Need to see this compute facility actually being accessible to small-town engineers and students, not just the usual IIT/IIM crowd. Hope it's truly inclusive.
The energy layer mention is critical and often overlooked. AI data centers consume massive power. India's push for renewable energy aligns perfectly with this need. A sustainable AI ecosystem could be our unique global advantage.
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.