India, March 2 -- In January 2025, the AI landscape witnessed a seismic shift with the emergence of DeepSeek R1, a large language model (LLM) developed in China.

Unlike previous AI breakthroughs, DeepSeek R1 did not stun the world with superior capabilities but rather with its resource efficiency.

While leading models from OpenAI, Google, and Meta relied on tens of thousands of GPUs for training, DeepSeek R1 reportedly required only a few thousand, challenging the long-standing belief that compute power is the ultimate determinant of AI success.

DeepSeek's success raises an important question: Can India develop a similar AI disruptor?

To answer this, we must examine the four key factors contributing to AI success: innovation, compute ...