Anupam Mittal, the founder of Shaadi.com and a prominent judge on Shark Tank India, has sparked an important conversation about India's approach to artificial intelligence. His perspective challenges the prevalent notion that India must compete with global tech giants in developing large language models and foundational AI systems.
The Model Race Versus Application Race
The "model race" refers to the competition among tech companies and nations to build increasingly sophisticated foundational AI models like GPT-4, Gemini, or Claude. These models require enormous computational resources, billions of dollars in investment, and access to massive data centers. Companies like OpenAI, Google, and Anthropic have spent years and substantial capital developing these systems.
Mittal's argument centers on a practical reality: India may not have the same financial resources or infrastructure as Silicon Valley or Chinese tech giants to compete in this capital-intensive race. However, this doesn't mean India should sit on the sidelines of the AI revolution.
India's Real Competitive Advantage
The entrepreneur's perspective highlights where India can genuinely excel in the AI ecosystem. The country has long been recognized for its software development talent, problem-solving capabilities, and cost-effective innovation. Rather than pouring resources into building models that may never match those created by well-funded American or Chinese companies, India can leverage existing models to create applications that solve real problems.
This approach recognizes that the true value in AI often lies not in the model itself, but in how it's applied. Just as India became a software services powerhouse without manufacturing most computer hardware, it can become an AI application leader without building every foundational model.
Building on Existing Infrastructure
The accessibility of AI models has democratized the technology landscape. Open-source models and API access to commercial models mean that developers anywhere can build sophisticated applications without creating the underlying AI from scratch. Indian startups and enterprises can use models from OpenAI, Meta's Llama, or other providers as building blocks for solutions tailored to Indian problems.
Consider sectors where India faces unique challenges: agriculture efficiency, vernacular language processing, financial inclusion, healthcare access in rural areas, and education at scale. These problems require contextual understanding and localized solutions that global tech companies may not prioritize.
The Economic Reality
Building and training large language models requires infrastructure that costs hundreds of millions of dollars. The computational power needed for training runs can consume as much electricity as a small city. For a developing economy balancing multiple priorities, directing resources toward application development rather than foundational research may yield better returns.
Indian developers can create AI-powered tools for local businesses, develop vernacular AI assistants, build diagnostic tools for healthcare, or create educational platforms that work in low-bandwidth environments. These applications can generate revenue, create jobs, and solve pressing problems without requiring the massive capital expenditure of model development.
The Talent Advantage
India produces hundreds of thousands of engineering graduates annually and has a proven track record in software development. This talent pool is well-suited for application development, which requires understanding user needs, system integration, and practical problem-solving rather than the specialized research expertise needed for foundational model development.
Many Indian professionals already work at leading AI companies globally, demonstrating the country's capability in this space. Channeling this talent toward building applications rather than competing in the model race allows India to play to its strengths.
Creating an AI Ecosystem
Mittal's vision suggests building a vibrant ecosystem of AI applications, startups, and services. This approach can create more immediate economic value and employment opportunities than a long-term bet on competing with tech giants in model development. As companies build applications, they create data, expertise, and infrastructure that strengthen India's position in the global AI landscape.
The strategy doesn't mean abandoning research entirely but rather prioritizing practical applications while maintaining research efforts at academic and government institutions. This balanced approach allows India to contribute to AI advancement while ensuring the technology delivers tangible benefits to its population.
India's path in the AI revolution may look different from that of the United States or China, but different doesn't mean inferior. By focusing on applications, solving local problems, and leveraging existing technologies creatively, India can carve out a significant and valuable role in the global AI ecosystem without winning the expensive and resource-intensive model race.
