India Bets on AI Tokens, Risks Coding Jobs

For 25 years, India's strength was software services; now, it's aiming to lead in AI tokens, the essential components for artificial intelligence agents. The Indian government is incentivizing foreign companies with a 20-year tax holiday to establish inference hubs, crucial for meeting the 24/7 demands of AI agents processing quintillions of tokens. This strategic shift raises questions about the future of India's 6 million coders.

While automation will create new roles, upskilling the workforce for complex AI guardrail development, rather than basic coding, presents a significant challenge. India's $200 billion services export industry faces a crisis as clients will value intelligence auditing over syntax. This pivot also positions India as a "virtual battery" for US hyperscalers seeking affordable power for AI computations, challenging Chinese competitors.

Companies like Alphabet, Microsoft, and Amazon are planning token-producing factories in India, with states like Andhra Pradesh offering substantial incentives. However, this may not offset the potential loss of traditional coding jobs, leaving policymakers to consider the future for millions of young people. Projections estimate India could produce 12 quadrillion tokens annually by 2030, generating billions in revenue, but this economic gain might not replace lost employment opportunities.

To secure its AI future, India needs high-end AI chips for research, potentially by negotiating terms with hyperscalers. For instance, requiring companies to contribute a percentage of their high-bandwidth memory chips to a national research cloud could benefit Indian universities. Additionally, creating pathways for AI supervisors and managers to oversee autonomous agents globally could provide new employment avenues for graduates.