At the India AI Summit in New Delhi, Dario Amodei warned that artificial intelligence could surpass most human cognitive abilities within a “small number of years.”
Describing a “country of geniuses in a data center,” the Anthropic chief outlined rapid AI growth, safety partnerships with India, and economic disruption ahead.
In his address, Amodei highlighted the explosive progress in AI over the past decade, describing it as an exponential curve akin to “Moore’s Law for intelligence.”
He stated that society is now well advanced along this trajectory, with only a small number of years remaining before AI models surpass the cognitive capabilities of most humans for most things.
Key quotes and points from the talk include:
- “We are now well advanced on that curve and there are only a small number of years for AI models to surpass the cognitive capabilities of most humans for most things.”
- AI is approaching the equivalent of “a country of geniuses in a data center“—a set of AI agents more capable than most humans at most tasks, able to coordinate at superhuman speed.
- This unprecedented capability level could enable breakthroughs like curing long-incurable diseases and lifting billions out of poverty.
- However, it also poses massive challenges, including risks of misuse (e.g., in biosecurity or weapons), economic disruption, and the need for responsible development.
Amodei praised India’s energy, ambition, and technical potential in AI, suggesting the country could play a central role in realizing benefits (potentially driving up to 25% economic growth) while addressing global risks through safety-focused partnerships. Anthropic is expanding operations in India (e.g., Bengaluru) to collaborate with local builders and enterprises.
He emphasized the urgency of commercial applications alongside technical advances, noting the palpable collaborative energy in India unlike anywhere else.
The keynote drew attention online, including some mixed reactions to Amodei delivering parts of it while reading from his phone. Overall, it underscores Anthropic’s focus on both AI’s transformative upside and the critical need for safeguards amid accelerating timelines to superhuman-level systems.
