DeepMind executive says AI can democratize global scientific research

Yesterday 10:50

By: Dakir Madiha

DeepMind executive says AI can democratize global scientific research

Artificial intelligence has the potential to decode the “recipe book of life” hidden within human DNA and position India as a global hub for advanced scientific research, according to Pushmeet Kohli, vice president of research at Google DeepMind.

In an interview with Moneycontrol, Kohli said AI’s greatest promise lies in its ability to democratize science by allowing researchers with limited resources to compete with the world’s most heavily funded laboratories. He pointed to AlphaFold, the Nobel Prize winning protein structure prediction system developed by DeepMind, as evidence of that shift. Tasks that once required years of laboratory work and millions of dollars in funding can now be completed in seconds with a single query, he said.

Kohli described the decoding of the human genome as the next major frontier for AI in biology. While scientists have long been able to read DNA sequences, understanding the full implications of those sequences remains incomplete. He compared proteins to the building blocks of life and said the next step is to interpret the genome itself, which he called the instruction manual governing human development, disease susceptibility and potential treatments.

The long term ambition, he said, is to determine how every possible change in genetic code could influence human health, paving the way for more precise diagnostics and targeted therapies. In January, DeepMind introduced AlphaGenome in the journal Nature, an AI model designed to analyze DNA sequences within the non coding genome, which makes up about 98 percent of human DNA. Researchers believe that this vast portion of the genome, once considered “junk DNA,” plays a critical regulatory role.

Kohli also argued that India is well positioned to lead high level AI research without requiring its top scientists to move abroad. He encouraged Indian researchers to apply AI to pressing domestic challenges in healthcare, agriculture and language technologies, areas where scalable solutions could deliver wide societal impact.

Google DeepMind’s laboratory in Bengaluru contributes to the development of Gemini, the company’s foundational AI model, while also building India specific applications. Agriculture has emerged as a key priority. Teams are working to organize large volumes of Indian agricultural data to create unified landscape models that could support planners and farmers in making informed decisions about crop management and resource allocation.

AlphaFold has already been adopted by more than 3.3 million researchers worldwide and cited in over 40,000 academic papers, according to Google DeepMind. Kohli said this broad uptake illustrates how AI tools can level the playing field for scientists in developing countries, enabling access to advanced computational capabilities without the need for extensive laboratory infrastructure.

India is scheduled to host the AI Impact Summit on February 19 and 20 in New Delhi, marking the first global AI summit to take place in countries of the Global South. Kohli said such initiatives could further strengthen collaboration and accelerate the application of artificial intelligence in science and public policy.