The world today is at a threshold moment where AGI, Artificial General Intelligence, is on the horizon, maybe in the next five to eight years, believes Demis Hassabis, CEO, Google DeepMind. And this is what makes the IndiaAI summit more important as we start seeing more autonomous and genetic AI systems much more capable, he says.
For context, AGI, Hassabis says, is about creating a system that can exhibit all the cognitive capabilities humans possess, including creativity and long term planning.
Today’s AI models, says Hassabis, are very impressive, but they still have many flaws and things they cannot do, including consistency across different tasks that you would want from a general system. “When I look at current systems and what is missing for them to be considered general intelligence, I would point to continual learning. Today’s systems are trained, and then they are essentially frozen and put out into the world. What you would want instead is for those systems to continue learning from experience, to adapt to the context they are in, and perhaps personalize themselves to the tasks you have for them,” he adds.
They also struggle with long term coherent planning. They can plan in the short term, but not over the kind of extended horizons that humans can, sometimes spanning years, he argues.
AI for one, will surely, open a lot of horizons, he argues. “I think we can revolutionize drug discovery and human health and also areas like material science and climate change with the help of AI. But, of course, it also comes with many risks, too. AI is a dual-purpose technology,” he adds. The biggest risk that world over all countries are worried about, is cyber security.
3 February 2026
Get the latest issue of Outlook Business
Experts point out that cyber criminals are learning technology quickly, and using deepfakes. Cybersecurity incidents in India rose from 10.29 lakh in 2022 to 22.68 lakh in 2024. Hassabis says that to mitigate some of the risks, it's going to need international cooperation. In the near term, biosecurity and cybersecurity are key concerns. Current systems are already quite capable in cyber domains. It is crucial to ensure that defensive applications of AI remain stronger than offensive ones, he adds.
“International collaboration is essential because AI is a digital technology that cannot be contained by borders. Open source development offers significant benefits but also raises questions. If vulnerabilities are discovered in open-source systems, recalling or patching them can be challenging. These are new kinds of issues that require global coordination,” he adds.
