My India Times News

In a revealing interview aired on CBS News’ 60 Minutes on April 22, Demis Hassabis, the CEO and co-founder of Google DeepMind, predicted that artificial general intelligence (AGI)—machines with human-level cognitive capabilities—could be realized within the next five to ten years. Yet, while expressing optimism about the rapid strides in AI, he offered a measured assessment: current systems still fall short in key areas, particularly imagination and originality.

Hassabis, who is also a Nobel Prize-winning researcher, clarified that although AI systems have become astonishingly advanced, they do not possess consciousness or any true sense of self-awareness. He pointed out that today’s large language models and other advanced AI systems are still fundamentally limited to what they’ve been trained on—an enormous, yet finite, corpus of human-generated information.

“You can think of them as the average of all human knowledge. That’s what they’ve been trained on. But they still struggle to formulate truly novel questions or hypotheses that haven’t been previously considered,” Hassabis explained during the interview.

AI’s Creative Limitations: Still Imitating, Not Inventing

Hassabis emphasized that AI, despite its remarkable capacity to simulate understanding, cannot yet generate genuine insights or intuitions. “They lack imagination,” he remarked, underlining one of the most profound boundaries still separating machine intelligence from the human mind. According to him, current AI systems don’t yet have the ability to ask groundbreaking questions or conceive theories outside the bounds of human precedent.

This creative gap is one of the primary reasons Hassabis urges caution when speculating about the potential of AI to achieve self-awareness or conscious thought. Instead of attempting to replicate human consciousness, he advocates for leveraging AI to build intelligent tools that assist humanity in its quest to understand the brain.

“Let’s first build useful intelligent systems and apply them to problems like decoding the brain. Only then should we even begin to think about machines with self-awareness,” Hassabis advised.

Revolutionizing Medicine: AI’s Role in Ending Disease

While maintaining a cautious stance on AGI and AI consciousness, Hassabis was unequivocally enthusiastic about the transformative potential of AI in healthcare and drug development. He expressed confidence that AI could play a pivotal role in bringing about the end of many diseases in the next decade.

Currently, the process of developing a single drug takes an average of 10 years and costs billions of dollars. According to Hassabis, AI could dramatically accelerate and reduce the costs of this process. By using advanced AI models to simulate molecular behavior and predict interactions, researchers might be able to identify promising compounds in mere weeks rather than years.

“We could shrink that timeline down from years to months—or maybe even weeks,” Hassabis said. “That’s a game-changer for human health.”

This potential is already being realized through DeepMind’s work on AlphaFold, an AI system developed under Hassabis’ leadership that can predict protein structures with high accuracy. This technology has opened up new frontiers in molecular biology and drug discovery.

Robotics: The Next Major AI Breakthrough

Another major frontier for AI, according to Hassabis, lies in robotics. He suggested that the coming years will see significant breakthroughs in this space, particularly the emergence of humanoid robots capable of performing useful tasks in the real world.

Robots today still struggle with complex physical environments and often lack the cognitive flexibility required to adapt to novel scenarios. But as AI models become more powerful and data-rich, the dream of robots that understand and interact with the world like humans is slowly becoming more plausible.

“We’re on the brink of witnessing a shift where robots begin to demonstrate real-world utility,” Hassabis predicted. “It won’t be long before we see practical demonstrations of these systems performing everyday tasks.”

A Vision Rooted in Science and Ethics

Hassabis’ vision for AI development is deeply informed by his unique academic background. With degrees from Cambridge, MIT, and Harvard, and a doctorate in neuroscience, his approach integrates computational power with a fundamental understanding of how the brain works.

His work, especially in partnership with fellow researcher John Jumper, culminated in the development of AlphaFold—a landmark AI system that revolutionized molecular biology. The duo’s contribution to science was recognized in 2024 with the Nobel Prize in Chemistry, awarded for unlocking the structure of proteins—a critical advance for understanding diseases and developing cures.

Despite these achievements, Hassabis remains grounded. He consistently emphasizes the need for responsible development, ethical AI deployment, and the importance of using AI to augment human capabilities, not replace them.

“AI is a tool, not a replacement for human thinking. It’s a partner in discovery, not the discoverer,” he remarked.

The Road to AGI: Promise, Perils, and Preparation

As AI systems edge closer to AGI, the global community faces unprecedented questions about the future of intelligence, labor, and even life itself. Hassabis’ cautious optimism serves as a guiding light in this rapidly changing landscape. While acknowledging the inevitability of AGI within a decade, he urges policymakers, scientists, and technologists to prepare thoughtfully and ethically.

He advocates for continued investment in AI safety research, public engagement, and interdisciplinary dialogue. “The question is not just when AGI arrives,” he noted, “but how we ensure it benefits everyone—and doesn’t lead us astray.”

In the end, Hassabis sees AI not as a threat, but as an extraordinary opportunity to amplify human intelligence and solve problems once thought intractable. But to truly fulfill that promise, humanity must proceed not just with ambition, but with imagination—the very trait AI has yet to master.