Google DeepMind CEO Warns AGI Could Arrive by 2035: 'Society Isn't Ready'

AGI Closer Than Expected

Demis Hassabis, CEO of Google DeepMind, recently revealed that Artificial General Intelligence (AGI) could be achieved within the next 5 to 10 years, potentially by 2035. AGI refers to AI systems that possess human-level intelligence, capable of thinking, reasoning, and understanding the world in nuanced and deep ways.

The Warning About Readiness

In an interview with Time, Hassabis expressed concerns that society is not prepared for the advent of AGI. He emphasized the critical need for international collaboration among governments, companies, and research labs to ensure that AGI development is safe and controllable. DeepMind’s research has also identified AGI as a possible existential threat, warning that mishandling could lead to catastrophic consequences for humanity.

Potential Benefits of AGI

Despite these risks, Hassabis highlighted the tremendous promise AGI holds, especially in healthcare. He stated that AGI could potentially eradicate diseases within the next decade, referencing recent breakthroughs in protein folding and drug discovery made possible by AI technologies. Unlike current AI, AGI would not only be more intelligent but would also possess capabilities equivalent to human creativity, logic, and intuition at digital speeds.

Proposed Solutions for Safe Development

To manage the risks and harness the benefits, Hassabis proposed establishing an international research hub akin to CERN, alongside an oversight body similar to the International Atomic Energy Agency (IAEA), to monitor AGI development globally.

The race towards AGI is accelerating rapidly. While the potential rewards are immense, including possibly ending disease, the risks to humanity are also significant, underscoring the urgency to prepare and regulate accordingly.