In a recent interview with TIME, Demis Hassabis, CEO of Google DeepMind and recent Nobel Prize in Chemistry laureate, shared his insights on the rapidly advancing field of Artificial General Intelligence (AGI). Hassabis, a long-time advocate for the potential of AGI to revolutionize various aspects of society, discussed both the optimistic and pessimistic scenarios for its impact.
The Promise of AGI
Hassabis envisions a future where AGI could lead to significant breakthroughs in various fields, including medicine, energy, and materials science. “If done properly and responsibly, it will be the most beneficial technology ever invented,” he stated. Potential applications include curing diseases, developing new energy sources like fusion or advanced batteries, and creating new materials such as superconductors. Hassabis believes that AGI could help address some of the world’s most pressing issues, such as climate change, by providing technological solutions that might be more effective than collective human action alone.
Risks and Challenges
However, Hassabis also emphasized the significant risks associated with AGI development. He highlighted two primary concerns: the potential for malicious actors to misuse AGI for harmful purposes and the risk of losing control over autonomous AGI systems. To mitigate these risks, Hassabis advocates for international cooperation and standards in AGI development and deployment. “We need international standards around how these systems get built, what designs and goals we give them, and how they’re deployed and used,” he stressed.
The Evolution of Google’s Principles
When Google acquired DeepMind in 2014, Hassabis secured a commitment that Google wouldn’t use DeepMind’s technology for military purposes. However, this pledge has since been modified. Hassabis explained that the changing geopolitical landscape and the widespread availability of AI technology have led to a reevaluation of their principles. While maintaining that the benefits of their technology must substantially outweigh the risks, DeepMind now works with governments and regulated industries, focusing on areas like cyber defense and biosecurity where they can make a significant positive impact.
Timeline and Implications
Hassabis discussed the varying predictions for AGI’s arrival, emphasizing that the definition of AGI is crucial. He believes that true AGI would possess human-like cognitive capabilities, such as the ability to generate novel hypotheses like Einstein’s theory of general relativity. While some predict AGI within the next few years, Hassabis suggests it may take longer, potentially within the next decade. The implications of AGI are profound, potentially leading to significant labor market disruptions and necessitating new political philosophies to address issues of wealth distribution and purpose in a post-scarcity world.
Future Vision
Hassabis envisions a future where AGI leads to “radical abundance,” solving fundamental problems like energy scarcity through advancements in fusion, optimal batteries, or superconductors. This could transform the economic landscape, potentially moving beyond the current zero-sum game driven by resource constraints. However, he acknowledges that this would require new political and philosophical approaches to address ongoing issues like inequality and the distribution of power.
As the development of AGI continues to accelerate, Hassabis’s insights highlight both the immense potential benefits and the complex challenges that lie ahead. The path include not only technological advancements but also the need for concerted global action and innovative political and social strategies to ensure that AGI benefits humanity as a whole.
