Google DeepMind CEO Demis Hassabis had appeared to hint at simulation theory, but he now says that the universe might not necessarily be simulated.
Demis Hassabis has hinted at a reality that may not be a contrived simulation as popularized by philosopher Nick Bostrom, but one deeply rooted in computation and information theory. His insights, stemming from the groundbreaking achievements of DeepMind’s AI systems, like AlphaGo and AlphaFold, offer a compelling glimpse into a potentially computable universe.
“Are we in a simulation?” Hassabis was asked in an interview. “Not in the way that Nick Bostrom and people talk about,” he replied. “I don’t think this is some kind of game, even though I wrote a lot of games. But I do think that ultimately underlying physics is information theory,” he added.
“So, I do think we’re in a computational universe, but it’s not just a straightforward simulation. The fact that these systems are able to model real structures in nature is quite interesting and telling. And I’ve been thinking a lot about our work we’ve done with AlphaGo and AlphaFold in these types of systems.”
“Maybe at some point I’ll write up a scientific paper about what I think that really means in terms of what’s actually going on here in reality.”
Earlier this month, Hassabis had posted “Nature to simulation at the press of a button, does make you wonder…” on X while sharing how Google’s latest AI model could create computer code to represent everyday objects, like trees and spider webs. It had appeared then that Hassabis had been hinting at simulation theory, which says that the universe is a giant simulation, much like how we simulate different worlds in computer games.
But it seems that Hassabis’ position is more nuanced. He avoids the simplistic notion of a “simulation”, while proposing something far more fundamental: a universe whose very fabric is woven from information and computation. This resonates with a growing body of scientific thought that sees information as a primary building block of reality, alongside matter and energy. This perspective is intriguing, especially considering DeepMind’s successes. AlphaGo mastered the game of Go through complex algorithms, demonstrating the power of computation to achieve human-level expertise, while AlphaFold revolutionized protein structure prediction, accurately modeling these complex molecules based on computational analysis. These achievements, Hassabis suggests, hint at a deeper connection between computation and the underlying structure of reality. And one would look forward to the paper he talks about writing — given his background in computer games, biology, AI, and a Nobel Prize to boot, his perspective on the nature of reality would carry more gravitas than most.