On November 3, a symposium event titled “Reality was Whatever Happened: Octavia Butler AI and Other Possible Worlds,” organized by U of T Associate Professor Beth Coleman, dived into the intersection of human cognition and learning and artificial intelligence (AI). Distinguished speakers discussed the multifaceted nature of reality and potential AI perils.
One conversation in the event titled “Temporal Relativism” explored the essence of time with Lee Smolin, a theoretical physicist and co-founder of the Perimeter Institute for Theoretical Physics, who was interviewed by Maurice Jones, a curator and AI regulation researcher at Concordia University.
Unravelling time: A physicist’s pursuit
Lee Smolin believes that time exists at a level encapsulating the most basic building blocks of our universe, the fundamental particles, and the fundamental forces governing them.
To Smolin, exploring the fundamental nature of time has been a deeply ingrained endeavour since the age of 17. His initial inspiration stemmed from an essay by Albert Einstein, wherein the sage physicist posited that understanding quantum gravity — a theory that aims to describe gravity at the ‘quantum level,’ as in on the smallest scale, and to provide a framework for understanding the nature of space and time — entails a dual challenge.
The first challenge involves comprehending the intricacies of quantum mechanics: the branch of physics exploring the behaviour of matter and light by establishing the properties of atoms, molecules, and their fundamental particles. The second is integrating these intricacies with general relativity: the theory describing gravitational force as the curvature of spacetime by objects.
The essence of time
Smolin offers two interconnected definitions of time. His perspective describes time as a fundamental, ubiquitous, and ever-evolving phenomenon, and defines time as something that generates novel events and contains, what Smolin describes as a “thick present.”
Smolin uses the term ‘thick present’ to describe how only present events have the potential to generate future events. These events are connected — as Smolin puts it, the past has exhausted its potential for generating future events.
Once an event has exhausted its potential to generate new events, it becomes the past. Essentially, Smolin reasons that while the past is determined, as it has already happened, the future is actually not yet determined. The past doesn’t completely determine the future; rather, there is room for novel events to be generated in the thick present.
The machines of the future
Regarding the rise of artificial intelligence, Smolin advocates a different approach from anyone who puts emphasis on creating machines that attempt to faithfully predict what will happen in the future. Smolin suggests constructing machines that help us build “a future that we’ve never imagined before.”
Smolin draws an analogy with how babies interact with the world. Babies, when encountering people, don’t attempt to predict who they will meet, he says. Instead, they engage in a sequential manner, asking questions like “Who is that?” after meeting each person. Smolin suggests that this kind of iterative approach should guide our endeavours in building the AI machines of the future.