Waterloo-led team creates virtual brain

“Spaun” learns patterns and uses new knowledge


Spaun learning (CTNWaterloo/YouTube)

A team led by a scientist at the University of Waterloo has created the world’s largest functional brain model and it’s the first to cross “the brain-behaviour gap,” Prof. Chris Eliasmith tells Canada.com. “Spaun” can learn patterns and then uses that new knowledge to answer questions. The team’s research was published Thursday in Science. Tenille Bonoguore of Waterloo communications explains more:

Spaun uses only its brain to understand images it sees, and “move” its virtual arm. In short, it uses the same processes that your brain does to pick up a cup of coffee.

“It has been interesting to see the reactions people have had to Spaun. Even seasoned academics have not seen brain models that actually perform so many tasks. Models are typically small, and focus on one function,” says Eliasmith, who is jointly appointed to the departments of philosophy and systems design engineering, and cross-appointed to computer science at the University of Waterloo.

Currently, other large-scale brain models (over a million neurons or so) do one thing—mimic lots of neurons. Spaun does that too (it has 2.5 million neurons), but more importantly, it is constructed so those neurons actually think about patterns it encounters and make things happen in its environment (and you can watch videos of it working).

Chris Eliasmith’s book “How To Build A Brain” will be published this fall by Oxford University Press.


Sign in to comment.