RUBI the robot tutor interacting with children in San Diego. RUBI is an acronym for Robot Using Bayesian Inference.
Credit: University of California at San Diego
SYDNEY: Can we really create a computer which can discern complex patterns, recognise facial expressions, decode meaning from speech and even understand our emotions? It's only a matter of time, says a leading researcher.
That’s the vision that drives Terrence Sejnowski, a leading authority on computational neuroscience who is trying to understand the inner-workings of the human brain – how we learn, form memories and interact with our environment – in order to develop computer devices that can emulate these processes.
“We call this reverse engineering,” said Sejnowski, who currently holds the Francis Crick Chair at the Salk Institute for Biological Studies near San Diego, California. “The goal is to get computers to communicate with humans – we want them to talk to us, to understand our emotional state and to communicate with us on an emotional level.”
Daunting challenge
On March 9, Sejnowski will explore this daunting challenge at length when he delivers the 2011 Graeme Clark Oration in Melbourne, an annual event organised by the ICT for Life Sciences Forum – a collaborative network of Melbourne-based researchers interested in the convergence between biology, computing and engineering.
“He has brought computer and computational sciences closer to neural science than almost anyone else on the planet,” said Iven Mareels, dean of the Melbourne School of Engineering at the University of Melbourne.
“His research has implications for medicine and engineering,” he added. “We all want machines that can assist us at a higher and higher level of abstraction, and machine-human interfaces that are more natural.”
Parallel architecture
Sejnowski will detail what researchers need to understand in order to create the next-generation computers based on the parallel architecture of the human brain. He will discuss how the brain behaves in disorders such as autism and schizophrenia, and the rise of social robots – one of which, named RUBI, is already serving as a teacher’s aide in a preschool classroom in San Diego.
“It’s hard to predict the future, but it won’t be too long before the computer in front of you will be a social robot,” he told Cosmos Online “You’ll be able to talk to it, flirt with it, or even get angry and yell at it – and it will understand you and your emotions,” he said, adding that he and his colleagues are developing algorithms that will improve a robot’s ability to learn gradually by practice – something that humans are quite good at.
Sejnowski’s multidisciplinary research approach is part of the reason he was invited to speak at the event, which honours Graeme Clark, the Australian scientist responsible for developing the world’s first bionic ear – a multi-channel cochlear implant that has helped restore hearing to more than 200,000 people across the globe.
The next big thing
Clark gave the inaugural oration in 2008 and was succeeded by genomics pioneer Craig Venter in 2010. “The idea is to get a speaker who has done something important, or who we feel is on the cusp of achieving something great,” said event organiser and ICT for Life Sciences spokesperson Luan Ismahil. “He’s considered by many to be the next big thing in neuroscience.”
THE 2011 GRAEME CLARK ORATION
• Who: Terrence Sejnowski, a leading authority on computational neuroscience who currently holds the Francis Crick Chair at the Salk Institute for Biological Studies
• What:The 2011 Graeme Clark Oration – an annual event in honour of Melbourne-based researcher Graeme Clark who developed the first multi-channel cochlear implant in the mid-1970s.
• Where:Melbourne Convention and Exhibition Centre
• When: 6.15 pm on Wednesday 9 March 2011
