The Martlet

Updated Feb. 7, 2012

September 30, 2010. A wayward lunch bag shuts down the New York Stock Exchange (NYSE).

The impetus for the shutdown? The NYC’s secretly-located, multi-million dollar surveillance system that uses artificial intelligence (AI) to simultaneously monitor 3 000 video cameras. The system is capable of detecting an object’s size and shape, along with the length of time it has remained motionless, in order to alert authorities to potential threats, such as a bomb. Or a lunch bag.

What else can it do? It can identify and track an individual based on a distinctive combination of appearance and gait, rather like a technologically-constructed fingerprint. Sound a bit too Big Brother-esque?

“This can be a good thing, or a not-so-good thing,” says Dr. Alexandra Branzan Albu, an associate professor in UVic’s Department of Electrical and Chemical Engineering, and Café Scientifique’s most recent guest lecturer. “It’s not the technology that is threatening. It is the way we choose to use this technology.”

Albu spoke about AI at the latest Café Scientifique session on Jan. 24 at the Strathcona Hotel.

The idea that super-intelligent computers might be capable of developing human emotions and consciousness has been a source of great debate amongst philosophers and computer scientists alike. Think Will Smith amidst a sea of metallic soldiers in Alex Proyas’ 2004 feature film, I, Robot.

“There’s no proof for that,” says Albu. “There’s no proof for this prediction that AI will ever gain consciousness.”

AI is already fully enmeshed in our everyday lives — spam filtration and voice recognition systems being prime examples.

Albu’s research focuses on developing novel algorithms for a variety of computer vision-related applications. What exactly is an algorithm? According to Oxford’s Online Dictionary, it is a process or set of rules to be followed in calculations or other problem-solving operations, especially by a computer. AI works by mimicking human thought processes — perception, cognition and action — to analyze data, detect patterns and initiate action.

Albu’s current projects include gait analysis for identification- and pathology-related purposes, as well as the underwater surveillance system Neptune and the School of Environmental Studies’ Mountain Legacy Project, which are both aimed at understanding climate change.

“This is what excites me; to work with people from environmental studies; to work with doctors, physiotherapists, because then I feel that I can make a difference,” says Albu.

Albu pulls up a real-time video of a barren ocean floor. Bits of debris known as sea snow fall methodically across the screen. “Watch that and fall asleep,” says Albu, a sea of heads nodding in agreement.

She then pulls up a frame in which a fish darts across the screen, highlighting the value of a system that is capable of screening out activity-less frames, saving researchers countless wasted (and mind-numbing) hours of surveillance.

The project that Albu seems most passionate about is VIPERS, a software program that uses special effects to help youth grasp the concepts of coding, while having fun.

“We need more kids to embrace careers in engineering and computer science, so we are building software for people to start enjoying programming in computer vision,” says Albu.

But while kids may reap the benefits of this technological progress, AI can feel more like a threat to adults in the midst of their careers. When asked if the concerns raised by radiologists in a recent Economist article about losing their jobs to AI technologies were legitimate, Albu concedes that we may see AI systems dominating the field, reserving only the more complex images for human analysis.

“But they could always use their brains in different ways,” says Albu, who is keen to focus on the positive. “I think that AI opens up amazing opportunities for new jobs.”

“Yeah, there are always washrooms to be cleaned,” whispers a member of the audience, a retired Canada Post letter carrier.

It becomes clear from the murmurs of disappointment when Albu tries to sum up her presentation that her quiet voice isn’t the only reason that the guests are leaning in towards her, sitting on the edges of their seats — at least those lucky enough to have found a chair at the popular event — the audience is in fact, deeply engrossed.