A PhD student in the Faculty of Engineering, Ibtihel Amara is a bridge builder. But not the kind of bridge one usually associates with engineers. Affiliated with the Department of Electrical and Computer Engineering, and the Centre for Intelligent Machines (CIM), Amara is looking to bridge the “trust gap” between humans and intelligent machines.
“It is difficult for us humans to blindly agree with and depend upon decisions without having explanations, reasons, or a measurable degree of transparency of the conclusions given by an AI system,” says Amara.
Amara’s work has garnered the interest – and support – of Borealis AI, a Royal Bank of Canada research institute focusing on state-of-the-art and ethical AI. Amara was recently named one of 10 recipients of the recipients of the 2019-2020 Borealis AI Graduate Fellowship Awards.
In its announcement Borealis AI said the Fellowships provide financial support to “exceptional domestic and international graduate students to carry out fundamental research as they pursue their masters and PhDs in various fields of AI.”
Amara is working on ways to allow modern AI systems to “express their level of uncertainty, which can assist humans to evaluate the degree to which these outputs should be ‘trusted.’”
“AI systems are known to provide competent, accurate predictions and decisions,” she says, “but these outputs should offer a ‘guidance’ measure for humans to further assess the degree of precision and whether to use these according to the given situation.”
Amara’s work addressing uncertainty and integrating reliability and trust into modern AI systems is critical in applications like medical diagnosis, self-driving cars and even financial banking applications.
“AI can offer people a better quality of life through quicker, more informed decisions leading to greater productivity… every human can benefit from these systems,” she says.
From flights of fancy to field of study
The seeds of Amara’s fascination in AI were sown in her youth.
“During my childhood, science fiction movies and cartoons fascinated me most – especially ones about intricate gadgets or human-like robots that provided decisions and recommendations to save the world,” she says. “When one of my elementary teachers asked me what I would be when I grew up, I answered with confidence that one day I want to be a scientist constructing those intelligent machines”
At McGill, and at CIM in particular, Amara has found like-minded people, who share her love of “constructing those intelligent machines.”
“My main mentor is my research supervisor, Professor James J. Clark. He is like my personal ‘Jedi master’ who helps me in providing insights about the AI field and especially sculpting my research skills,” she says. “Being a part of the CIM community, and exchanging ideas with faculty members, researchers and students has been instrumentals in improving my research performance, my problem-solving expertise and social skills.”
“Ms Amara is perpetually enthusiastic and positive, and raises the spirits of everyone she meets,” says Clark. “In addition to being an excellent researcher, she contributes enormously to the CIM community. For example, for the past two years she served as the emcee for the annual CIM Student Research Showcase event. She also was key in arranging a weekly evening social event for graduate students in the centre, sharing baked goods and board games… She is a wonderful person.”
This article was originally published in the McGill Reporter.