Commenting on Canadians’ need to use more AI, the Honourable Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, stated on CBC‘s Power and Politics recently, “Technology moves at the speed of innovation, but adoption moves at the speed of trust.”
Speaking as someone who has spent the better part of the past 20 years researching AI ethics, I’ll add the following proviso, rooted in ethical inquiry: AI adoption ought to move at the speed of trustworthiness.
Indeed, I wrote about this very issue in 2018 as lead author on a G7 paper titled “Accountability in AI: Promoting Greater Social Trust”—it was Canada’s contribution to the first G7 discussion on Responsible AI.
As we wrote in our 2018 submission, chasing “trust” in AI without properly considering its “trustworthiness” puts us all at risk of over (or under) trusting the technology. It treats the trustworthiness of the technology as a foregone conclusion. A quick scan of the daily news tells us that’s not a great assumption—we have many good reasons to question the trustworthiness of the many AI-powered applications in our lives today. However, we are regularly told to trust applications that we should not, underscoring the need for greater focus on calibrating our trust.
In her fantastic post (no surprises there), Teresa Scassa highlights that there are no AI ethics experts on the AI Task Force. I suspect that’s by design. Ethics tends towards a slow thinking process, one that doesn’t always sit well among those predisposed to Agile mindsets—the consultation is a “sprint” after all. Still, the right AI ethics experts are well-positioned to help calibrate our trust, and those among us who focus on AI policy could no doubt think quick enough to accommodate, and contribute meaningfully to, even the quickest sprints.
I couldn’t agree more with Teresa Scassa that a national AI strategy, especially given today’s geopolitical reality, deserves careful thought, certainly more than a sprint. The online survey contains twenty nine (29) questions, most of which raise important questions (a few are demographic) that cover a lot of ground. That said, we’ll be doing an AI ethics sprint of our own at CRAiEDL, and plan to contribute to the process the Government of Canada has chosen. I’d encourage everyone to pick a question or two and make your voice heard.
Canada's AI Task Force Could Use Some Ethics Expertise
