How a new foundation wants to help non-profits lead the conversation on responsible AI
Why It Matters
Technology companies are not likely to prioritize responsible development and dissemination of artificial intelligence, and government regulation in Canada has been slow to catch up to this fast-paced sector.
A new Toronto-based foundation is aiming to help more Canadian non-profits benefit from artificial intelligence.
Through government funding, the Human Feedback Foundation hopes to subsidize training opportunities for non-profit leaders, staff and board members.
Executive director Elena Yunusov previously led the marketing function at RBC Borealis, the bank’s internal AI and data innovation arm.
The non-profit sector “has incentives aligned in a really unique way for AI to be adopted responsibly,” given that the financial bottom line is not the sector’s primary priority, she said.
“Responsible AI won’t come naturally from the industry, and it won’t come naturally from government – only because government will take a lot of time to get it right,” Yunusov said.
“What do we have left then? The non-profit sector.”
Yunusov had been working in artificial intelligence at RBC long before ChatGPT burst onto the scene. In the wake of OpenAI’s chatbot, she observed many “red flags”: large technology companies wrestling with their own ethics teams, a rush to commercialize, and unclear government regulation, she said.
“The competitive environment doesn’t really favour spending a year or two on model validation and testing,” Yunusov added.
Yunusov also pointed out the rise of so-called synthetic data, which trains artificial intelligence models. This is data that is not real, but reflects reality, and is used to test and train models.
Despite not working in the non-profit sector before, Yunusov launched the Human Feedback Foundation under the Linux Foundation.
“The dream for the Foundation is to be able to fund open-source projects that prioritize human-centric AI,” over private models, she said.
The Linux Foundation does not, however, support the Human Feedback Foundation financially. To develop the fund, Yunusov is currently applying for funding through the Digital Supercluster, a research and development fund dedicated to digital transformation, partly funded by the federal government’s Ministry of Innovation, Science and Economic Development.
One of the streams of funding that the Digital Supercluster is focused on is commercializing AI, having recently announced substantial investments in AI solutions for agriculture, mining and healthcare.
Yunusov hopes this funding will open up access to upskilling opportunities for non-profit staff and board members through subsidizing costs, as well as helping them embed AI into their operations through a dedicated adoption framework.
She has also been building a community of engineers and developers nationwide, without whom responsible AI is not possible, she said.
“We’ve seen all kinds of ‘AI for good’ initiatives – and they tend to be initiatives,” she said. “There is not a system-wide change that I’m seeing.”
Tell us this made you smarter | Contact us | Report error
