What Now? AI, Episode 2: Safe and Accountable 

As AI becomes more integrated into our lives, how do we make sure it aligns with our values while minimizing risks?

In the second episode of What Now? AI, hosts Beth Coleman and Rahul Krishnan are joined by University of Toronto experts Gillian Hadfield and Roger Grosse as they tackle critical questions surrounding AI safety, regulation and alignment. 

Listen to episode two on Apple, Spotify, SoundCloud, iHeartRadio and Amazon. Watch episode two on YouTube. 

Grosse, an associate professor of computer science in the Faculty of Arts & Science and a founding member of the Vector Institute, joined the technical staff on the alignment team at Anthropic, an AI safety and research company based in San Francisco, during a sabbatical last year.

He calls working on AI research and systems while investigating safety a “difficult needle to thread.” 

“As you move up the ladder of different AI capabilities, new requirements start kicking in – in terms of keeping the models secure from bad actors and being able to make sure they won’t intentionally carry out harmful plans,” says Grosse, a faculty affiliate at the Schwartz Reisman Institute for Technology and Society.

Hadfield, a professor of law and strategic management in the Faculty of Law and the inaugural Schwartz Reisman Chair in Technology and Society, has proposed a national registry for large AI models. She thinks companies should disclose to governments what they’re building, the data being used and the AI model’s capabilities. 

“This is a unique moment in human history,” says Hadfield, who holds a CIFAR AI Chair at the Vector Institute for AI and served as a senior policy adviser to OpenAI from 2018 to 2023. “I think this is the first time that you have such a powerful technology that is being developed almost exclusively within private technology companies, so the public and the academic sector don’t have full visibility into how the technology is working.” 

About the hosts: 

Beth Coleman is an associate professor at U of T Mississauga’s Institute of Communication, Culture, Information and Technology and the Faculty of Information. She is also a research lead on AI policy and praxis at the Schwartz Reisman Institute for Technology and Society. Coleman authored Reality Was Whatever Happened: Octavia Butler AI and Other Possible Worlds using art and generative AI. 

Rahul Krishnan is an assistant professor in U of T’s department of computer science in the Faculty of Arts & Science and department of laboratory medicine and pathobiology in the Temerty Faculty of Medicine. He is a Canada CIFAR Chair at the Vector Institute, a faculty affiliate at the Schwartz Reisman Institute for Technology and Society and a faculty member at the Temerty Centre for AI Research and Education in Medicine (T-CAIREM).