Ramin Shadmehr (PhD Candidate)

A valuable and important framework "1. We must be actively critical to ensure AI tools protect and respect the complexities of digital privacy and don’t compromise it to deliver “optimization.” 2. We must be able to opt-in and opt-out, or at least understand how our use of systems contributes to our own surveillance and manipulation. 3. We must speak up to make sure AI addresses the voices of those on the edges, for whom the “typical” will never work. Without intervention, we risk AI creating a kind of amplified, fractal echo-chamber, advancing homogenization and being capable of enacting more discrimination at a faster rate. 4. We must build in protections for those who measure beyond a certain standard deviation for an “average” assumed through data that perpetuates statistical discrimination. 5. We must proceed with eyes wide open for the human and environmental costs of using and growing AIs and the impact on climate change (another complex problem that disproportionately impacts those who already face discrimination and inequities). 6. We must be able to follow the results of AI to a source so that we can avoid and track its own mistakes, its hallucinations. 7. We must maintain the “human in the loop” if we are to have difficult conversations that move us beyond a sanitized AI output of censorship. 8. We must address the failure of AIs to avoid biased, stereotyped output—an upstream data problem." #genai #inclusive #accessible #ai

10

1 Comment

So much potential for modern AI. So much to still understand and best practices to establish. Universities have a vital role to play.This is just the start from McGill’s Collaborative for AI & Society (CAIS)! https://lnkd.in/eTfxWkWi

27

This week I got to do one of my favourite exercises with my students here at St. Lawrence College. We use design thinking to solve some pretty big, student-focused, issues. Here are some of the challenges I gave to my students and their solutions: • A noisy and often garbage-littered cafeteria: Sound absorbing panels, smaller tables, more garbage receptacles, department challenges to gamify keeping it tidy, fun posters as a reminder • Increasing food costs: An area to collect uneaten food, a student-focused app that shows where students can get discounts, drop in prices for uneaten food at the cafeteria and Larry's •College way-finding: Redo the maps to make them easier to understand, a "Waze-style" map app to help students find classrooms and services, lines on the floor or coloured dots on the classroom door numbers to identify wings Other random ideas were to get students and faculties to "adopt an area" to keep clean. Students would get to paint or otherwise take ownership of the garbage and recycle bins. Real-time AI generated updates sent to college-wide app to keep students informed about services etc. including wait times, events, etc. #designthinking #AI #college #problemsolving School of Business at St. Lawrence College

5