Artificial intelligence is increasingly used to make decisions that affect our lives, from job applications to loan approvals to what content we see online. But these systems can sometimes treat people unfairly, especially those from underrepresented groups. Researchers have developed many ways to make machine learning fairer. However, these solutions are rarely used by companies, since they can reduce performance and are not yet required by most laws. Meanwhile, these systems often rely on data contributed by users themselves. This opens up a new possibility: what if the people affected by biased AI could take action collectively? In this talk, I’ll start by explaining what machine learning is and how it can potentially become unfair. I will then share initial results about how a small, coordinated minority can shift the behavior of AI systems toward fairer outcomes.

Licensed to the public under https://creativecommons.org/licenses/by/4.0/

about this event: https://cfp.cttue.de/tdf4/talk/MKSYLS/