[The 84th TrustML Young Scientist Seminar] Talk by Simo Hosio (University of Oulu) "Rethinking Data Origins: Crowdsourcing for Ethical and Human-Centric AI Solutions"

Date and Time:
December 3, 2024: 1:30 pm - 2:30 pm (JST)
Venue: Online and Meeting Room 1 at the RIKEN AIP Nihonbashi office*
*Meeting Room 1 is available to AIP researchers only

Title:
Rethinking Data Origins: Crowdsourcing for Ethical and Human-Centric AI Solutions

Speaker:
Simo Hosio (Associate Professor at the University of Oulu)

Abstract:
Machine learning is embedded in our daily lives. Yet, the general public neither understands (nor wants to consider) the origins of training data. Ethical and economic questions drive an often-overlooked discussion in ML research: fairness, transparency, and ethical practices are essential, but issues like worker well-being in data annotation ot other potential uses for their labour bandwith are rarely emphasized. In my talk, I explore crowdsourcing platforms not only as sources of training data for the ML community but also as foundational elements in AI-driven well-being solutions, such as healthcare applications. I consider both AI-driven and traditional database-driven solutions. I will discuss our past projects and highlight promising future research and collaboration opportunities at the intersection of HCI, ML, and human computation.

Bio:
Simo Hosio is an Associate Professor at the University of Oulu where he leads the Crowd Computing Research Group. He is currently a Visiting Researchers at the University of Tokyo in the Interactive Intelligent Systems Laboratory. His background is in Ubiquitous Computing and HCI, with keen interest in human-centric thinking and crowdsourcing. (website)