Did AI Do That? Mitigating the potential risks to user expression online
This highly interactive session will focus on what we can do, from a content governance perspective, to mitigate the risks AI and ML technologies may pose to user expression online. The workshop will have 4 discussion leads: Anima Anandkumar (NVDIA) , Thiago Oliva (OSB Policy), Nic Suzor (QUT, Oversight Board Member), Jillian York (EFF). The discussion leads will introduce themselves and key question that they want to explore in breakout groups. Each breakout group will report back at the end of the session.
Potential breakout group questions include:
How much should we expect of AI to understand local context, nuance and systemic inequality?
Can AI help develop effective content moderation or appeals systems that work at scale?
How can platforms provide better explanations to users about why their content has been flagged automatically? What would you like to see to understand the confidence and basis of predictions?