ONLINE ASSEMBLY 3: Rejecting discriminatory algorithms: Is Unbiased AI Possible? 

Who gets to design the systems that increasingly govern us, and what would it mean to design them differently?

AI is already making decisions about our lives: who gets an interview, who receives welfare, who is considered “high risk”, what we see online, and who gets access to public space. From AI-powered job interviews to algorithmic welfare systems and racist chatbots, supposedly neutral technologies can reproduce – and intensify – existing inequalities.

But how do algorithms actually work? Where does algorithmic bias come from? And can we build technologies that don’t reproduce the systems of power that already exist?

Building on case studies explored through our in-person assemblies, we will investigate how algorithmic decision-making is already affecting our rights and everyday lives, and explore how an intersectional feminist ai may be possible.

EXPERTS

1. Technical / critical AI expert

Someone who can demystify how algorithmic systems work, explain where bias enters the process, and challenge the idea that algorithms are inherently objective.

• Cathy o’neil (emailed 04/09)
• Someone from data tank (already in contact)
• https://aiforensics.org/
• AlgorithmWatch
• Raphaële Xenidis

2. Lived experience / public-facing voice

Someone who can make the issue tangible through concrete examples – particularly AI hiring/interview systems, automated decision-making at work, welfare algorithms or other experiences of algorithmic discrimination. Particularly looking for AI comedic influencers who could open with their videos etc & help us advertise.

3. Feminist technology / open-source AI practitioner

Someone working at the intersection of feminism, technology and open-source AI who can offer alternative models for how these systems could be designed, governed and owned.

• Digital Fems
• EDRi
• Feminist AI
• https://radicaldata.org/