In this episode of the Mechanical Ink podcast, Schalk Neethling shares his experience attending a workshop on ethical reflection in AI research reviews at MozFest 2023. The workshop explored how research reviews are conducted. Participants were grouped into six breakout rooms to discuss different example research projects. Schalk joined the group that focused on a VPN app study. The session raised critical questions about the study's funding source, assumed institutional knowledge, and data filtering practices. The group also discussed concerns about data backups, the end goal of the study, and the baseline against which results would be measured. This session highlighted the importance of a diverse review committee to ask challenging questions from different perspectives.
Schalk also touches on the Kwanele project, Bountiful features, and the AI Futures Lab, which demonstrated their work during the MozFest Trustworthy AI Working Group Demo Day.
Show notes can be found on Substack.
Additional reading
- Lies, Damn Lies, and carbon calculators
- Co-maintaining openness
- ChatGPT, Galactica, and the Progress Trap
- The complicated road to Open Source Sustainability
- CommonVoice
- Article19
- Accessibility and Green Software [podcast]
- AI Takes Over GitHub Analysis: The Future of Data Science is Here
- European Commission Funding for Open-Source Federated Cloud Platform
- Erase browser history: can AI reset the browser battle?
- You Are Not a Parrot And a chatbot is not a human.
- Signal CEO: We “1,000% won’t participate” in UK law to weaken encryption
- There is no standard’: investigation finds AI algorithms objectify women’s bodies
- RESF and B Corp: Open-Source Projects Get a New Home
Recorded and produced by Mechanical Ink.