Last Week in AI Bias in Twitter & Zoom, LAPD Facial Recognition, GPT-3 Exclusivity
Oct 1, 2020
Dive into the heated discussion on the algorithmic bias issues plaguing Twitter and Zoom. Explore the controversial use of facial recognition by the LAPD, revealing surprising stats. Hear skepticism from AI experts about trusting technology in healthcare, especially post-COVID. Delve into the debate on introducing robotic companions in UK care homes – a potential remedy for loneliness, but with a need for human connection. Plus, learn about OpenAI's exclusive deal with Microsoft regarding GPT-3 and its implications for the future of AI.
AI Snips
Chapters
Transcript
Episode notes
Algorithmic Bias Issues on Twitter and Zoom
- Twitter's image-cropping algorithm was criticized for often excluding Black people.
- Zoom's virtual background feature similarly removed a Black professor's head.
User Experience in AI Design
- Twitter's cropping algorithm issue highlights problems beyond AI bias, including user experience.
- Giving users more control over AI-driven features could improve user experience.
LAPD's Use of Facial Recognition
- The LAPD used facial recognition software 30,000 times in the past decade, despite prior denials.
- This lack of transparency raises concerns about civil liberties and informed public discourse.