Unintended consequences of technology: Enabling NextGen discrimination
summary
Under concrete examples, big companies’ systems reproduce and amplify bias: Amazon trains an AI resume screener on decade-old male-dominated CVs and scrapped the tool after it systematically downgraded women; Microsoft’s chatbot Tay learned slurs in 24 hours from users; Zoom’s face-detection algorithm erases racialized faces on video; a grainy Michigan surveillance match produces a false positive wrongful arrest; workforce stats show only ~18% of top AI conference authors are women and corporate AI teams hover around 10–15% women and ~2–4% Black staff.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 0:38Technology in times of crisis
- 2:08NextGen discrimination
- 3:49Artificial Intelligence and Machine Learning
- 5:55Amazons Resume Screening Tool
- 6:44Why was Amazons Resume Screening Tool sexist
- 8:10Ability Bias
- 9:05Predictive technologies
- 12:25Facial recognition
- 17:13COVID19 pandemic
- 18:24Facebooks original motto
- 18:51We need a yes and approach
- 20:47People are empowered to make good decisions
- 21:57Our mission
- 23:03Questions
related talks
35mBad computers, bad people
38mSocial Implications of Bias in Machine Learning
47mThe tech future is diverse
34m(Un)Ethical Machines? Why AI Bias Is a Problem and What to Do About It
56mAn AI with an Agenda: How Our Cognitive Biases Leak Into Machine Learning
34mAI and Gender: Building Fair AI Systems for the Future
47mEthics & AI: Identifying & preventing bias in predictive models
43mDiversity: Two steps forward, one step back (just keep it going)