Machine Ethics - moving from Value to Values-driven insights
summary
Deploys a humanoid robot prototype and a drone as digital front doors in US bank and healthcare pilots, swapped a white robot for a pink one after bias surfaced, and finds elderly users can’t relate to humanoids; runs a 2018 voice‑assistant emotion test across Siri, Alexa and Bixby and quotes Siri’s responses; built RPA conversational bots for call centers and rolled out credit‑bureau models to 15 APAC markets; argues ML can’t infer human feelings, shows crime‑prediction bias from policing data, and proposes scoring AI for values and bias resilience.
AI-written summary · about
chapters (12)
- 0:00Machine ethics introduction
- 1:06Data-driven to embedded ethics
- 3:57Beyond statistics: intent & decisions
- 6:04Stancil Petrov's intuition
- 8:53Machines and human emotions
- 11:20BANI world & ethical AI
- 13:04Stacy's complexity matrix
- 15:10Bias in credit scoring
- 17:16Fairness in ML algorithms
- 19:21AI for uncertain scenarios
- 20:25Bias & transparency in AI
- 21:06Ethical challenges & humanized AI
related talks
45mEthical use of AI: seeking, understanding, and countering the risks
16mMachine Intelligence With Humans in Mind
36mKeynote: Ethics in AI: Building Fair and Just Intelligence Systems
42mBrand Humanizing: A humane turn on AI
35mDesigning Inclusive and Accessible Experiences in the Era of AI
19mMoral Outsourcing: Humanity in the age of AI
03mMaking good robots
26mHow to develop useful and unbiased AI