Why the Future of AI Is Feminist
summary
Generative models produce only 21% women generated overall (12% for powerful, ~20% for successful, ~60% for “beautiful”; Delhi prompts show 6% women in power; Midjourney shows 92% women for beauty), and systems reproduce harms like the Apple Card credit bias, Amazon’s recruiting that filtered out women, worse skin‑cancer detection on darker skin, and >90% of deepfake pornography victims being women. Proposed fixes include funded feminist AI research, diversity committees, EU regulation, and applied tools like an AI meeting analyzer and women‑focused health diagnostics.
AI-written summary · about
chapters (8)
- 0:00Introduction and welcome
- 1:07Present AI is not feminist
- 8:17Power structures in AI development
- 14:01Why future AI must be feminist
- 17:34Political and organizational change
- 21:25Feminist AI use cases and examples
- 24:55Call to action and collaboration
- 25:16Q&A begins
related talks
34mAI and Gender: Building Fair AI Systems for the Future
47mHow gender imbalance is holding back the future of AI
34m(Un)Ethical Machines? Why AI Bias Is a Problem and What to Do About It
41mAll About Biases in Artificial Intelligence
22mGender Bias in Artificial Intelligence
44mPanel: AI Feminism
11mAddressing Privacy, Bias, and Fairness in AI Systems
29mUnintended consequences of technology: Enabling NextGen discrimination