Reducing Bias in AI with Open Source
summary
Claims that large models skew toward Western views because training sets and tokenizers favor English, with one example showing the same sentence takes 7 tokens in English and 16 in Vietnamese. Reports of Kenyan annotators paid about $2 per hour feeding problematic labels into models accompany examples: facial recognition misidentifies people, hiring algorithms reproduce gender bias, and tokenization raises cost barriers. Proposals include publishing datasets on Hugging Face, building a funded foundation to coordinate GPU donations and audit datasets so models become more accountable.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 3:19AI bias concerns
- 7:01AI bias examples
- 10:58Open source for bias reduction
- 19:32Challenges in open AI models
- 28:27Open floor discussion on solutions and ideas
- 42:44Summary
related talks
43mReducing Bias in AI with Open Source
11mMachines of Love & Grace – Building Truly Inclusive AI
32mRoll for Initiative: how to make the world of AI a more ethical place
05mWhat Do We Mean When We say Open Source AI?
26mBuilding ethical AI: Transparency, bias mitigation, and trust
05mToward Ethical AI
19mMoral Outsourcing: Humanity in the age of AI
40mCulturally-Inclusive AI: Enhancing User Experience With Global Perspectives