Putting AI to Work on Our Biggest Challenges
summary
Generative AI amplifies existing data gaps: models train mostly on Western English sources so cultural errors and bias persist, and solutions require data.org style capacity building plus paid, traceable data supply chains like Manu Chopra’s Indian-language efforts. Companies report concrete impact—Mastercard operates in 200 markets across 186 billion transactions and says it has $40 billion defended over three years—while Amazon cites having trained 21 million engineers and Sweden’s mammogram AI shows ~30% better detection than a single radiologist.
AI-written summary · about
chapters (13)
- 0:00AI's dual impact
- 1:55Standards, solutions, scale
- 4:07Generative AI strengths and risks
- 6:13Data bias and culture
- 8:39Bias in large language models
- 10:23Data.org's inclusive AI mission
- 12:31Adoption cycles and education gaps
- 15:18AI's societal impact and misinformation
- 17:28Guardrails for AI safety
- 19:12Open vs closed AI systems
- 21:59AI amplifying societal biases
- 23:25Human oversight in AI decisions
- 24:10Call for AI governance
related talks
25mSynergizing AI Innovation
53mThe Evolution and Impact of Generative AI: Transforming Creativity and Operations
20mHarnessing AI: Business Impact Beyond 2025
27mChallenges of Generative AI
28mEthical AI?!
30mGenerative AI in action: from POC to business value
20mBeyond the Buzz: Discovering High-Impact Use Cases in the Generative AI Era
20mGetting business value from AI: Everyday AI, extraordinary people