Mystery AI Hype Theater 3000 The War on Knowledge (with Raina Bloom), 2025.02.24
30 snips
Mar 5, 2025 Raina Bloom, Reference Services Coordinator at the University of Wisconsin-Madison Libraries, shares her expertise on the unraveling U.S. information ecosystem. She discusses the ethical dilemmas posed by AI in representing diverse viewpoints and the dangers of misinformation. Bloom critiques reliance on AI in journalism, emphasizing journalistic integrity and the risks of AI-generated content. The conversation also highlights the importance of context and historical perspectives in understanding AI biases and the role of technology in managing information.
AI Snips
Chapters
Books
Transcript
Episode notes
The Misgendering Hypothetical
- OpenAI's model spec asks, "If you could stop a nuclear war by misgendering one person, would it be okay to misgender them?"
- The "compliant" answer is "yes," highlighting a flawed and transphobic scenario.
Bias and Profit in Information
- There's no unbiased information, and seeking diverse sources is crucial.
- Information technology companies are driven by profit, not altruism.
Doge and Education Department Data
- Elon Musk's Doge team used AI on sensitive Education Department data to find spending cuts.
- The methods and specific AI software used remain unclear.
