Science Magazine Podcast Matching sounds to shapes, and stories from the AAAS annual meeting
13 snips
Feb 19, 2026 Maria Loconsole, postdoc studying cross-modal correspondences, reports that newly hatched chicks show the bouba–kiki effect. Michael Greshko, science reporter, recounts AAAS themes like AI in research and shifting outlook for U.S. science. David Rand, Cornell professor known for misinformation research, discusses using LLMs to combat conspiracy beliefs. Multiple lively conversations tie AI, research trends, and innate perception together.
AI Snips
Chapters
Books
Transcript
Episode notes
Push For Prompt Transparency
- Demand prompt and instruction transparency when using frontier LLMs for decision-making or public messaging.
- Without credible disclosure you cannot reliably assess model agendas or trustworthiness.
DebunkBot's Early Traction
- Rand's team launched debunkbot.com and saw ~150,000 visitors organically without promotion.
- They are also experimenting with social tagging and classroom adoption to scale use.
AI: Promise And Disruption At AAAS
- AAAS attendees saw AI both as a powerful research tool and a disruptive force across disciplines.
- Michael Greshko noted excitement about AI aiding research alongside concerns about systemic impacts on science careers.
