Episode 131: What's Wrong With AI (with Molly Crockett)

Cognitive scientist Molly Crockett joins the show to talk about what it's like to be known in the field as an "AI skeptic." We talk about where Molly is actually skeptical, but also what she thinks that label gets wrong. Along the way, we cover a bunch of other things too: Molly's research in social neuroscience, computational modeling, and moral psychology; Molly and Mickey's disagreements about empathic AI; and Molly's case for what she's calling "thick empathy." Finally, Molly makes a pitch for non-alcoholic beer, which sadly falls on deaf ears.

Special Guest: Molly Crockett.

Links:

OSF | The pitfalls of pay-to-play morality

AI Surrogates and illusions of generalizability in cognitive science: Trends in Cognitive Sciences

Replication for Language Models

GUIDE-LLM: Reporting checklist for studies with large language models in the behavioral and social sciences

Do Artifacts Have Politics? on JSTOR

CITP Seminar:Janet Vertesi - Where Fairness Fails: How to Confront Material Political Economy of AI - YouTube

\Reckoning with the Political Economy of AI: Avoiding Decoys in Pursuit of Accountability

UK quietly increases AI emissions forecast 100-fold – POLITICO

The carbon and water footprints of data centers and what this could mean for artificial intelligence: Patterns