Simulated Minds - Using AI to Deepen Research Insights | Ep. 24 - podcast episode cover

Simulated Minds - Using AI to Deepen Research Insights | Ep. 24

Aug 23, 202453 min
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Episode description

In this episode of ChatEDU (Simulated Minds - Using AI to Deepen Research Insights), Matt and Liz explore the use of AI in social science research, focusing on a study from NYU and Stanford about using large language models to simulate research participants.

First in our Intro: Matt and Liz discuss a trend of women using AI to fact-check men's height claims on dating apps. They then explore contrasting organizational approaches to AI adoption:

  • AI Bans in Organizations: Matt shares an anecdote about a friend's workplace banning all access to generative AI, while Liz discusses similar bans in various organizations.

  • Moderna's AI Adoption: Liz highlights Moderna's extensive use of custom GPTs across their organization, with 750 custom GPTs created and 40% of users actively using them weekly.

Then in the News:

  • Grok2 Image Generation Tool: Matt introduces Elon Musk's new AI model, Grok2 which lacks typical content filters. They discuss its potential misuse in school settings and the challenges it presents for educators and parents.

  • Student Attitudes Toward AI in Education: Liz shares findings from a global AI student survey by the Digital Education Council, revealing that many students believe overuse of AI in teaching devalues education and negatively impacts academic performance.

Beneath the Surface: Matt and Liz examine a research paper on using large language models to predict results of social science experiments:

  • Discussion of the study's methodology using AI to simulate research participants based on demographic profiles

  • Exploration of potential applications in cost-effective pilot studies and augmenting traditional research methods

  • Demonstration of the approach using Liz's doctoral dissertation on student self-efficacy in high school mathematics

  • Simulation of a research participant named Amber and comparison of AI-generated responses to actual research data

Short Stack Chat Back: A simulated Charles Darwin summarizes this week’s Beneath the Surface in three short sentences.

Have a question or comment? Please email Matt and Liz at chatedu@edadvance.org 

Related Links:

For the best experience, listen in Metacast app for iOS or Android