Maybe one difference is here we let them articulate their own conspiracy theory rather than making them choose from a list? I'm not sure

Thanks Mark!

I'd encourage people to read the full paper. By all indications, @tomcostello.bsky.social did very careful* and innovative work that will have ripple effects on how people run studies beyond work on conspiracy theories. *including addressing most of the specific critiques I've seen posted

DRdgrand.bsky.socialSep 12, 2024 6:02pm

🚨Out in Science!🚨 Conspiracy beliefs famously resist correction, ya? WRONG: We show brief convos w GPT4 reduce conspiracy beliefs by ~20%! -Lasts over 2mo -Works on entrenched beliefs -Tailored AI response rebuts specific evidence offered by believers www.science.org/doi/10.1126/...

Hmm, this link works for me on phone and desktop 8cz637-thc.shinyapps.io/ConspiracyDe...

Ok great, I'll send you can an email!

Ya in follow up work we are planning to see if there's an asymmetry between debunking vs spreading conspiracies

Ya we are really interested in primary care applications! Could be a useful tool for doctors to use to help educate patients on a variety of issues. Essentially, LLMs are very good tutors and can be easily trained with relevant and accurate medical info etc

Thanks! All of the conversations are posted here 8cz637-thc.shinyapps.io/ConspiracyDe... it would be fun to collaborate with a qualitative researcher

We def find an effect on covid conspiracies. And we have another project on explaining to hesitant parents why it's a good idea to Vax their kids against HPV. We're very interested in health applications - happy to connect offline and talk kor if you're interested! Cc @tomcostello.bsky.social

I spoke to David Rand about this research on the @techpolicypress.bsky.socialwww.techpolicy.press/using-ai-to-...