Lightnews — Scholar-powered news

Tom Costello

@tomcostello.bsky.social

3.7K followers 220 following 150 posts

research psychologist. beliefs, AI, computational social science. prof at american university.

Posts Media Videos Starter Packs

Tom Costello @tomcostello.bsky.social · 19h

I’m going to be in Montreal for a few days starting tomorrow for COLM — anyone also at the conference / interested in meeting up, let me know!

Reposted by Tom Costello

Florian Foos @florianfoos.bsky.social · Jul 28

This is a valid point, I think. The question is always what type of alternative information gathering processes AI chatbots replace. In the case of medical "self diagnosis", there is some reason to believe that common alternative mechanisms aren't superior.

Tom Costello @tomcostello.bsky.social · Jul 17

Is there a strong case for AI helping, rather than harming, the accuracy of people's beliefs about contentious topics? In this
@nature.com Nature Medicine piece (focusing on vaccination), I argue the answer is YES. And it boils down to how LLMs differ from other sources of information.

Tom Costello @tomcostello.bsky.social · Jul 24

Maybe you see this as all too rosy, which is fair and maybe even true, but warnings and dismissals (alone) are bad tools, if nothing else. future isn't set. So yes, I believe we should actively articulate and defend a positive vision in order to reduce harms + capture gains.

Tom Costello @tomcostello.bsky.social · Jul 24

Targeted ads have gone too far

Tom Costello @tomcostello.bsky.social · Jul 24

Also, incentives are not static; if revenue continues to come from usage fees (rather than ads), maybe helping users reach reliable answers is indeed a profitable/competitive approach. open question. plus i don't imagine these big companies want to replay social media era mistakes

Tom Costello @tomcostello.bsky.social · Jul 24

So the problem is incentives. I agree. The incentives are aligned with building the models in the first place, too (hence my first sentence in that quote). Should we not try to identify and bolster a positive vision that underscores potential returns to cooperation, democracy, etc?

Tom Costello @tomcostello.bsky.social · Jul 24

Thanks for sharing!

Reposted by Tom Costello

Scott McGrath @smcgrath.phd · Jul 16

Thomas Costello argues that as patients move from WebMD to AI, we might be slightly optimism. Unlike former tools, LLMs can synthesize vast, shared knowledge, potentially helping users converge on more accurate beliefs.

The major caveat is: as long as the LLMs are not trained on bad data.

Large language models as disrupters of misinformation - Nature Medicine

As patients move from WebMD to ChatGPT, Thomas Costello makes the case for cautious optimism.

Tom Costello @tomcostello.bsky.social · Jul 17

Open link here: www.nature.com/articles/s41...

Large language models as disrupters of misinformation

Nature Medicine - As patients move from WebMD to ChatGPT, Thomas Costello makes the case for cautious optimism.

Tom Costello @tomcostello.bsky.social · Jul 17

Is there a strong case for AI helping, rather than harming, the accuracy of people's beliefs about contentious topics? In this
@nature.com Nature Medicine piece (focusing on vaccination), I argue the answer is YES. And it boils down to how LLMs differ from other sources of information.

Reposted by Tom Costello

Robert Califf @robcaliff.bsky.social · Jul 16

This is an interesting take on LLMs and the RCT in vaccine hesitancy is fascinating.

www.nature.com/articles/s41...

Large language models as disrupters of misinformation - Nature Medicine

As patients move from WebMD to ChatGPT, Thomas Costello makes the case for cautious optimism.

Tom Costello @tomcostello.bsky.social · Jul 9

More on that front soon, actually...

Tom Costello @tomcostello.bsky.social · Jul 9

I think this is interesting, and it would be worthwhile to convene a group and expose them to this chatbot interaction (perhaps it would be much less effective when social dynamics are involved), but I think the active ingredient is strong arguments + evidence. LLMs can surface good arguments.

Reposted by Tom Costello

Tom Costello @tomcostello.bsky.social · Jul 9

Conspiracies emerge in the wake of high-profile events, but you can’t debunk them with evidence because little yet exists. Does this mean LLMs can’t debunk conspiracies during ongoing events? No!

We show they can in a new working paper.

PDF: osf.io/preprints/ps...

Tom Costello @tomcostello.bsky.social · Jul 9

You mean given everything with the Epstein files?

Reposted by Tom Costello

David Rand @dgrand.bsky.social · Jul 9

I'm very excited about this new WP showing that LLMs effectively countered conspiracies in the immediate aftermath of the 1st Trump assassination attempt, and that treatment also reduced conspiratorial thinking about the subsequent 2nd assassination attempt

Tom Costello @tomcostello.bsky.social · Jul 9

Conspiracies emerge in the wake of high-profile events, but you can’t debunk them with evidence because little yet exists. Does this mean LLMs can’t debunk conspiracies during ongoing events? No!

We show they can in a new working paper.

PDF: osf.io/preprints/ps...

Tom Costello @tomcostello.bsky.social · Jul 9

yeah, I think that talking to an LLM that is prompted to behave like ChatGPT is likely to amplify whichever tendencies already exist in a person (so can weird beliefs, like has been reported). But our studies give the LLM a very specific goal (e.g., debunking), so it is not 1:1 in a meaningful way

Tom Costello @tomcostello.bsky.social · Jul 9

We also find this intervention succeeds for vaccine skepticism:

bsky.app/profile/dgra...

David Rand @dgrand.bsky.social · Apr 18

🚨New WP🚨
Using GPT4 to persuade participants significantly reduces climate skepticism and inaction
-Sig more effective than consensus messaging
-Works for Republicans
-Evidence of persistence @ 1mo
-Scalable!
PDF: osf.io/preprints/ps...
Try the bot: www.debunkbot.com/climate-change
Here’s how 👇

Tom Costello @tomcostello.bsky.social · Jul 9

Do these effects succeed for non-conspiracy beliefs, like climate attitudes? yes!

bsky.app/profile/dgra...

David Rand @dgrand.bsky.social · Apr 18

🚨New WP🚨
Using GPT4 to persuade participants significantly reduces climate skepticism and inaction
-Sig more effective than consensus messaging
-Works for Republicans
-Evidence of persistence @ 1mo
-Scalable!
PDF: osf.io/preprints/ps...
Try the bot: www.debunkbot.com/climate-change
Here’s how 👇

Tom Costello @tomcostello.bsky.social · Jul 9

Second, why are debunking dialogues so effective? Good arguments and evidence! (and, for unfolding conspiracies, saying "no one knows what's going on, you should be epistemically cautious" may be a strong argument)

bsky.app/profile/tomc...

Tom Costello @tomcostello.bsky.social · Feb 18

Last year, we published a paper showing that AI models can "debunk" conspiracy theories via personalized conversations. That paper raised a major question: WHY are the human<>AI convos so effective? In a new working paper, we have some answers.

TLDR: facts

osf.io/preprints/ps...

Tom Costello @tomcostello.bsky.social · Jul 9

Some other recent papers from our group on AI debunking:

First, does this work if people think they're talking to a human being? yes!

bsky.app/profile/gord...

Gordon Pennycook @gordpennycook.bsky.social · May 15

Recent research shows that AI can durably reduce belief in conspiracies. But does this work b/c the AI is good at producing evidence, or b/c ppl really trust AI?

In a new working paper, we show that the effect persists even if the person thinks they're talking to a human: osf.io/preprints/ps...

🧵

Tom Costello @tomcostello.bsky.social · Jul 9

Huge thanks to my brilliant co-authors: Nathaniel Rabb (who split the work with me and is co-first author), Nick Stagnaro, @gordpennycook.bsky.social, and @dgrand.bsky.social

We're eager to hear your thoughts and feedback!

Tom Costello @tomcostello.bsky.social · Jul 9

Also, the treatment succeeded for both Democrats and Republicans, who endorsed slightly different conspiratorial explanations of the assassination attempts (see figure below for a breakdown)

Tom Costello @tomcostello.bsky.social · Jul 9

(the most notable part?): The effect was durable and preventative. When we recontacted participants 2 months later after the second assassination attempt, those from the tx group were ~50% less likely to endorse conspiracies about this new event! The debunking acted as an "inoculation" of sorts.

Tom Costello @tomcostello.bsky.social · Jul 9

Did this work? Yes. The Gemini dialogues significantly reduced conspiracy beliefs compared to controls who chatted about an irrelevant topic or just read a fact sheet (d = .38). The effect was robust across multiple measures.

Key figure attached