Thread by @margaret_crane1

Given the interest with this tweet, I wanted to share some thoughts on how #ChatGPT also can be useful for brainstorming ideas (research, clinical, or anything really)... here are some examples, and a few caveats

https://twitter.com/margaret_crane1/status/1620529025860132864

Brainstorm project ideas with #ChatGPT by asking it for future directions in your field. This is pretty spot on for #impsci. You can drill down by asking follow up questions (tell me more about X).

H/T to @daniadamsPhD for this prompt

Ask #ChatGPT for what could go wrong in a project as a premortum. This response is a great explanation for the null results in my diss (testing opinion leaders as a dissemination strategy), and would have been helpful to consider before starting the project.

Generating ideas for exposure therapy with #ChatGPT. #CBTWorks Great starting point to customize with a client

The big caveat is that #ChatGPT is only as evidence-based as the info it pulls from. I found it to be pretty accurate in my field, but you can't blindly trust it. Knowing a bit about a topic helps you search better, and be an informed consumer of its response

For those of us in the dissemination space, this is certainly something to follow. It has huge potential as a new source of information (and hopefully not misinformation).

@emollick inspired a lot of these ideas. Check out his blog for more! oneusefulthing.substack.com/p/the-mechanic…