this entire book was the most helpful thing I read in all of grad school, especially the A-B-T framework for how to frame your contribution and outline an abstract/intro/talk

press.uchicago.edu/ucp/books/bo...

thanks for sharing Linda!

great minds!!!!

interesting!!

I hope this paper changes our view of the types of leaders we should select and promote, as well as the type of leaders we try to be.

anyways, that's enough musings. here's the link!
dx.doi.org/10.1037/pspa...

we wrote this paper at a fever pitch, mostly as a way for me to block out the job market blues. we have replicated these effects many, many times, with a variety of samples and methods.

I've never worked on a paper this polarizing! people either reacted with "duhh?!" or "no f*cking way" 😅

we find that communal traits like honesty, openness, care, and compassion define the ideal leader. we find evidence across eight studies (and five supplemental studies) that followers prefer communal leaders and will choose communality even at the expense of agency.

turns out, agentic traits like dominance, assertiveness, and even competence loom large in the minds of people when asked about the TYPICAL leader. but things shift dramatically when you ask about the IDEAL leader or the person people want to work for.

🧠 Interacting with AI changes how we see ourselves—when people read AI-generated advice before generating their own, they rated their own advice as less authentic.

in sum - use these tools with caution! 🚧

⚖️ When we told participants the advice was GPT-generated, they rated the same advice less favorably, revealing a persistent bias against AI in personal decision-making contexts.

💡 ChatGPT-generated advice was rated as higher quality, more effective, and even more authentic than human-generated advice—but only when people didn’t know it was from ChatGPT.

we replicated this with Claude and Gemini!

key findings summarized by the one and only ChatGPT. open-access link below!

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

in line with past work, we document AI's ability to generate quality content as well as people's aversion to content written by these tools.

but what excited me most about this project is that we also found that reading ChatGPT-generated advice shaped SELF-evaluations.

this has been on my TBR for so long! I need to actually read it this year!

oh wow - thanks Ben!

thanks so much Vanessa!! 🥹

thanks Pam!!

ohh fun! excited to read this in more detail.

really cool work!