Rui Pei

🚨📢 Excited to share our new paper out today in @NatureHumBehav: “Bridging the empathy perception gap fosters social connection” nature.com/articles/s4156… A thread

🧍‍♀️↔️🧍‍♂️Social connection buffers us from #stress, but social disconnection is prevalent and increasing among young people. Field interventions show promising results in improving the social networks of undergraduate students. Read more in Ch5 #WHR2025 👉 worldhappiness.report/ed/2025/connec…

This work has important practical implications: even the most caring community can feel lonely if its members don’t see how much others care. We hope this inspires new ways to foster empathy, belonging, and social courage.

Many interventions focus on what people should do (eg. skills training) or what institutions should provide (socializing opportunities). We contribute to a growing line of research on reducing social disconnection by introducing a perception-based approach

🚨📢 Excited to share our new paper out today in @NatureHumBehav: “Bridging the empathy perception gap fosters social connection” nature.com/articles/s4156… A thread

@akaPrateekshit yesss!

Should I post videos from my science comedy show over here? Would people be interested in watching them?

Repetitive ads are annoying, but it's generally understood that more exposure -> better intended outcomes. Is this always true? In a new paper, we find that too much exposure can backfire by reducing perceived value and message engagement. Link: academic.oup.com/abm/article-ab…

🧑‍💻 World Happiness Report 2026 will focus on the relationship between social media and subjective #wellbeing, featuring a chapter from @JonHaidt, author of The Anxious Generation. Find out more and submit your own chapter proposal by Sunday 15 June👇 worldhappiness.report/news/submit-a-…

@akaPrateekshit SANS can't come soon enough!!

hey that's me! come play with me at Comic SANS I promisseeeee it will be fun and silly okay? okay.

LLMs tend to provide overly "safe" social interactions. If people get used to interacting with LLMs, interacting with real humans will be even riskier and harder news.ucsc.edu/2025/03/ai-emp…