Gregory Eady

New preprint! We developed new measurement tools to study moralization in ~2B X & Reddit posts and ~5M traditional media texts. Key finding: moralization increased markedly on social media from 2013-2021; more than traditional media; associated with multiple user dynamics 🧵👇

Today (w/ @UniofOxford @Stanford @MIT @LSEnews) we’re sharing the results of the largest AI persuasion experiments to date: 76k participants, 19  LLMs, 707 political issues. We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more 🧵

Excited to report that this paper with @Jacob_Morrier is now published in @polanalysis! Interested in LLMs? Give it a read! "Measuring the Quality of Answers in Political Q&As with Large Language Models", bit.ly/44PMwNP

Most important tech blog this year: OpenAI engineer and ex-founder of $3.5B Segment wrote a tell all post about how OpenAI works internally. From obsession with X, devout use of Slack to engineering culture and tech stack. A peek under the hood of a generational company.

How harmful is GenAI around elections? Will it trigger a misinformation apocalypse and upend elections? I am happy to finally be able to share @Sacha_Altay’s & my answers to these and other questions on which we have been working for a year and which is out via @knightcolumbia.

When researchers label social media posts using only their text and ignore multimodal content (e.g., videos, images, links), how much do they miss? Our answer: Not much. New paper with @JamesBisbee @Jonathan_Nagler @j_a_tucker at @CSMaP_NYU doi.org/10.1017/psrm.2…