1/ Excited to report we have a new paper out @Nature today! The bottom line: training data for LLMs does not just fall from the sky - it is created in the context of existing social political institutions - and that has consequences for LLM output.

nature.com/articles/s4158… 2/ More specifically, in countries with lower levels of media freedom, when you query models in the language of that country about politics in that country, you are more likely to get a pro-regime/leadership answer than if you ask in a different language.

📢📣 We've got a new paper out from @CSMaP_NYU today in @NatureComms 📢

Russia's foreign influence campaign on Twitter in 2016 caused widespread concern. But who was exposed? How effective was it? 🧵👇

1/

nature.com/articles/s4146… Russia's efforts to intervene with the 2016 election have been widely documented by news media and investigators. But researchers' understanding of exactly how influential these campaigns were has been largely restricted due to data limitations. 2/

📢📣New publication @ScienceAdvances on echo chambers! Most users don't follow political elites on Twitter. But those who do show overwhelming preferences for ideological congruity.

Co-authored w/@mwojcieszak @CasAndreu @imxudong @Jonathan_Nagler

🧵1/
science.org/doi/10.1126/sc… Social media facilitates exposure to different perspectives & w/diverse people. It may also lead to insular online communities, where users only find info consistent w/their views. It is feared that such “echo chambers” fuel extremity & exacerbate inter-party hostility.

2/

📢We’ve got a new @CSMaP_NYU paper in @journalsafetech
Fraud & conspiracy narratives proliferated on social media around the 2020 election. Our analysis finds YouTube was more likely to recommend fraud videos to users already skeptical about legitimacy of the election. 🧵👇1/ @CSMaP_NYU @journalsafetech There's growing scholarly consensus that social media algorithms have little influence on online echo chambers, in which users only see content reaffirming pre-existing views. But what if that content is undermining democratic confidence? 2/

Just read through the entire list of 963 Americans banned from Russia. Some quick, not entirely systematic thoughts in a short 🧵

1/ washingtonpost.com/world/2022/05/… It is very, very heavy on members of Congress - both House and Senate, and both sides of the aisle

2/

📣A Senate Judiciary subcommittee will hold a hearing at 2pm today on platform transparency/understanding the impact of social media.

The conversation about social media data access should start with what questions we want to answer with those data.🧵1/
judiciary.senate.gov/meetings/platf… At @CSMaP_NYU, we've been researching social media and politics for more than a decade. But from the beginning, we've done so with one hand tied behind our backs because social media companies tightly control the data necessary to study the platforms' impact.

2/

After weeks of war, much has been written about the success and failures of Russia’s propaganda. It's worth stepping back to consider the various audiences for Russia’s disinformation and examine where it's working and where it's not. 🧵

1/
slate.com/technology/202… First, where it's not: Ukraine & the West. Putin has spread a long line of lies -- denying Ukrainian statehood, claiming Nazis rule the country. Now they say Ukrainians are bombing their own people. These falsehoods quickly fell apart, largely because true info won out.

2/

Important piece in @nytimes today about new tactics from Russian gov't to deal with online opposition. A short 🧵 on a conceptual framework for understanding these developments

nytimes.com/2021/10/22/tec…

1/ In a 2018 Comparative Politics article with @SergeySanovich @DenisStukal, we introduced a three part framework for thinking about how non-democratic regimes respond to online opposition.

2/

Very pleased that @JEPS_ed will be publishing a special issue in honor of Becky Morton that will be guest edited by @rkwrice @TorreyShine and me.

Please consider submitting a paper! A short thread with some instructions regarding submission 👇

1/ We are looking for papers that reflect the many contributions Becky made to the field of experimental research in political science. This could include, but is not limited to:

2/

For anyone interested in the larger context behind social media’s impact on yesterday’s events specifically, and democracy generally, I’m reupping this post from 2017.

A quick thread:

wapo.st/3nrXTDQ

1/ The post summarizes a @JoDemocracy article by me @mollyeroberts @p_barbera @Yannis_Theo entitled “From Liberation to Turmoil: Social Media and Democracy”:

bit.ly/3hYcgyD

2/

Now available from @CUP_PoliSci , Social Media and Democracy: The State of the Field and Prospects for Reform, edited by @persily and me. [A Thread on the Chapters] 1/17

rb.gy/gtlh0a Book contains 13 jam-packed chapters on the most pressing topics on the democracy-related challenges of social media and what to do about them. 2/17

rb.gy/gtlh0a