Yue Zhao’s Post

👏 🎉 Excited to announce my latest first-authored work, Political-LLM, collaborating with 30+ AI researchers & political science researchers from world-leading Universities and top Industries. Special thanks to Prof. Yue Zhao from USC and Prof. Zhengzhong Tu from TAMU for sharing and posting this research. Amid the rapid advancements in AI technology, large language models (LLMs) have found extensive applications in political science, including election forecasting, sentiment analysis, policy evaluation, legislative analysis, and international diplomacy. However, there remain significant unanswered questions about how to systematically understand the use of LLMs in the broader field of political science and address the technical challenges they face. To tackle these challenges, our team has developed the Political-LLM framework, offering a comprehensive summary of LLM applications in computational political science and outlining future directions for their development.

I recently edited The Global Public Opinion on Artificial Intelligence survey (GPO-AI), which examined opinions about artificial intelligence (AI) in 21 countries. Created by the Schwartz Reisman Institute for Technology and Society in collaboration with Policy, Elections & Representation Lab at the University of Toronto - Munk School of Global Affairs & Public Policy, I also enjoyed speaking to the research lead, Peter John Loewen, about the implications. So many issues that I'll leave for another post. What I can say for sure is this: public trust in AI varies widely and needs some attention from policy makers. Reports like this can provide policy makers with a method for identifying the lowest-hanging fruit for their efforts to get the ball rolling. We will get there. Loved working on this, and it's intended to be for public consumption, so please dive in! https://lnkd.in/dk4-cZc6

Data, Data, Data everywhere! What do we know about data, who owns it, who controls it and who governs it and how we also need a decolonial lens to understand the data flow from the Global South to Global North. Manuel Castells in his seminal work The Rise of Networked Society (1996) talked about this power struggle and emergence of ‘informational economy’. In modern days, we need a refreshed look at data grab. Colonialism has not disappeared – it has taken on a new form. In the new world order, Big Tech companies are grabbing our most basic natural resource – our data – exploiting our labour and connections, and repackaging our information to track our movements, record our conversations and discriminate against us. Join Professor Nick Couldry, Professor Myria Georgiou of Department of Media and Communications at The London School of Economics and Political Science (LSE) and Professor Ulises A Mejias as we uncover this new power struggle. Check out the link below. #DataGrab #BookLaunch #DataFlow #LSE #PartofLSE

🌍 Exciting News for Democracy Research! 🗳️ We are thrilled to announce that the Peace Research Center Prague is part of a transdisciplinary project "Strengthening Democratic Resilience Through Digital Twins" (TWIN4DEM). Led by Erasmus School of Social and Behavioural Sciences (ESSB) the project received a €3 million Horizon Europe grant. Together with 11 European partners, this project will use cutting-edge Computational Social Science (CSS) methods like natural language processing and dynamic simulations to investigate how democracies decline and, more importantly, how we can prevent it. By creating the first-ever digital twins of political systems in Czechia, France, Hungary, and the Netherlands, the team aims to provide new insights into the causes and effects of democratic erosion. A huge thank you to our partners: Erasmus School of Social and Behavioural Sciences (ESSB), Université catholique de Lille, GESIS - Leibniz Institute for the Social Sciences, Fundazione Bruno Kessler, Linnaeus University, Babes-Bolyai University, MTA Társadalomtudományi Kutatóközpont, Eticas AI, Democracy International and DBC diadikasia. Looking forward to seeing how this innovative project unfolds over the next three years! 🚀 https://lnkd.in/eyzDUVPz. #HorizonEurope #DemocracyResearch #DigitalTwins #ComputationalSocialScience #ErasmusUniversity #TWIN4DEM #DemocraticResilience #Innovation

Information (whether true or false) has a social network effect that forms groups. This phenomenon will increasingly be influenced by AI which will intensify our information networks that determine our beliefs and behaviour even more. This is termed info-determinism. This will also further challenge expert’s authority - a key problem we are facing as university educators. Rejecting knowledge has become political - “no one can tell me what to think”. Many reject expertise and facts. To counter the intensification AI will bring - we need regulation and a “computer politics” that safeguards information. Like banning AI impersonations. For the moment I think AI is likely already shaping crucial information networks for our students and where we’re heading doesn’t appear progressive. A good read on info determinism here 👇 https://lnkd.in/ehsNp_Jw

While researching for one of my master's projects in Audiovisual and Multimedia Communication at the UC Dynamics of Social Media and Digital Media, I came across the article "From Poisons to Antidotes: Algorithms as Democracy Boosters" by Cavaliere & Romeo (2022). The text provided me with a critical analysis of the role of algorithms in democracy, a crucial issue in the digital society we live in. The article argues that algorithms can strengthen democracy, as long as they are implemented in a transparent, accessible, and auditable way. This would enable more efficient communication between government and the population, promoting citizen engagement and participation in political life. However, the text also acknowledges the challenges that accompany this promise. Algorithmic manipulation and discrimination are real concerns that need to be addressed to ensure fairness and equity in the use of these tools. The reading made me reflect on the role of algorithms in society and inspired me to seek solutions to the challenges they present. I believe that discussing this topic is fundamental for the future of democracy and digital communication. #AI #EU Cavaliere, P., & Romeo, G. (2022). From Poisons to Antidotes: Algorithms as Democracy Boosters. European Journal of Risk Regulation, 13(3), 421–442. doi:10.1017/err.2021.57 https://lnkd.in/dDJw93-Q

As a community college librarian deeply involved in teaching AI literacy, digital literacy, and media literacy, I found this article to be both alarming and a poignant illustration of our current challenges. As educators and librarians, we are on the front lines of ensuring that our community is equipped with the necessary skills to critically assess and question the validity of the information they encounter. Our role extends beyond traditional literacy; it is about fostering an informed citizenry capable of navigating the complexities of a digital world saturated with AI-generated content. This responsibility highlights why the skills we teach—critical thinking, source evaluation, and digital literacy—are more essential than ever. As the article exemplifies, without these competencies, the public is at risk of being swayed by fabricated realities that could influence their decisions at the polls. In embracing these challenges, we reinforce the value of librarians in our society, proving that our work is vital in guarding against the threats posed by emerging technologies. Let us continue to advocate for the importance of media literacy education and empower our communities to discern truth in an increasingly artificial world. #DigitalLiteracy #AILiteracy #MediaLiteracy #Librarians #Education #Democracy #InformationLiteracy https://lnkd.in/gkd6iZUm

The ways in which researchers can artificially inflate their reference counts are growing. The ways in which researchers can artificially inflate their reference counts are growing. https://lnkd.in/gNdp5Jch

A new #Rutgers project designed to advocate to STEM-educated New Jersey state legislators (who often champion evidence-informed policies and serve as trusted sources of scientific information for their colleagues) for evidence-informed AI regulation has been launched by two Rutgers faculty members: Anna M. Dulencin, director of the Rutgers Eagleton Institute’s Science and Politics Program, and SC&I Professor of Communication Itzhak Yanovitzky. “Scientists must work together with policymakers and the public to shape sound AI policy that harnesses its potential to benefit individuals and society while placing checks on its potential to cause harms. There is already a great deal of insight from across scientific disciplines and professional practice fields that can productively inform policy discourse but do not effectively reach federal and state policymakers,” they wrote. Read more: https://lnkd.in/e2uxAmKQ #RutgersResearch #RutgersExcellence

AI in Politics: TalkBot Generates Answers, Sparks Controversy #AIdisruptingpoliticalnorms #AIincampaignstrategies #AIinpolitics #AIgeneratedanswers #engagementfromopponents #ethicsofusingAIinpolitics #polarizationinsociety #TalkBot #technologyandtrustininstitutions #usingtechnologyinpolitics