-
Nicer Than Humans: How do Large Language Models Behave in the Prisoner's Dilemma?
Authors:
Nicoló Fontana,
Francesco Pierri,
Luca Maria Aiello
Abstract:
The behavior of Large Language Models (LLMs) as artificial social agents is largely unexplored, and we still lack extensive evidence of how these agents react to simple social stimuli. Testing the behavior of AI agents in classic Game Theory experiments provides a promising theoretical framework for evaluating the norms and values of these agents in archetypal social situations. In this work, we i…
▽ More
The behavior of Large Language Models (LLMs) as artificial social agents is largely unexplored, and we still lack extensive evidence of how these agents react to simple social stimuli. Testing the behavior of AI agents in classic Game Theory experiments provides a promising theoretical framework for evaluating the norms and values of these agents in archetypal social situations. In this work, we investigate the cooperative behavior of three LLMs (Llama2, Llama3, and GPT3.5) when playing the Iterated Prisoner's Dilemma against random adversaries displaying various levels of hostility. We introduce a systematic methodology to evaluate an LLM's comprehension of the game rules and its capability to parse historical gameplay logs for decision-making. We conducted simulations of games lasting for 100 rounds and analyzed the LLMs' decisions in terms of dimensions defined in the behavioral economics literature. We find that all models tend not to initiate defection but act cautiously, favoring cooperation over defection only when the opponent's defection rate is low. Overall, LLMs behave at least as cooperatively as the typical human player, although our results indicate some substantial differences among models. In particular, Llama2 and GPT3.5 are more cooperative than humans, and especially forgiving and non-retaliatory for opponent defection rates below 30%. More similar to humans, Llama3 exhibits consistently uncooperative and exploitative behavior unless the opponent always cooperates. Our systematic approach to the study of LLMs in game theoretical scenarios is a step towards using these simulations to inform practices of LLM auditing and alignment.
△ Less
Submitted 19 September, 2024; v1 submitted 19 June, 2024;
originally announced June 2024.
-
Modeling the amplification of epidemic spread by individuals exposed to misinformation on social media
Authors:
Matthew R. DeVerna,
Francesco Pierri,
Yong-Yeol Ahn,
Santo Fortunato,
Alessandro Flammini,
Filippo Menczer
Abstract:
Understanding how misinformation affects the spread of disease is crucial for public health, especially given recent research indicating that misinformation can increase vaccine hesitancy and discourage vaccine uptake. However, it is difficult to investigate the interaction between misinformation and epidemic outcomes due to the dearth of data-informed holistic epidemic models. Here, we employ an…
▽ More
Understanding how misinformation affects the spread of disease is crucial for public health, especially given recent research indicating that misinformation can increase vaccine hesitancy and discourage vaccine uptake. However, it is difficult to investigate the interaction between misinformation and epidemic outcomes due to the dearth of data-informed holistic epidemic models. Here, we employ an epidemic model that incorporates a large, mobility-informed physical contact network as well as the distribution of misinformed individuals across counties derived from social media data. The model allows us to simulate various scenarios to understand how epidemic spreading can be affected by misinformation spreading through one particular social media platform. Using this model, we compare a worst-case scenario, in which individuals become misinformed after a single exposure to low-credibility content, to a best-case scenario where the population is highly resilient to misinformation. We estimate the additional portion of the U.S. population that would become infected over the course of the COVID-19 epidemic in the worst-case scenario. This work can provide policymakers with insights about the potential harms of exposure to online vaccine misinformation.
△ Less
Submitted 29 January, 2025; v1 submitted 17 February, 2024;
originally announced February 2024.
-
Misinformation and Polarization around COVID-19 vaccines in France, Germany, and Italy
Authors:
Gianluca Nogara,
Francesco Pierri,
Stefano Cresci,
Luca Luceri,
Silvia Giordano
Abstract:
The kick-off of vaccination campaigns in Europe, starting in late December 2020, has been followed by the online spread of controversies and conspiracies surrounding vaccine validity and efficacy. We study Twitter discussions in three major European languages (Italian, German, and French) during the vaccination campaign. Moving beyond content analysis to explore the structural aspects of online di…
▽ More
The kick-off of vaccination campaigns in Europe, starting in late December 2020, has been followed by the online spread of controversies and conspiracies surrounding vaccine validity and efficacy. We study Twitter discussions in three major European languages (Italian, German, and French) during the vaccination campaign. Moving beyond content analysis to explore the structural aspects of online discussions, our investigation includes an analysis of polarization and the potential formation of echo chambers, revealing nuanced behavioral and topical differences in user interactions across the analyzed countries. Notably, we identify strong anti- and pro-vaccine factions exhibiting heterogeneous temporal polarization patterns in different countries. Through a detailed examination of news-sharing sources, we uncover the widespread use of other media platforms like Telegram and YouTube for disseminating low-credibility information, indicating a concerning trend of diminishing news credibility over time. Our findings on Twitter discussions during the COVID-19 vaccination campaign in major European languages expose nuanced behavioral distinctions, revealing the profound impact of polarization and the emergence of distinct anti-vaccine and pro-vaccine advocates over time.
△ Less
Submitted 6 February, 2024;
originally announced February 2024.
-
Online misinformation is linked to early COVID-19 vaccination hesitancy and refusal
Authors:
Francesco Pierri,
Brea Perry,
Matthew R. DeVerna,
Kai-Cheng Yang,
Alessandro Flammini,
Filippo Menczer,
John Bryden
Abstract:
Widespread uptake of vaccines is necessary to achieve herd immunity. However, uptake rates have varied across U.S. states during the first six months of the COVID-19 vaccination program. Misbeliefs may play an important role in vaccine hesitancy, and there is a need to understand relationships between misinformation, beliefs, behaviors, and health outcomes. Here we investigate the extent to which…
▽ More
Widespread uptake of vaccines is necessary to achieve herd immunity. However, uptake rates have varied across U.S. states during the first six months of the COVID-19 vaccination program. Misbeliefs may play an important role in vaccine hesitancy, and there is a need to understand relationships between misinformation, beliefs, behaviors, and health outcomes. Here we investigate the extent to which COVID-19 vaccination rates and vaccine hesitancy are associated with levels of online misinformation about vaccines. We also look for evidence of directionality from online misinformation to vaccine hesitancy. We find a negative relationship between misinformation and vaccination uptake rates. Online misinformation is also correlated with vaccine hesitancy rates taken from survey data. Associations between vaccine outcomes and misinformation remain significant when accounting for political as well as demographic and socioeconomic factors. While vaccine hesitancy is strongly associated with Republican vote share, we observe that the effect of online misinformation on hesitancy is strongest across Democratic rather than Republican counties. Granger causality analysis shows evidence for a directional relationship from online misinformation to vaccine hesitancy. Our results support a need for interventions that address misbeliefs, allowing individuals to make better-informed health decisions.
△ Less
Submitted 12 July, 2022; v1 submitted 21 April, 2021;
originally announced April 2021.
-
Human Mobility in Response to COVID-19 in France, Italy and UK
Authors:
Alessandro Galeazzi,
Matteo Cinelli,
Giovanni Bonaccorsi,
Francesco Pierri,
Ana Lucia Schmidt,
Antonio Scala,
Fabio Pammolli,
Walter Quattrociocchi
Abstract:
The policies implemented to hinder the COVID-19 outbreak represent one of the largest critical events in history. The understanding of this process is fundamental for crafting and tailoring post-disaster relief. In this work we perform a massive data analysis, through geolocalized data from 13M Facebook users, on how such a stress affected mobility patterns in France, Italy and UK. We find that th…
▽ More
The policies implemented to hinder the COVID-19 outbreak represent one of the largest critical events in history. The understanding of this process is fundamental for crafting and tailoring post-disaster relief. In this work we perform a massive data analysis, through geolocalized data from 13M Facebook users, on how such a stress affected mobility patterns in France, Italy and UK. We find that the general reduction of the overall efficiency in the network of movements is accompanied by geographical fragmentation with a massive reduction of long-range connections. The impact, however, differs among nations according to their initial mobility structure. Indeed, we find that the mobility network after the lockdown is more concentrated in the case of France and UK and more distributed in Italy. Such a process can be approximated through percolation to quantify the substantial impact of the lockdown.
△ Less
Submitted 13 May, 2020;
originally announced May 2020.
-
Evidence of economic segregation from mobility lockdown during COVID-19 epidemic
Authors:
Giovanni Bonaccorsi,
Francesco Pierri,
Matteo Cinelli,
Francesco Porcelli,
Alessandro Galeazzi,
Andrea Flori,
Ana Lucia Schmidt,
Carlo Michele Valensise,
Antonio Scala,
Walter Quattrociocchi,
Fabio Pammolli
Abstract:
In response to the COVID-19 pandemic, National governments have applied lockdown restrictions to reduce the infection rate. We perform a massive analysis on near real-time Italian data provided by Facebook to investigate how lockdown strategies affect economic conditions of individuals and local governments. We model the change in mobility as an exogenous shock similar to a natural disaster. We id…
▽ More
In response to the COVID-19 pandemic, National governments have applied lockdown restrictions to reduce the infection rate. We perform a massive analysis on near real-time Italian data provided by Facebook to investigate how lockdown strategies affect economic conditions of individuals and local governments. We model the change in mobility as an exogenous shock similar to a natural disaster. We identify two ways through which mobility restrictions affect Italian citizens. First, we find that the impact of lockdown is stronger in municipalities with higher fiscal capacity. Second, we find a segregation effect, since mobility restrictions are stronger in municipalities for which inequality is higher and where individuals have lower income per capita.
△ Less
Submitted 11 April, 2020;
originally announced April 2020.
