Jeffrey T. Hancock

Chenyan Jia, Michelle S. Lam, Minh Chau Mai et al. · stanford

Can we design artificial intelligence (AI) systems that rank our social media feeds to consider democratic values such as mitigating partisan animosity as part of their objective functions? We introduce a method for translating established, vetted social scientific constructs into AI objective functions, which we term societal objective functions, and demonstrate the method with application to the political science construct of anti-democratic attitudes. Traditionally, we have lacked observable outcomes to use to train such models, however, the social sciences have developed survey instruments and qualitative codebooks for these constructs, and their precision facilitates translation into detailed prompts for large language models. We apply this method to create a democratic attitude model that estimates the extent to which a social media post promotes anti-democratic attitudes, and test this democratic attitude model across three studies. In Study 1, we first test the attitudinal and behavioral effectiveness of the intervention among US partisans (N=1,380) by manually annotating (alpha=.895) social media posts with anti-democratic attitude scores and testing several feed ranking conditions based on these scores. Removal (d=.20) and downranking feeds (d=.25) reduced participants' partisan animosity without compromising their experience and engagement. In Study 2, we scale up the manual labels by creating the democratic attitude model, finding strong agreement with manual labels (rho=.75). Finally, in Study 3, we replicate Study 1 using the democratic attitude model instead of manual labels to test its attitudinal and behavioral impact (N=558), and again find that the feed downranking using the societal objective function reduced partisan animosity (d=.25). This method presents a novel strategy to draw on social science theory and methods to mitigate societal harms in social media AIs.

Maurice Jakesch, Jeffrey Hancock, Mor Naaman

Human communication is increasingly intermixed with language generated by AI. Across chat, email, and social media, AI systems suggest words, complete sentences, or produce entire conversations. AI-generated language is often not identified as such but presented as language written by humans, raising concerns about novel forms of deception and manipulation. Here, we study how humans discern whether verbal self-presentations, one of the most personal and consequential forms of language, were generated by AI. In six experiments, participants (N = 4,600) were unable to detect self-presentations generated by state-of-the-art AI language models in professional, hospitality, and dating contexts. A computational analysis of language features shows that human judgments of AI-generated language are hindered by intuitive but flawed heuristics such as associating first-person pronouns, use of contractions, or family topics with human-written language. We experimentally demonstrate that these heuristics make human judgment of AI-generated language predictable and manipulable, allowing AI systems to produce text perceived as "more human than human." We discuss solutions, such as AI accents, to reduce the deceptive potential of language generated by AI, limiting the subversion of human intuition.

Yutong Zhang, Dora Zhao, Jeffrey T. Hancock et al.

As large language model (LLM)-enhanced chatbots become increasingly expressive and socially responsive, many users begin forming companionship-like bonds with them. This study investigates how using AI companions relates to psychological well-being. We collected self-reported data from 1,131 U.S. adults who use CharacterAI, including survey responses and 4,664 chat sessions (464,687 messages) from 237 participants. By triangulating self-reported usage, relationship descriptions, and real chat histories, we identify patterns of engagement and associated outcomes. Smaller social networks were associated with reporting companionship as the primary chatbot use (beta = -0.03, 95% confidence interval (CI) [-0.05, -0.01]), which in turn was associated with lower well-being (beta = -0.48, 95% CI [-0.70, -0.25]). For self-reported companionship usage, this association was stronger when interactions were intensive (beta = -0.31, 95% CI [-0.56, -0.06]) and highly disclosive (beta = -0.38, 95% CI [-0.63, -0.14]). These results suggest that the association between AI companionship and well-being is not uniform and depends on how chatbots are used and users' offline social environments.

Tiziano Piccardi, Martin Saveski, Chenyan Jia et al.

There is widespread concern about the negative impacts of social media feed ranking algorithms on political polarization. Leveraging advancements in large language models (LLMs), we develop an approach to re-rank feeds in real-time to test the effects of content that is likely to polarize: expressions of antidemocratic attitudes and partisan animosity (AAPA). In a preregistered 10-day field experiment on X/Twitter with 1,256 consented participants, we increase or decrease participants' exposure to AAPA in their algorithmically curated feeds. We observe more positive outparty feelings when AAPA exposure is decreased and more negative outparty feelings when AAPA exposure is increased. Exposure to AAPA content also results in an immediate increase in negative emotions, such as sadness and anger. The interventions do not significantly impact traditional engagement metrics such as re-post and favorite rates. These findings highlight a potential pathway for developing feed algorithms that mitigate affective polarization by addressing content that undermines the shared values required for a healthy democracy.

Mitchell L. Gordon, Michelle S. Lam, Joon Sung Park et al.

Whose labels should a machine learning (ML) algorithm learn to emulate? For ML tasks ranging from online comment toxicity to misinformation detection to medical diagnosis, different groups in society may have irreconcilable disagreements about ground truth labels. Supervised ML today resolves these label disagreements implicitly using majority vote, which overrides minority groups' labels. We introduce jury learning, a supervised ML approach that resolves these disagreements explicitly through the metaphor of a jury: defining which people or groups, in what proportion, determine the classifier's prediction. For example, a jury learning model for online toxicity might centrally feature women and Black jurors, who are commonly targets of online harassment. To enable jury learning, we contribute a deep learning architecture that models every annotator in a dataset, samples from annotators' models to populate the jury, then runs inference to classify. Our architecture enables juries that dynamically adapt their composition, explore counterfactuals, and visualize dissent.

Jess Hohenstein, Dominic DiFranzo, Rene F. Kizilcec et al.

Artificial intelligence (AI) is now widely used to facilitate social interaction, but its impact on social relationships and communication is not well understood. We study the social consequences of one of the most pervasive AI applications: algorithmic response suggestions ("smart replies"). Two randomized experiments (n = 1036) provide evidence that a commercially-deployed AI changes how people interact with and perceive one another in pro-social and anti-social ways. We find that using algorithmic responses increases communication efficiency, use of positive emotional language, and positive evaluations by communication partners. However, consistent with common assumptions about the negative implications of AI, people are evaluated more negatively if they are suspected to be using algorithmic responses. Thus, even though AI can increase communication efficiency and improve interpersonal perceptions, it risks changing users' language production and continues to be viewed negatively.

Pranav Khadpe, Ranjay Krishna, Li Fei-Fei et al.

With the emergence of conversational artificial intelligence (AI) agents, it is important to understand the mechanisms that influence users' experiences of these agents. We study a common tool in the designer's toolkit: conceptual metaphors. Metaphors can present an agent as akin to a wry teenager, a toddler, or an experienced butler. How might a choice of metaphor influence our experience of the AI agent? Sampling metaphors along the dimensions of warmth and competence---defined by psychological theories as the primary axes of variation for human social perception---we perform a study (N=260) where we manipulate the metaphor, but not the behavior, of a Wizard-of-Oz conversational agent. Following the experience, participants are surveyed about their intention to use the agent, their desire to cooperate with the agent, and the agent's usability. Contrary to the current tendency of designers to use high competence metaphors to describe AI products, we find that metaphors that signal low competence lead to better evaluations of the agent than metaphors that signal high competence. This effect persists despite both high and low competence agents featuring human-level performance and the wizards being blind to condition. A second study confirms that intention to adopt decreases rapidly as competence projected by the metaphor increases. In a third study, we assess effects of metaphor choices on potential users' desire to try out the system and find that users are drawn to systems that project higher competence and warmth. These results suggest that projecting competence may help attract new users, but those users may discard the agent unless it can quickly correct with a lower competence metaphor. We close with a retrospective analysis that finds similar patterns between metaphors and user attitudes towards past conversational agents such as Xiaoice, Replika, Woebot, Mitsuku, and Tay.

Myle Ott, Claire Cardie, Jeff Hancock

Consumers' purchase decisions are increasingly influenced by user-generated online reviews. Accordingly, there has been growing concern about the potential for posting "deceptive opinion spam" -- fictitious reviews that have been deliberately written to sound authentic, to deceive the reader. But while this practice has received considerable public attention and concern, relatively little is known about the actual prevalence, or rate, of deception in online review communities, and less still about the factors that influence it. We propose a generative model of deception which, in conjunction with a deception classifier, we use to explore the prevalence of deception in six popular online review communities: Expedia, Hotels.com, Orbitz, Priceline, TripAdvisor, and Yelp. We additionally propose a theoretical model of online reviews based on economic signaling theory, in which consumer reviews diminish the inherent information asymmetry between consumers and producers, by acting as a signal to a product's true, unknown quality. We find that deceptive opinion spam is a growing problem overall, but with different growth rates across communities. These rates, we argue, are driven by the different signaling costs associated with deception for each review community, e.g., posting requirements. When measures are taken to increase signaling cost, e.g., filtering reviews written by first-time reviewers, deception prevalence is effectively reduced.