Merrick Osborne’s research while affiliated with Minnesota School of Business and other places

Artificial intelligence (“AI”) has the potential to vastly improve human decision-making. In line with this, researchers have increasingly sought to understand how people view AI, often documenting skepticism and even outright aversion to these tools. In the present research, we complement these findings by documenting the performance of LLMs in the personal advice domain. In addition, we shift the focus in a new direction—exploring how interacting with AI tools, specifically large language models, impacts the user’s view of themselves. In five preregistered experiments (N = 1,722), we explore evaluations of human- and ChatGPT-generated advice along three dimensions: quality, effectiveness, and authenticity. We find that ChatGPT produces superior advice relative to the average online participant even in a domain in which people strongly prefer human-generated advice (dating and relationships). We also document a bias against ChatGPT-generated advice which is present only when participants are aware the advice was generated by ChatGPT. Novel to the present investigation, we then explore how interacting with these tools impacts self-evaluations. We manipulate the order in which people interact with these tools relative to self-generation and find that generating advice before interacting with ChatGPT advice boosts the quality ratings of the ChatGPT advice. At the same time, interacting with ChatGPT-generated advice before self-generating advice decreases self-ratings of authenticity. Taken together, we document a bias towards AI in the context of personal advice. Further, we identify an important externality in the use of these tools—they can invoke social comparisons of me vs. the machine.

Since 2022, generative AI (genAI) has rapidly become integrated into workplaces. Though organizations have made commitments to use this technology "responsibly", how organizations and their employees prioritize responsibility in their decision-making remains absent from extant management theorizing. In this paper, we examine how product managers - who often serve as gatekeepers in decision-making processes - implement responsible practices in their day-to-day work when using genAI. Using Institutional Theory, we illuminate the factors that constrain or support proactive responsible development and usage of genAI technologies. We employ a mixed methods research design, drawing on 25 interviews with product managers and a global survey of 300 respondents in product management-related roles. The majority of our respondents report (1) widespread uncertainty regarding what "responsibility" means or looks like, (2) diffused responsibility given assumed ethical actions by other teams, (3) lack of clear incentives and guidance within organizations, and (4) the importance of leadership buy-in and principles for navigating tensions between ethical commitments and profit motives. However, our study finds that even in highly uncertain environments, absent guidance from leadership, product managers can "recouple" ethical commitments and practices by finding responsibility "micro-moments". Product managers seek out low-risk, small-scale actions they can take without explicit buy-in from higher-level managers, such as individual or team-wide checks and reviews and safeguarding standards for data. Our research highlights how genAI poses unique challenges to organizations trying to couple ethical principles and daily practices and the role that middle-level management can play in recoupling the two.

Technological innovations have become a key driver of societal advancements. Nowhere is this more evident than in the field of machine learning (ML), which has developed algorithmic models that shape our decisions, behaviors, and outcomes. These tools have widespread use, in part, because they can synthesize massive amounts of data to make seemingly objective recommendations. Yet, in the past few years, the ML community has been raising the alarm on why we should be cautious in interpreting and using these models: they are created by humans, from data generated by humans, whose psychology allows for various biases that impact how the models are developed, trained, tested and interpreted. As psychologists, we thus face a fork in the road; Down the first path, we can continue to use these models without examining and addressing these critical flaws, and rely on computer scientists to try to mitigate them. Down the second path, we can turn our expertise in bias towards this growing field, collaborating with computer scientists to mitigate the deleterious outcomes associated with these models. This paper serves to light the way down the second path by identifying how extant psychological research can help examine and mitigate bias in ML models.

The current research explores how local racial diversity affects Whites' efforts to structure their local communities to avoid incidental intergroup contact. In two experimental studies (N = 509; Studies 1a-b), we consider Whites' choices to structure a fictional, diverse city and find that Whites choose greater racial segregation around more (vs. less) self-relevant landmarks (e.g., their workplace and children's school). Specifically, the more time they expect to spend at a landmark, the more they concentrate other Whites around that landmark, thereby reducing opportunities for incidental intergroup contact. Whites also structure environments to reduce incidental intergroup contact by instituting organizational policies that disproportionately exclude non-Whites: Two large-scale archival studies (Studies 2a-b) using data from every U.S. tennis (N = 15,023) and golf (N = 10,949) facility revealed that facilities in more racially diverse communities maintain more exclusionary barriers (e.g., guest policies, monetary fees, dress codes) that shield the patrons of these historically White institutions from incidental intergroup contact. In a final experiment (N = 307; Study 3), we find that Whites' anticipated intergroup anxiety is one driver of their choices to structure environments to reduce incidental intergroup contact in more (vs. less) racially diverse communities. Our results suggest that despite increasing racial diversity, White Americans structure local environments to fuel a self-perpetuating cycle of segregation.