SPEAKER

Prof. Daniel Elfenbein
Professor of Organization and Strategy
Olin Business School
Washington University in St. Louis

ABSTRACT

Research on the governance of new ventures has traditionally focused on mitigating incentive misalignment between founders and investors through boards’ incentive design, monitoring, and control rights. We identify a distinct governance problem: managing cognitive misalignment—systematic differences in beliefs about an opportunity’s underlying value. Cognitive misalignment arises from the inherent challenges of learning under uncertainty. As founders and boards gather information, their assessments of an opportunity’s merits may diverge because they interpret evidence differently and bring distinct biases to learning and decision making. We examine how boards address this challenge through cognitive governance, defined as how boards allocate their limited attention between independently evaluating the opportunity and learning about the founder’s confidence bias, and how this allocation shapes how boards interpret founder input when exercising decision rights. We use a computational model of founder–board interaction to explore when investors are better served by focusing more attention on understanding founders than independently evaluating opportunities, how optimal governance depends on the prevalence of overconfidence in the founder population, and why boards create more value when working with moderately overconfident founders. The results yield testable predictions for board composition, investor-founder matching, and portfolio construction.

SPEAKER

Prof. Leon Musolff
Assistant Professor of Business Economics and Public Policy
The Wharton School
The University of Pennsylvania

ABSTRACT

We study how productivity effects evolve across successive generations of AI coding tools and across stages of the software production process, using data on more than 100,000 GitHub developers combined with internal telemetry. In an event-study design with matched controls, we find that each successive generation of AI coding tools—autocomplete, interactive coding agents, and autonomous coding agents—significantly increases coding activity, with respective cumulative effects of 40%, 140%, and 180%. However, these effects attenuate sharply across production stages: for lines of code, the effect is 1,630%, but falls to 180% for commits, and 30% for releases. We show that this pattern is consistent with a multi-stage production process in which AI automates code generation but downstream tasks—review, integration, and coordination—remain human-bottlenecked. Aggregate data from three software marketplaces corroborate this result: large developer-level gains in task productivity only translate into modest increases in software output.

SPEAKER

Prof. Cynthia Lee
Distinguished Professor Emeritus
Management and Organizational Development
Northeastern University

ABSTRACT

While workplace territoriality research has received considerable scholarly attention and bloomed over the past decade, critical gaps remain in three areas: the conceptualization of territorial behaviors (with extant work focusing on territorial marking and defending, overlooking territorial expansion), the drivers of those behaviors, and how AI reshapes workplace boundary dynamics. This talk presents three research papers that address these omissions and advance territoriality theory. The first paper, published in the Journal of Applied Psychology, extends the territoriality construct by introducing a territoriality expanding dimension. Drawing on conservation of resources and regulatory focus theory, it adopts a resource-based perspective to propose a double-edged effect on job performance, mediated by all three forms of territoriality and information exchange, and moderated by individual regulatory focus. The second, an ongoing study, examines the intersection of job insecurity and territoriality, exploring how boundaryless career orientation and territorial behavior together shape job search activity. The third paper examines these dynamics through the lens of AI adoption, investigating whether AI presents a threat that triggers territorial defense or an opportunity that drives territorial expansion, and how each outcome influences workplace loneliness at varying degrees of psychological safety.

Together, these papers invite future work to move beyond its current theorizing of territoriality as a static, defensive behavior and reconceptualize it as a dynamic and proactive strategy individuals adopt in response to prevailing organizational challenges such as mass layoffs, AI disruption, and beyond.

SPEAKER

Professor Simon Jäger
Associate Professor of Economics and Public Affairs
Princeton University

ABSTRACT

Discrete choice experiments are widely used to estimate workers’ willingness to pay (WTP) for job amenities under the assumption that workers evaluate each attribute in isolation. We test this assumption by embedding an amenity with a known market price — a well-known monthly public transport pass — in a large-scale choice experiment with 6,000 German workers. Stated WTP for the pass is roughly twice its market price, while WTP for other amenities matches prior estimates.  A complementary belief-elicitation experiment shows that advertising any amenity causally shifts beliefs about unlisted characteristics of the employer, including in unstructured text that we analyze with an LLM.  Posted wages similarly signal the amenity bundle so that wage variation, the money metric for WTP calculation, is itself contaminated by belief spillovers. In general, workers infer that higher pay comes with better amenities but also more stressful work environments. The belief spillovers we document imply that individual-amenity WTP estimates capture perceived bundles rather than isolated attributes. We discuss implications for the measurement of non-wage compensation and the estimation of monopsony power.

SPEAKER

Prof. Jordan Siegel
Professor of Strategy
Ross School of Business
University of Michigan

ABSTRACT

This study examines the extent to which rival social movements centered on gender-related issues have been associated with changes in firms’ managerial gender composition.  Using the rise of the MeToo movement (2019-2021) and the subsequent Anti-MeToo countermovement (2021-2023) in South Korea as an empirical setting, we focus on changes in women’s representation across managerial levels across firms.  We combine our statistical results from the Korea Labor Institute’s Workplace Panel Survey data with evidence from our qualitative fieldwork conducted during 2025-2026.  Results show that these social movements are associated with changes in firms’ hiring and promotion, that larger firms, in particular, increased their representation of women at all hierarchical levels of management during the peak era of the MeToo movement (2019-2021), and that during the anti-MeToo movement (2021-2023), there was a separating equilibrium among the relatively larger firms.  One subset of larger firms continued to uniformly increase their representation of women at all hierarchical levels of management despite the anti-MeToo social movement.  In contrast, another subset of larger firms continued to increase their representation of women at the very top levels of their organization, while at the same time increasing their representation of males/decreasing their representation of women from the entry managerial level to the beginning of the senior managerial levels.  The latter pattern is consistent with the possibility that some firms adopted a form of organizational compromise in response to competing societal pressures generated by the MeToo movement and the subsequent Anti-MeToo countermovement.  These two subsets of larger firms are similarly populated.  Lastly, the average change in the representation of women across the totality of all managerial levels was significantly positive for larger firms even during the Anti-MeToo period.

SPEAKER

Dr. Donghyun Suh
Economist Research
Department
Bank of Korea

ABSTRACT

This paper develops a model of hierarchical production organizations to study the effects of technological change on income distribution, focusing on top labor incomes. The model features workers with different skill levels who interact with machines. Machine complexity determines how machines are organized inside the hierarchy and, through that channel, whether they augment or substitute for workers. Two main findings emerge. First, if machines only perform sufficiently simple tasks, they augment low-skilled workers and attenuate the “superstar effect” by flattening the upper tail of the income distribution. Second, if machines become sufficiently complex, then they substitute for low-skilled workers and augment high-skilled workers, strengthening the superstar effect. Lastly, I examine future AI systems that automate managerial tasks performed by high-skilled workers. AI managers reduce inequality within and across occupations, with the largest gains for the least skilled. However, this equalizing effect need not survive superintelligence; once machines surpass all humans and supervise everyone, further advances can widen inequality. The results highlight the importance of machine complexity and supervision costs for understanding the distributional effects of technology.

SPEAKER

Prof. Michal Kosinski
Associate Professor of Organizational Behavior
Graduate School of Business
Stanford University

ABSTRACT

Large Language Models (LLMs) trained to predict the next word in a sentence have surprised their creators by displaying emergent properties, ranging from a proclivity for biases to an ability to write computer code and solve mathematical tasks. This talk discusses the results of several studies evaluating LLMs’ performance on tasks typically used to study human psychological processes. Findings indicate that as LLMs increase in size and linguistic ability, they can navigate false-belief scenarios, sidestep semantic illusions, and tackle cognitive reflection tasks. This talk explores what these emergent properties reveal about the nature of intelligence—human and artificial—and what they might mean for the future of technology, society, and our understanding of the mind itself.

SPEAKER

Prof. Milan Miric
Dean’s Associate Professorship in Business Administration
Associate Professor of Data Sciences and Operations
Marshall School of Business
University of Southern California

ABSTRACT

Entrepreneurship is highly uncertain, and many ventures are likely to fail. Thus, potential entrepreneurs likely consider the impact founder experience may have on their future career. Given that gender has substantial impacts on both career progression and entrepreneurial performance, we hypothesize that female and male entrepreneurs may have substantially different outcomes when they re-enter traditional employment, potentially serving as a hidden roadblock to female participation in entrepreneurship. Using data from LinkedIn’s Economic Graph Research Program, we find that when re-entering traditional employment after founding a firm former, female founders are more likely to regress and less likely to advance in their next role in traditional employment than former male founders. This significant gender penalty for female entrepreneurs, however, obscures important variation that allows us to point to potential mechanisms. We find that male entrepreneurs are substantially more likely to advance in their career relative to similar female entrepreneurs when their startup showed significant growth, but that the gender penalty for female entrepreneurs is dependent upon the gender composition of the hiring firm. Female entrepreneurs hired by firms with greater female representation are more likely to advance in their career relative to men whereas female entrepreneurs hired by firms with less female representation are less likely to advance relative to similar male entrepreneurs.

SPEAKER

Prof. Ricardo Alonso
Professor of Management
London School of Economics and Political Science

ABSTRACT

We study how a privately informed decision maker uses recommendations from an AI system. We treat the AI as a black box guaranteeing a minimum level of performance when its advice is followed. A conservative Bayesian decision maker combines the AI recommendation with her signal to maximize the worst-case performance consistent with that guarantee. By mapping this robust inference problem into a Bayesian persuasion problem, we characterize the conditions under which the decision maker ignores the AI, combines its recommendation with her own information (“complementary AI”), or fully delegates to it (“substitutive AI”). By reinterpreting the robust inference problem as an optimal reliance problem, we derive explicitly the value of the AI’s recommendation. We show that the value of AI depends on the form of expertise. For example, in a classification problem, a prescriptive expert—one who knows which class is more likely—typically either ignores the AI or delegates to it, whereas a proscriptive expert—one who knows which class is less likely—always finds AI complementary and benefits more from its advice. We characterize which types of experts benefit the most from Human-AI interaction.

SPEAKER

Professor Robert Akerlof
Professor of Economics
UNSW Business School

ABSTRACT

Standard economic theory assumes that people have perfect information processing ability: they can derive all logical consequences of information immediately. In reality, however, people struggle to work out the consequences of information. In this paper, we offer a theory of the reasoning process—how people go about “connecting the dots.” In our theory, an agent may have all of the information needed to draw a conclusion yet they still fail to see it. Our key assumption is that agents have limited “working memory.” This constrains the number of pieces of information—“dots”— an agent can think about and connect at once. We explore how agents analyze pieces of information and work their way to a big picture—or, “narrative”—and how narratives, in turn, shape the agent’s view of the parts. We show that limited working memory makes sense of why people struggle with tradeoffs, suffer from choice overload, are influenced by defaults, engage in satisficing, selectively attend to attributes, are subject to primacy effects, and may vacillate between interpretations. We apply the framework to political persuasion, showing how supplying a simple narrative can lock in an interpretation of ambiguous evidence.