Events

We examine how artificial intelligence impacts three core pillars of collaboration—performance enhancement, expertise integration, and social engagement—through a pre-registered field experiment with 791 professionals at Procter & Gamble, a global consumer packaged goods company. Working on real product innovation challenges, professionals were randomly assigned to work either with or without AI, and either individually or with another professional in new product development teams. Our findings show that (1) AI significantly enhances performance: individuals with AI matched the performance of teams without AI, suggesting that AI can effectively replicate certain benefits of human collaboration. Moreover, (2) AI helps bridge functional silos: without AI, R&D professionals tended to suggest more technical solutions, while Commercial professionals leaned toward commercially-oriented proposals. Professionals using AI produced more balanced solutions, regardless of their professional background. (3) AI’s language-based interface prompted more positive self-reported emotional responses among participants, suggesting it can fulfill part of the social and motivational role traditionally offered by human teammates. Finally, decomposing the innovation process suggests that AI primarily enhances the quality of generated ideas, shifting the distribution of creative output upward, while human judgment retains value in evaluative selection. This finding highlights the multiple and complementary roles that human and AI partners can play in new product development tasks and creative problem solving. More generally, our results suggest that AI adoption in knowledge work affects not only performance but also how expertise and sociality appear within teams, offering insights into the impact of GenAI on collaborative work within organizations.

Contact person: Marina Chugunova

Subscription to the invitation mailing list and more information on the seminar page.

The idea that inventions are solutions to specific technical problems is foundational to innovation research—yet, so far, there is no systematic way of measuring which problem an invention addresses. The problem that a battery overheats, for instance, may be solved through inventions in electrochemistry, materials science, or mechanical engineering. In existing patent data, these solutions would appear unrelated, because the data are organized around the technology employed, not the problem addressed. This missing problem dimension means we cannot observe which technical problems attract competing solutions from different fields, how broadly firms search across alternative solution paths, or how inventive labor is divided between those who recognize problems and those who develop solutions. This project constructs a new dataset that makes the link between inventions and the problems they address visible at scale.

Contact person: Elisabeth Hofmeister

Subscription to the invitation mailing list and more information on the seminar page.

More information

We use quasi-random local variation in the number of young men who died as a result of World War I to estimate the impact of this demographic shock on innovation and structural change in France. Our analysis shows that excess mortality led to an increase in patenting activity in counties with high pre-war education levels, driven predominantly by innovations in labor-saving technologies. Our estimates imply that an additional 6,000 patents were filed in the 15 years following the war, amounting roughly to the average annual number of patents filed pre-war.We find a positive association between war-related mortality and wage growth as well as with the adoption of machines in the agricultural sector, providing additional evidence that incentives to escape labor scarcity are driving the innovation response to mortality.

Contact person: Dominik Asam

Subscription to the invitation mailing list and more information on the seminar page.

We develop a method to estimate producers’ productivity beliefs in settings where output quantities and input prices are unobservable, and we use it to evaluate allocative efficiency in the market for science. Our model of researchers’ labor supply shows that their willingness to pay for their two key inputs, funding and time, reveals their underlying productivity beliefs. We estimate the model’s parameters using data from a nationally representative survey of research-active professors from all major fields of science. We find that the distribution of research productivity is highly skewed. Using these estimates, we assess the market’s allocative efficiency by comparing actual input allocations to optimal allocations given various objectives. Overall, the market for science is moderately efficient at maximizing output and researchers’ utility: actual input levels are positively correlated with the optimal levels implied by the model. However, the wedge between researchers’ actual and optimal input levels is often significant and difficult to predict. Our estimates imply that total budgets would need to increase by roughly 40% under actual allocations in order to achieve the same growth in scientific output that we predict under alternative allocations of the current budget. Scaling to the population level, this equates to billions of dollars in funding — there are substantial gains from developing new ways of identifying and supporting talented scientists.

Contact Person: Jordan Bisset

Subscription to the invitation mailing list and more information on the seminar page.

10:00–10:15  Opening

10:15–11:00  Interview Guide Priorities, Data Sources, Investment Estimationcore Discussion: Testing the Central Hypothesis, Scenario Ranking

11:00–12:00  Input Talk: Dr. Alexei Slobodianski (Thermo Fisher Scientific)(online)

12:00–13:00  Lunch

13:00–13:45  Synthesis: Revising Scenario Hierarchy

13:45–14:30  Methodology: Interview Guide Priorities, Data Sources, Investment Estimation

14:30–14:45  Closing Remarks

More information

More information