Artificial intelligence, idea evaluation and social relationships in innovation management
|Researcher:||Linus Dahlander (ESMT Berlin)|
Innovation is the source of competitiveness for large technology-based companies. To succeed in the competitive innovation race companies need to understand how innovative ideas travel through the organization, what facilitate the flow of ideas and how innovation can be done more quickly at a reasonable cost. How companies manage their flow of ideas is thus at the core of this challenge. This project investigates how individuals evaluate ideas proposed for innovation and how artificial intelligence systems affect the managers doing such evaluation.
The project intends to provide causal evidence on how managers evaluate ideas, and would thus like to conduct field experiments within Ericsson where internal employees generate ideas for innovation and senior managers evaluate them. The field experimental approach has three major advantages that can advance research on innovation where experiments have been rare to date. First, behavior and outcomes are observed in a natural setting rather than an artificial laboratory environment, which enhances external validity. Second, the behavior in well-designed field experiments has consequences ensuring that individuals have real stakes in the decisions they make. Third, randomization of treatment and control status enables causal inference. And causal inference is important to help managers make better decisions. For instance, firemen are not causing fires just because they go to where fires are raging.
We will randomly assign ideas to senior managers and will ask them to vote on the idea. To analyze the role of social relationships for evaluation outcomes, we will additionally randomize the amount of information available on the employee who proposed the idea. Similarly, we will vary information on whether an AI-system was used to prescreen the ideas before the manager receives it. The experimental design ensures that we can casually explain how social relationships and AI-systems impacts on which ideas managers select.