COURNOT Seminar – Bernard Sinclair-Desgagné (Skema Business School – Côte d’Azur University)
![](https://www.beta-economics.fr/uploads/2021/04/Compo_presentation_BETA_2021_04_29-e1619762308835.png)
Le 22/10/2021
De 14:00 à 15:30
Détails de l'événement :
« Handling the unknown unknowns of innovation »
Abstract : Innovation is replete with unknown unknowns, widely acknowledged ones occur early on at the stage of invention, when exploration leads naturally to navigating uncharted territory. Notable additional ones might also show up at the diffusion stage, when the targeted users prove to be largely skeptical of the asserted benefits of new products or technologies. These unknown unknowns constitute major hurdles which R&D project managers, marketers and policy makers have to deal with. Yet, despite substantial advances in the economics of uncertainty, decision theory and contract theory, the understanding and handling of unknown unknowns remain elusive to both applied research scholars and practitioners. In this seminar, I will present two recent developments of mine on the matter.
The first development draws on the documented tendency of some people to shy away from using algorithms, even when such machines would likely enhance their livelihood or job performance. The literature refers to this behavior as algorithm aversion. It is often imputable to people’s intrinsic reluctance to surrender control to a machine they see as a “black box”. To capture what goes on, i.e. the choice to delegate or not decision-making to an entity endowed with superior but abstruse cognitive ability, I will build a first model of a benevolent agent who can control what are unknown unknowns to the principal. This will yield new empirical propositions and policy recommendations for technology diffusion.
The second development has to do with product innovation and the generation of ideas. I will present and illustrate a systematic approach which I recently proposed for thinking out of the box. This approach relies on Formal Concept Analysis (FCA), a recent data-mining technique which is increasingly used to represent and treat knowledge. In practice, it involves using standard spreadsheets. In theory, it brings to light the logic of discovery.