It is now generally accepted that people are not always rational and often make decisions that are not aligned with their own best interest. How can we improve humans’ decision-making? A decade ago, Richard Thaler, a 2017 Nobel Laureate in Economics, has introduced the term “choice architecture” to describe different tools that can be used to improve humans’ decision-making without limiting their choices. Choice architecture can be used to motivate employees to save more for retirement, tax-payers to pay their taxes on time, diners to consume healthier food, consumers to make better purchases and more. Thaler’s choice-architecture is based on insights from behavioral economics and psychology. In the present study, we ask whether current methods of choice-architecture may be improved if they rely on computational models.

We specifically ask this question in the field of operant learning – learning from rewards and punishments, because in this field, scientists now describe learning and decision making using tools from computer science and artificial intelligence (AI). In the natural sciences, quantitative models underlay the development of engineering. Therefore, we ask whether quantitative models can revolutionize the field of choice architecture into choice engineering, defined as the use of quantitative models to shape choice behavior. To address this question, we launch an academic competition in which scientists will compete on the ability of their models to shape choice behavior.