The German Centre for Research and Innovation- DWIH New Delhi organized the year’s first Science Circle Lecture on Thursday, 17th January 2019, on the theme: Simple Heuristics for a Complex World, at the Embassy of the Federal Republic of Germany, New Delhi. The lecture was delivered by Prof. Dr. Gerd Gigerenzer, Director of the Harding Centre for Risk Literacy at the Max Planck Institute for Human Development in Berlin. People from different walks of life including, but not limited to students, professors and industrialists were in attendance at the event.
Addressing the eclectic audience, Dr. Renate Schimkoreit, HoD, Economic and Global Affairs at the German Embassy, New Delhi, observed, “We spend most of our lives solving problems.”
Dr. Matthias Kiesselbach, Chairman of DWIH New Delhi went on to elaborate how decision making is a subject of interest in various disciplines including psychology, mathematics and economics and how heuristics play a decisive role in the process.
With this, he set the stage for Dr. Gerd Gigerenzer. At the very outset, Dr. Gigerenzer refuted the common notion that logic is the cornerstone of decision making. He cited examples from machine learning, including failures of Google Flu Trends and IBM’s Watson in Healthcare and Finance to highlight the success of heuristics in situations of uncertainty where, unlike in case of risk, fine-tuned optimization policies rarely work. Instances from sports and financial sectors added weight to his argument of less being more when it comes to complex problem solving.
He also added that people use heuristics owing to the accuracy-effort trade off.
Towards the end of his lecture, Prof. Gigerenzer warned against the hazards of relying too much on past data to build predictive models, a phenomenon referred to as the Turkey Illusion.
Themes ranging from the validity of intuition to the role of experience in decision making based on simple heuristics came up in the question answer round at the conclusion of the lecture. The talk charted a path for the evolving areas of AI and machine learning to profit from simple, robust heuristics rather than complex, but often futile analysis of multiple variables.