"Curiosity, Risk & Reward:
Shaping Autonomous Intelligence"
13-20 March 2020
Exploring the unknown is an important element of learning. Exploring the unknown also involves risks and rewards. The scientific study of learning and development in animals, humans, artificial intelligent agents and entire societies has investigated the roles of risk and reward in many ways. A century of intense research along these themes has led to the establishing of powerful scientific paradigms: behaviourism in psychology and neuroscience; optimal control theory in signal engineering; reinforcement learning as one of the pillars of modern machine learning; decision support systems in AI. A current trend is to also include into the picture the purest motivation for exploration: curiosity. Curiosity arguably triggers exploration without being tied to an expected specific reward. This IK explores (being itself both curiosity- and reward-driven!) the roles of curiosity, risk and reward (CRR for short) for the shaping of intelligence, both individual and social. The effects of CRR mechanisms can be observed on all levels from synaptic plasticity to conservatism vs. progressivism in societies. A wide spectrum of courses will illuminate how CRR is fundamental for long-term learning and development, how it serves as a basis for and potentially conflicts with decision-making, how creative curiosity can be captured in AI systems, how it is related to its antagonists caution and fear, and how it facilitates participatory behaviour patterns in societies.