Abstract:
BILAI - Bilateral AI is a project that aims to develop bilateral AI systems with broader skill acquisition and problem solving capabilities than current systems that are applicable to a wide range of tasks. Neuro-symbolic AI, which combines sub-symbolic and symbolic AI methods, is still in its infancy, while it is one of the most promising directions to advance AI, but has with AlphaGo and AlphaFold already two big success stories, the later being awarded with the Nobel prize. In contrast to these specific problem solutions (narrow AI), BILAI pursues to improve essential properties of AI systems using neuro-symbolic approaches (broad AI). Enhancing essential properties of AI systems by bilateral approaches has a very high innovative potential. In this talk, we briefly present this project, in which experts from both AI research directions work in an unprecedented manner together, and its associated activities.