Challenges for Achieving Human-Level Intelligence

Kai-Uwe Kühnberger
Institute of Cognitive Science
University of Osnabrück, Germany
kkuehnbe@uos.de

ABSTRACT

Current state-of-the-art techniques in artificial intelligence for modeling human-level intelligence [1] face several problems ranging from the profusion of knowledge representation formalisms to a diffuse variety of reasoning techniques and the gap between symbolic and sub-symbolic approaches. In this talk, I will present some ongoing research endeavors of the AI group in Osnabrueck, by arguing that modeling a broad range of higher cognitive abilities requires certain non-standard computational paradigms. These paradigms include modules for analogical reasoning, the dynamic rewriting of ontological background knowledge, and neuro-symbolic integration.

Reference: Leake, D. et al. (eds.) 2006. AI magazine: Special Issue on Achieving Human-Level Intelligence through Integrated Systems and Research, AAAI.

AI at Universität Osnabrück
Colloquia Series page