The missing neurocognitive and artificial general intelligence bases of robocup reasearch: What still needs to be done before 2050?
Tarik Hadzibeganovic, M.W.M.L. van den Noort, Sergio Alejandro Cannas, M.P.C. Bosch, Gudberg Konrad Jonsson, Magnusson · Radboud Repository (Radboud University) · 2008
There has been much recent interest in cognitively motivated paradigms for the design of humanoid soccer playing robots.In addition, some proto-attempts have been made to bring the 3D simulation and humanoid RoboCup leagues closer to each other.In the present paper, we outline the basic concepts of our Neurocognitive Artificial General Intelligence (NAGI) framework for the future 3D Soccer Simulation League.The aim of NAGI is to develop teachable AI soccer playing agents embodied in virtual environments such as Second Life.Once virtually embodied agents attain approximately human-like intelligence, the transfer of agent modeling knowledge to the humanoid robots should take place.