Keywords: Goal, computational modelling
Abstract: Current artificial intelligence models severely suffer from the problems of poor
generality. Deep neural networks trained on quantities of data can only handle a
specific task, sometimes even under many constraints. The reason for this phe-
nomenon lies in the fact that current models can not model expressive concepts
like functionality, rationality goals and etc. In this article, we will mainly focus on
how to model goals in a computational framework and we hope this will contribute
to a more generalized artificial intelligence model.
Submission Number: 146
Loading