Concept Learning for Safe Autonomous AI

Sophisticated autonomous AI may need to base its behavior on fuzzy concepts such as well-being or rights. These concepts cannot be given an explicit formal definition, but obtaining desired behavior still requires a way to instill the concepts in an AI system. To solve the problem, we review evidence suggesting that the human brain generates its concepts using a relatively limited set of rules and mechanisms. This suggests that it might be feasible to build AI systems that use similar criteria for generating their own concepts, and could thus learn similar concepts as humans do. Major challenges to this approach include the embodied nature of human thought, evolutionary vestiges in cognition, the social nature of concepts, and the need to compare conceptual representations between humans and AI systems.

[1]  Nick Bostrom,et al.  Superintelligence: Paths, Dangers, Strategies , 2014 .

[2]  Terry Regier,et al.  Word Meanings across Languages Support Efficient Communication , 2015 .

[3]  Timo Honkela,et al.  Tkk Reports in Information and Computer Science Gica: Grounded Intersubjective Concept Analysis a Method for Enhancing Mutual Understanding and Participation Gica: Grounded Intersubjective Concept Analysis a Method for Enhancing Mutual Understanding and Participation , 2022 .

[4]  L. Barsalou,et al.  Embodiment in Attitudes, Social Perception, and Emotion , 2005, Personality and social psychology review : an official journal of the Society for Personality and Social Psychology, Inc.

[5]  T. Abeles.  Moral Machines: Teaching Robots Right from Wrong , 2010 .

[6]  Terry Regier,et al.  Spatial terms across languages support near-optimal communication: Evidence from Peruvian Amazonia, and computational analyses , 2013, CogSci.

[7]  Marcello Guarini,et al.  Particularism and the Classification and Reclassification of Moral Cases , 2006, IEEE Intelligent Systems.

[8]  Nick Bostrom,et al.  Thinking Inside the Box: Controlling and Using an Oracle AI , 2012, Minds and Machines.

[9]  L. Cosmides,et al.  The Adapted mind : evolutionary psychology and the generation of culture , 1992 .

[10]  Timo Honkela,et al.  Subjects on objects in contexts: Using GICA method to quantify epistemological subjectivity , 2012, The 2012 International Joint Conference on Neural Networks (IJCNN).

[11]  Lior Shamir,et al.  Computer analysis of art , 2012, JOCCH.

[12]  S. Dehaene,et al.  Cultural Recycling of Cortical Maps , 2007, Neuron.

[13]  Roman V Yampolskiy,et al.  Responses to catastrophic AGI risk: a survey , 2014 .

[14]  Charles Kemp,et al.  How to Grow a Mind: Statistics, Structure, and Abstraction , 2011, Science.

[15]  Charles Kemp,et al.  The discovery of structural form , 2008, Proceedings of the National Academy of Sciences.

[16]  Ben Goertzel,et al.  Stages of Ethical Development in Artificial General Intelligence Systems , 2008, AGI.

[17]  Peter Gärdenfors,et al.  Conceptual spaces - the geometry of thought , 2000 .

[18]  Colin Allen,et al.  Prolegomena to any future artificial moral agent , 2000, J. Exp. Theor. Artif. Intell..

[19]  D. Chalmers.  The Singularity: a Philosophical Analysis , 2010 .

[20]  Eliezer Yudkowsky.  Artificial Intelligence as a Positive and Negative Factor in Global Risk , 2006 .

[21]  N. Kriegeskorte,et al.  Author ' s personal copy Representational geometry : integrating cognition , computation , and the brain , 2013 .

[22]  A Prince,et al.  Optimality: From Neural Networks to Universal Grammar , 1997, Science.

[23]  L. Cosmides,et al.  Cognitive adaptations for social exchange. , 1992 .

[24]  Timo Honkela,et al.  Simulating processes of concept formation and communication , 2008 .

[25]  G. Lakoff,et al.  Where mathematics comes from : how the embodied mind brings mathematics into being , 2002 .

[26]  Ben Goertzel,et al.  Nine Ways to Bias Open-Source AGI Toward Friendliness , 2012 .

[27]  R. Poldrack,et al.  Measuring neural representations with fMRI: practices and pitfalls , 2013, Annals of the New York Academy of Sciences.

[28]  G. Clore,et al.  Disgust as Embodied Moral Judgment , 2008, Personality and Social Psychology Bulletin.

[29]  H. Barrett,et al.  Modularity in cognition: framing the debate. , 2006, Psychological review.

[30]  Eliezer Yudkowsky.  Complex Value Systems are Required to Realize Valuable Futures , 2011 .