Bounded Recursive Self-Improvement

We have designed a machine that becomes increasingly better at behaving in underspecified circumstances, in a goal-directed way, on the job, by modeling itself and its environment as experience accumulates. Based on principles of autocatalysis, endogeny, and reflectivity, the work provides an architectural blueprint for constructing systems with high levels of operational autonomy in underspecified circumstances, starting from a small seed. Through value-driven dynamic priority scheduling controlling the parallel execution of a vast number of reasoning threads, the system achieves recursive self-improvement after it leaves the lab, within the boundaries imposed by its designers. A prototype system has been implemented and demonstrated to learn a complex real-world task, real-time multimodal dialogue with humans, by on-line observation. Our work presents solutions to several challenges that must be solved for achieving artificial general intelligence.

[1]  Kristinn R. Thórisson,et al.  Holistic Intelligence: Transversal Skills & Current Methodologies , 2009 .

[2]  T. Anguera,et al.  From Communication to Presence: Cognition, emotions and culture towards the ultimate communicative experience , 2006 .

[3]  Kristinn R. Thórisson From Constructionist to Constructivist A.I , 2009, AAAI Fall Symposium: Biologically Inspired Cognitive Architectures.

[4]  Allen Newell,et al.  Report on a general problem-solving program , 1959, IFIP Congress.

[5]  W. McGrew An ethological study of children's behavior , 1972 .

[6]  Kristinn R. Thórisson,et al.  Evaluating multimodal human-robot interaction: a case study of an early humanoid prototype , 2010, MB '10.

[7]  Bas R. Steunebrink,et al.  Towards an Actual Gödel Machine Implementation: a Lesson in Self-Reflective Systems , 2012 .

[8]  Tamas Madl,et al.  LIDA: A Systems-level Architecture for Cognition, Emotion, and Learning , 2014, IEEE Transactions on Autonomous Mental Development.

[9]  Pei Wang,et al.  Rigid Flexibility: The Logic of Intelligence , 2006 .

[10]  Ben Goertzel,et al.  Theoretical Foundations of Artificial General Intelligence , 2012, Atlantis Thinking Machines.

[11]  Kristinn R. Thórisson,et al.  Achieving Artificial General Intelligence Through Peewee Granularity , 2009 .

[12]  R. Sanz,et al.  Fridges, elephants, and the meaning of autonomy and intelligence , 2000, Proceedings of the 2000 IEEE International Symposium on Intelligent Control. Held jointly with the 8th IEEE Mediterranean Conference on Control and Automation (Cat. No.00CH37147).

[13]  Jürgen Schmidhuber,et al.  Reinforcement Learning with Self-Modifying Policies , 1998, Learning to Learn.

[14]  E. Schegloff,et al.  A simplest systematics for the organization of turn-taking for conversation , 1974 .

[15]  Kristinn R. Thórisson,et al.  Self-Programming: Operationalizing Autonomy , 2009 .

[16]  Marcus Hutter,et al.  Universal Artificial Intellegence - Sequential Decisions Based on Algorithmic Probability , 2005, Texts in Theoretical Computer Science. An EATCS Series.

[17]  E. W. Adams,et al.  Models of Man, Social and Rational: Mathematical Essays on Rational Human Behavior in a Social Setting , 1962 .

[18]  Kristinn R. Thórisson,et al.  Attention Capabilities for AI Systems , 2012, ICINCO.

[19]  Pei Wang,et al.  THE ASSUMPTIONS ON KNOWLEDGE AND RESOURCES IN MODELS OF RATIONALITY , 2011 .

[20]  John E. Laird,et al.  The Soar Cognitive Architecture , 2012 .

[21]  M. Bromberg,et al.  Analyse de la structure interactionnelle et des stratégies discursives dans un talk-show , 1993 .

[22]  Ben Goertzel,et al.  Proceedings of the Second Conference on Artificial General Intelligence , 2009 .

[23]  Jürgen Schmidhuber,et al.  Gödel Machines: Fully Self-referential Optimal Universal Self-improvers , 2007, Artificial General Intelligence.

[24]  Jürgen Schmidhuber,et al.  Resource-Bounded Machines are Motivated to be Effective, Efficient, and Curious , 2013, AGI.

[25]  Helgi Páll Helgason,et al.  General Attention Mechanism for Artificial Intelligence Systems , 2013 .

[26]  M. Magnússon Hidden Real-Time Patterns in Intra- and Inter-Individual Behavior: Description and Detection , 1996 .

[27]  H. Simon,et al.  A Behavioral Model of Rational Choice , 1955 .

[28]  M S Magnusson,et al.  Discovering hidden time patterns in behavior: T-patterns and their detection , 2000, Behavior research methods, instruments, & computers : a journal of the Psychonomic Society, Inc.

[29]  Joel Veness,et al.  A Monte-Carlo AIXI Approximation , 2009, J. Artif. Intell. Res..