We propose a new measure of intelligence for general reinforcement learning agents, based on the notion that an agent’s environment can change at any step of execution of the agent. That is, an agent is considered to be interacting with its environment in real-time. In this sense, the resulting intelligence measure is more general than the universal intelligence measure (Legg and Hutter, 2007) and the anytime universal intelligence test (Hernández-Orallo and Dowe, 2010). A major advantage of the measure is that an agent’s computational complexity is factored into the measure in a natural manner. We show that there exist agents with intelligence arbitrarily close to the theoretical maximum, and that the intelligence of agents depends on their parallel processing capability. We thus believe that the measure can provide a better evaluation of agents and guidance for building practical agents with high intelligence.
If the inline PDF is not rendering correctly, you can download the PDF file here.
Aanderaa, S. O. 1974. On k-tape versus (k -1)-tape real-time computation. In Karp, R. M., ed., Complexity of Computation (SIAM-AMS Proceedings), volume 7, 75-96. American Mathematical Society, Providence, Rhode Island.
Aaronson, S., and Ambainis, A. 2005. Quantum search of spatial regions. Theory of Computing 1:47-79.
Hernández-Orallo, J., and Dowe, D. L. 2010. Measuring universal intelligence: Towards an anytime intelligence test. Artificial Intelligence 174(18):1508-1539.
Hutter, M. 2005. Universal Artificial Intelligence: Sequential Decisions based on Algorithmic Probability. Springer, Berlin.
Hutter, M. 2007. On universal prediction and Bayesian confirmation. Theoretical Computer Science 384(1):33-48.
Lattimore, T., and Hutter, M. 2011. Time consistent discounting. In Kivinen, J.; Szepesv´ari, C.; Ukkonen, E.; and Zeugmann, T., eds., Proceedings of The 22nd International Conference on Algorithmic Learning Theory (ALT’11), volume 6925 of LNAI, 383-397. Espoo, Finland: Springer, Berlin.
Legg, S., and Hutter, M. 2007. Universal intelligence: A definition of machine intelligence. Minds & Machines 17(4):391-444.
Li, M., and Vit´anyi, P. M. B. 2008. An Introduction to Kolmogorov Complexity and Its Applications. Springer, New York, third edition.
Paul, W. J.; Seiferas, J. I.; and Simon, J. 1981. An information-theoretic approach to time bounds for on-line computation. Journal of Computer and System Sciences 23:108-126.
Rabaey, J. M.; Chandrakasan, A.; and Nikolic, B. 2003. Digital Integrated Circuits. Prentice Hall, second edition.
Rosenberg, A. L. 1967. Real-time definable languages. Journal of the Association for Computing Machinery 14(4):645-662.
Schmidhuber, J. 1997. A computer scientist’s view of life, the universe, and everything. In Freksa, C.; Jantzen, M.; and Valk, R., eds., Foundations of Computer Science: Potential - Theory - Cognition, volume 1337 of Lecture Notes in Computer Science. Springer, Berlin. 201-208.
Vitányi, P. M. B. 1980. On the power of real-time Turing machines under varying specifications. In Proceedings of the 7th International Colloquium on Automata, Languages and Programming (ICALP’80), volume 85 of Lecture Notes in Computer Science, 658-671. Springer, Berlin.