Comparing Humans and AI Agents
Comparing humans and machines is one important source of information about both machine and human strengths and limitations. Most of these comparisons and competitions are performed in rather specific tasks such as calculus, speech recognition, translation, games, etc. The information conveyed by these experiments is limited, since it portrays that machines are much better than humans at some domains and worse at others. In fact, CAPTCHAs exploit this fact. However, there have only been a few proposals of general intelligence tests in the last two decades, and, to our knowledge, just a couple of implementations and evaluations. In this paper, we implement one of the most recent test proposals, devise an interface for humans and use it to compare the intelligence of humans and Q-learning, a popular reinforcement learning algorithm. The results are highly informative in many ways, raising many questions on the use of a (universal) distribution of environments, on the role of measuring knowledge acquisition, and other issues, such as speed, duration of the test, scalability, etc.
KeywordsIntelligence measurement universal intelligence general vs specific intelligence reinforcement learning IQ tests
Unable to display preview. Download preview PDF.
- 1.Dowe, D.L., Hajek, A.R.: A non-behavioural, computational extension to the Turing Test. In: Intl. Conf. on Computational Intelligence & multimedia applications (ICCIMA 1998), Gippsland, Australia, pp. 101–106 (1998)Google Scholar
- 2.Gordon, D., Subramanian, D.: A cognitive model of learning to navigate. In: Proc. 19th Conf. of the Cognitive Science Society, 1997, vol. 25, p. 271. Lawrence Erlbaum, Mahwah (1997)Google Scholar
- 4.Hernández-Orallo, J.: A (hopefully) non-biased universal environment class for measuring intelligence of biological and artificial systems. In: Hutter, M., et al. (eds.) 3rd Intl. Conf. on Artificial General Intelligence, pp. 182–183. Atlantis Press, London (2010) Extended report at, http://users.dsic.upv.es/proy/anynt/unbiased.pdf Google Scholar
- 6.Hernández-Orallo, J., Dowe, D.L., España-Cubillo, S., Hernández-Lloreda, M.V., Insa-Cabrera, J.: On more realistic environment distributions for defining, evaluating and developing intelligence. In: Schmidhuber, J., Thórisson, K.R., Looks, M. (eds.) AGI 2011. LNCS(LNAI), pp. 81–90. Springer, Heidelberg (2011)Google Scholar
- 7.Legg, S., Hutter, M.: A universal measure of intelligence for artificial agents. In: Intl Joint Conf on Artificial Intelligence, IJCAI, vol. 19, p. 1509 (2005)Google Scholar
- 11.Sanghi, P., Dowe, D.L.: A computer program capable of passing IQ tests. In: 4th Intl. Conf. on Cognitive Science (ICCS 2003), Sydney, pp. 570–575 (2003)Google Scholar
- 13.Strehl, A.L., Li, L., Wiewiora, E., Langford, J., Littman, M.L.: PAC model-free reinforcement learning. In: ICML 2006, pp. 881–888. New York (2006)Google Scholar
- 14.Sutton, R.S., Barto, A.G.: Reinforcement learning: An introduction. The MIT press, Cambridge (1998)Google Scholar