Superintelligence is a hypothetical agent that possesses intelligence far surpassing that of the brightest and most gifted human minds. In light of recent advances in machine intelligence, a number of scientists, philosophers and technologists have revived the discussion about the potentially catastrophic risks entailed by such an entity. In this article, we trace the origins and development of the neo-fear of superintelligence, and some of the major proposals for its containment. We argue that total containment is, in principle, impossible, due to fundamental limits inherent to computing itself. Assuming that a superintelligence will contain a program that includes all the programs that can be executed by a universal Turing machine on input potentially as complex as the state of the world, strict containment requires simulations of such a program, something theoretically (and practically) impossible. This article is part of the special track on AI and Society.

Does semantic expression spread online from person to person? And if so, what kinds of expression are most likely to spread? To address these questions, we developed a nonexperimental, noninvasive method to detect and quantify contagion of semantic expression in massive online social networks, which we review and discuss here. Using only observational data, the method avoids performing emotional experiments on users of online social networks, a research practice that recently became an object of criticism and concern. Our model combines geographic aggregation and instrumental variables regression to measure the effect of an exogenous variable on an individual's expression and the influence of this change on the expression of others to whom that individual is socially connected. In a previous work, we applied our method to the emotional content of posts generated by a large sample of users over a period of three years. Those results suggest that each post expressing a positive or negative emotion can cause friends to generate one to two additional posts expressing the same emotion, and it also inhibits their use of the opposite emotion. Here, we generalize our method so it can be applied to contexts different than emotional expression and to different forms of content generated by the users of online platforms. The method allows us to determine the usage of words in the same semantic category spread, and to estimate a signed relationship between different semantic categories, showing that an increase in the usage of one category alters the usage of another category in one's social contacts. Finally, it also allows us to estimate the total cumulative effect that a person has on all of her social contacts.

The present work deals with the problem of information acquisition in a strategic networked environment. To study this problem, Kleinberg and Raghavan (FOCS 2005) introduced the model of query incentive networks, where the root of a binomial branching process wishes to retrieve an information -- known by each node independently with probability 1/n -- by investing as little as possible. The authors considered fixed-payment contracts in which every node strategically chooses an amount to offer its children paid upon information retrieval to convince them to seek the information in their subtrees. Kleinberg and Raghavan discovered that the investment needed at the root exhibits an unexpected threshold behavior that depends on the branching parameter b. For b>2, the investment is linear in the expected distance to the closest information (logarithmic in n, the rarity of the information), while, for 1<b<2, it becomes exponential in the same distance (i.e., polynomial in n). Arcaute et al. (EC 2007) later observed the same threshold behavior for arbitrary Galton-Watson branching processes.