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.

Student plagiarism is a major problem in universities worldwide. In this paper, we focus on plagiarism in answers to computer programming assignments, where students mix and/or modify one or more original solutions to obtain counterfeits. Although several software tools have been developed to help the tedious and time consuming task of detecting plagiarism, little has been done to assess their quality, because determining the real authorship of the whole submission corpus is practically impossible for markers. In this paper, we present a grammar evolution technique which generates benchmarks for testing plagiarism detection tools. Given a programming language, our technique generates a set of original solutions to an assignment, together with a set of plagiarisms of the former set which mimic the basic plagiarism techniques performed by students. The authorship of the submission corpus is predefined by the user, providing a base for the assessment and further comparison of copy-catching tools. We give empirical evidence of the suitability of our approach by studying the behavior of one advanced plagiarism detection tool (AC) on four benchmarks coded in APL2, generated with our technique.

In the last years, the computers have increased their capacityof calculus and networks - for the interconnectionof these machines - have been improved until obtaining theactual high rates of data transferring. The programs thatnowadays try to take advantage of these new technologies,cannot be written using the traditional techniques of programming,since most of the algorithms were designed forbeing executed in only one processor, in a nonconcurrentform, instead of being executed concurrently in a set of processors,working and communicating through a network.This work aims to present the ongoing development of a newmethod to simulate the Ferromagnetic Potts model, takinginto account these new technologies.

Recent large scale experiments have shown that the normalized information distance, an algorithmic information measure, is among the best similarity metrics for melody classification. This paper proposes the use of this distance as a fitness function which may be used by genetic algorithms to automatically generate music in a given pre-defined style. The minimization of this distance of the generated music to a set of musical guides makes it possible to obtain computer-generated music which recalls the style of a certain human author. The recombination operator plays an important role in this problem and thus several variations are tested to fine tune the genetic algorithm for this application. The superiority of the relative pitch envelope over other music parameters, such as the lengths of the notes, brought us to develop a simplified algorithm that nevertheless obtains interesting results.

Student plagiarism is a mayor problem in universities worldwide. In this paper,we focus on plagiarism in answers to computer programming assignments,where student mix and/or modify one or more original solutions to obtain counterfeits. Although several software tools have been implemented to help the tedious and time consuming task of detecting plagiarism, little has been done to assess their quality, because, in fact, determining the original subset of the whole solutionset is practically impossible for graders. In this article we present a Grammatical Evolution technique which generates benchmarks. Given a programming language, our technique generates a set of original solutions to an assignment, together with a set of plagiarisms of the former set which mimic the way in which students act. The phylogeny of the coded solutions is predefined, providing a base for the evaluationof the performance of copy-catching tools. We give empirical evidence of the suitability of our approach by studying the behavior of one state-of-the-art detection tool (AC) on four benchmarks coded in APL2, generated with this technique.

This paper describes Christiansen grammar evolution (CGE), a new evolutionary automatic programming algorithm that extends standard grammar evolution (GE) by replacing context-free grammars by Christiansen grammars. GE only takes into account syntactic restrictions to generate valid individuals. CGE adds semantics to ensure that both semantically and syntactically valid individuals are generated. It is empirically shown that our approach improves GE performance and even allows the solution of some problems are difficult to tackle by GE

This correspondence studies the influence of noise on the normalized compression distance (NCD), a measure based on the use of compressors to compute the degree of similarity of two files. This influence is approximated by a first order differential equation which gives rise to a complex effect, which explains the fact that the NCD may give values greater than 1, observed by other authors. The model is tested experimentally with good adjustment. Finally, the influence of noise on the clustering of files of different types is explored, finding that the NCD performs well even in the presence of quite high noise levels

This paper presents an agent-based model of an organization. The model is made of a social network—composed of the different organization workers—and a knowledge network. Workers are assigned tasks, for which they have to use information in the knowledge network. We have modeled the quality of the information by assigning each information item a probability of being wrong. Agents can interact with other agents, who can recommend to them new information items in the knowledge network for the task to be performed. Workers are assigned different information-seeking behavior (passive, active, and learning), governing the way in which they interact with each other. Moreover, indirect interaction is also possible, as a publicly accessible knowledge base contains each agent's preferred information items. The model was implemented in SDML, and its simulation shows that agents quickly learn to discern the better information items for the given task. However, group formation (agents' collaborating by exchanging information) takes longer to stabilize. Additionally, when the quality of items is changed, agents can quickly select the better new knowledge items, and organization performance improves again to a maximum that is only randomly disturbed.