Peeking behind the screen: the unsuspected power of the standard Turing Test

No computer that had not experienced the world as we humans had could pass a rigorously administered standard Turing Test. This paper will show that the use of ‘subcognitive’ questions allows the standard Turing Test to indirectly probe the human subcognitive associative concept network built up over a lifetime of experience with the world. Not only can this probing reveal differences in cognitive abilities, but crucially, even differences in physical aspects of the candidates can be detected. Consequently, it is unnecessary to propose even harder versions of the Test in which all physical and behavioural aspects of the two candidates had to be indistinguishable before allowing the machine to pass the Test. Any machine that passed the ‘simpler’ symbols-in symbols-out test as originally proposed by Turing would be intelligent. The problem is that, even in its original form, the Turing Test is already too hard and too anthropocentric for any machine that was not a physical, social and behavioural carbon copy of ourselves to actually pass it. Consequently, the Turing Test, even in its standard version, is not a reasonable test for general machine intelligence. There is no need for an even stronger version of the Test.     

Other bodies,other minds: A machine incarnation of an old philosophical problem

Any attempt to explain the mind by building machines with minds must confront the other-minds problem: How can we tell whether any body other than our own has a mind when the only way to know is by being the other body? In practice we all use some form of Turing Test: If it can do everything a body with a mind can do such that we can't tell them apart, we have no basis for doubting it has a mind. But what is “everything” a body with a mind can do? Turing's original “pen-pal” version of the Turing Test (the TT) only tested linguistic capacity, but Searle has shown that a mindless symbol-manipulator could pass the TT undetected. The Total Turing Test (TTT) calls instead for all of our linguistic and robotic capacities; immune to Searle's argument, it suggests how to ground a symbol manipulating system in the capacity to pick out the objects its symbols refer to. No Turing Test, however, can guarantee that a body has a mind. Worse, nothing in the explanation of its successful performance requires a model to have a mind at all. Minds are hence very different from the unobservables of physics (e.g., superstrings); and Turing Testing, though essential for machine-modeling the mind, can really only yield an explanation of the body.     

Imitation Versus Communication: Testing for Human-Like Intelligence

Turing’s Imitation Game is often viewed as a test for theorised machines that could ‘think’ and/or demonstrate ‘intelligence’. However, contrary to Turing’s apparent intent, it can be shown that Turing’s Test is essentially a test for humans only. Such a test does not provide for theorised artificial intellects with human-like, but not human-exact, intellectual capabilities. As an attempt to bypass this limitation, I explore the notion of shifting the goal posts of the Turing Test, and related tests such as the Total Turing Test, away from the exact imitation of human capabilities, and towards communication with humans instead. While the continued philosophical relevance of such tests is open to debate, the outcome is a different class of tests which are, unlike the Turing Test, immune to failure by means of sub-cognitive questioning techniques. I suggest that attempting to instantiate such tests could potentially be more scientifically and pragmatically relevant to some Artificial Intelligence researchers, than instantiating a Turing Test, due to the focus on producing a variety of goal directed outcomes through communicative methods, as opposed to the Turing Test’s emphasis on ‘fooling’ an Examiner.

Development of a single-wheeled inverted pendulum robot capable of climbing stairs

ABSTRACT

In this paper, we propose the design of a single-wheeled robot capable of climbing stairs. The robot is equipped with the proposed climbing mechanism, which enables it to climb stairs. The mechanism has an extremely simple structure, comprised of a parallel arm, belt, harmonic drive, and pulley. The proposed climbing mechanism has the advantage of not requiring an additional actuator because it can be driven by using a single actuator that drives the wheel. The robot is equipped with a control moment gyroscope to control the stability in a lateral direction. Experimental results demonstrate that the robot can climb stairs with a riser height of 12–13?cm and a tread depth of 39?cm at an approximate rate of 2 to 3 s for each step.     

A Note on Maxwell's Demon and Universal Computation

Abstract

An analogy between a Maxwellian Demon capable of regulating the passage of particles between two chambers and a Turing machine capable of manipulating a tape and its symbols, is made explicit. It is shown that a slightly modified Maxwell's Demon can simulate a Universal Turing Machine.     

Making the Right Identification in the Turing Test1

The test Turing proposed for machine intelligence is usually understood to be a test of whether a computer can fool a human into thinking that the computer is a human. This standard interpretation is rejected in favor of a test based on the Imitation Game introduced by Turing at the beginning of "Computing Machinery and Intelligence."     

The Cartesian Test for Automatism1

In Part V of his Discourse on the Method, Descartes introduces a test for distinguishing people from machines that is similar to the one proposed much later by Alan Turing. The Cartesian test combines two distinct elements that Keith Gunderson has labeled the language test and the action test. Though traditional interpretation holds that the action test attempts to determine whether an agent is acting upon principles, I argue that the action test is best understood as a test of common sense. I also maintain that this interpretation yields a stronger test than Turing's, and that contemporary artificial intelligence should consider using it as a guide for future research.     

Consequences of nonclassical measurement for the algorithmic description of continuous dynamical systems

Abstract

Continuous dynamical systems intuitively seem capable of more complex behavior than discrete systems. If analyzed in the framework of the traditional theory of computation, a.continuous dynamical system with countably many quasistable states has at least the computational power of a universal Turing machine. Such an analysis assumes, however, the classical notion of measurement. If measurement is viewed nonclassically, a continuous dynamical system cannot, even in principle, exhibit behavior that cannot be simulated by a universal Turing machine.     

Self-improving AI: an Analysis

Self-improvement was one of the aspects of AI proposed for study in the 1956 Dartmouth conference. Turing proposed a “child machine” which could be taught in the human manner to attain adult human-level intelligence. In latter days, the contention that an AI system could be built to learn and improve itself indefinitely has acquired the label of the bootstrap fallacy. Attempts in AI to implement such a system have met with consistent failure for half a century. Technological optimists, however, have maintained that a such system is possible, producing, if implemented, a feedback loop that would lead to a rapid exponential increase in intelligence. We examine the arguments for both positions and draw some conclusions.

Development and field test of the articulated mobile robot T2 Snake-4 for plant disaster prevention

Abstract

In this work, we develop an articulated mobile robot that can move in narrow spaces, climb stairs, gather information, and operate valves for plant disaster prevention. The robot can adopt a tall position using a folding arm and gather information using sensors mounted on the arm. In addition, this paper presents a stair climbing method using a single backward wave. This method enables the robot to climb stairs that have a short tread. The developed robot system is tested in a field test at the World Robot Summit 2018, and the lessons learned in the field test are discussed.     

The complexity of ranking simple languages

Ranking is the problem of computing for an input string its lexicographic index in a given (fixed) language. This paper concerns the complexity of ranking. We show that ranking languages accepted by 1-way unambiguous auxiliary pushdown automata operating in polynomial time is inNC ⁽²⁾. We also prove negative results about ranking for several classes of simple languages.C is rankable in deterministic polynomial time iffP=P ^#P , whereC is any of the following six classes of languages: (1) languages accepted by logtime-bounded nondeterministic Turing machines, (2) languages accepted by (uniform) families of unbounded fan-in circuits of constant depth and polynomial size, (3) languages accepted by 2-way deterministic pushdown automata, (4) languages accepted by multihead deterministic finite automata, (5) languages accepted by 1-way nondeterministic logspace-bounded Turing machines, and (6) finitely ambiguous linear context-free languages.This research was partially supported by the National Science Foundation under Grant DCR-8696097. A preliminary version of this paper was presented at the 3rd Annual Structure in Complexity Theory Conference, Washington, DC, June 1988.     

Computing Machines Can't Be Intelligent (...and Turing Said So)

According to the conventional wisdom, Turing (1950) said that computing machines can be intelligent. I don't believe it. I think that what Turing really said was that computing machines –- computers limited to computing –- can only fake intelligence. If we want computers to become genuinelyintelligent, we will have to give them enough initiative (Turing, 1948, p. 21) to do more than compute. In this paper, I want to try to develop this idea. I want to explain how giving computers more ``initiative' can allow them to do more than compute. And I want to say why I believe (and believe that Turing believed) that they will have to go beyond computation before they can become genuinely intelligent.     

Commentary on Alan M. Turing: The Applications of Probability to Cryptography

Abstract

In April 2012, two papers written by Alan Turing during the Second World War on the use of probability in cryptanalysis were released by GCHQ. The longer of these presented an overall framework for the use of Bayes's theorem and prior probabilities, including four examples worked out in detail: the Vigenère cipher, a letter subtractor cipher, the use of repeats to find depths, and simple columnar transposition. (The other paper was an alternative version of the section on repeats.) Turing stressed the importance in practical cryptanalysis of sometimes using only part of the evidence or making simplifying assumptions and presents in each case computational shortcuts to make burdensome calculations manageable. The four examples increase roughly in their difficulty and cryptanalytic demands. After the war, Turing's approach to statistical inference was championed by his assistant in Hut 8, Jack Good, which played a role in the later resurgence of Bayesian statistics.     

低资源语言的无监督语音关键词检测技术综述

目的 低资源(low-resource)语言的无监督的关键词检测技术近年来引起了广泛的研究兴趣.低资源语言由于缺乏足够的标注数据及相关的专家知识,使得传统的基于大词汇量语音识别系统的关键词检测技术无法使用.近年来,研究者试图寻找一种无监督的技术来完成针对低资源语言的语音关键词检测.方法 首先阐述了该技术目前面临的问题与挑战,然后介绍了该技术使用的主流的基于动态时间规整的算法框架,并从特征表示、模板匹配方法、效率提升等几个重要方面介绍了近几年来主要的研究成果,最后介绍了该任务常用的系统评价标准及目前所能达到的水平,讨论了未来可能的研究方向.结果 该任务的研究目前取得了很多成果,但仍处于实验室阶段,多系统融合策略导致系统庞大,而且目前还没有好的进行索引的方法,导致检测时间过长,对于低资源语音的关键词检测技术,还有很多研究工作要做.结论 期望通过对目前低资源语言的无监督的关键词检测技术做出一个全面的综述,从而给研究者的工作带来便利.     

图灵机和图灵测试

阿兰?图灵为人工智能学的诞生做出了重大的贡献,本文介绍了图灵机和图灵测试,图灵机对计算机的结构、可实现性和局限性都产生了深远的影响,而图灵测试为机器能否思考的争论双方找到了一种公认的判决准则。     

PessimalPrint: a reverse Turing test

Abstract. We exploit the gap in ability between human and machine vision systems to craft a family of automatic challenges that tell human and machine users apart via graphical interfaces including Internet browsers. Turing proposed [Tur50] a method whereby human judges might validate “artificial intelligence” by failing to distinguish between human and machine interlocutors. Stimulated by the “chat room problem” posed by Udi Manber of Yahoo!, and influenced by the CAPTCHA project [BAL00] of Manuel Blum et al. of Carnegie-Mellon Univ., we propose a variant of the Turing test using pessimal print: that is, low-quality images of machine-printed text synthesized pseudo-randomly over certain ranges of words, typefaces, and image degradations. We show experimentally that judicious choice of these ranges can ensure that the images are legible to human readers but illegible to several of the best present-day optical character recognition (OCR) machines. Our approach is motivated by a decade of research on performance evaluation of OCR machines [RJN96,RNN99] and on quantitative stochastic models of document image quality [Bai92,Kan96]. The slow pace of evolution of OCR and other species of machine vision over many decades [NS96,Pav00] suggests that pessimal print will defy automated attack for many years. Applications include `bot' barriers and database rationing. Received: February 14, 2002 / Accepted: March 28, 2002 An expanded version of: A.L. Coates, H.S. Baird, R.J. Fateman (2001) Pessimal Print: a reverse Turing Test. In: {\it Proc. 6th Int. Conf. on Document Analysis and Recognition}, Seattle, Wash., USA, September 10–13, pp. 1154–1158 Correspondence to: H. S. Baird     

Exact Algorithms for Linear Programming over Algebraic Extensions

Abstract. We study the computational complexity of linear programs with coefficients that are real algebraic numbers under a Turing machine model of computation. After reviewing a method for exact representation of algebraic numbers under the Turing model, we show that the fundamental tasks of comparison and arithmetic can be performed in polynomial time. Our technique for establishing polynomial-time algorithms for comparison and arithmetic is distinct from the usual resultant-based approaches, and has the advantage that it provides a natural framework for analysis of the complexity of computational tasks, such as Gaussian elimination, that involve a sequence of arithmetic operations. Our main contribution is to show that a variant of the ellipsoid method can be used to solve linear programming in time polynomial in the encoding size of the problem coefficients and the degree of any algebraic extension that contains those coefficients.     

A New Characterization of NP, P, and PSPACE with Accepting Hybrid Networks of Evolutionary Processors

We consider three complexity classes defined on Accepting Hybrid Networks of Evolutionary Processors (AHNEP) and compare them with the classical complexity classes defined on the standard computing model of Turing machine. By definition, AHNEPs are deterministic. We prove that the classical complexity class NP equals the family of languages decided by AHNEPs in polynomial time. A language is in P if and only if it is decided by an AHNEP in polynomial time and space. We also show that PSPACE equals the family of languages decided by AHNEPs in polynomial length.