Endgame studies have long served as a tool for testing human creativity and intelligence. We find that they can serve as a tool for testing machine ability as well. Two of the leading chess engines, Stockfish and Leela Chess Zero (LCZero), employ significantly different methods during play. We use Plaskett's Puzzle, a famous endgame study from the late 1970s, to compare the two engines. Our experiments show that Stockfish outperforms LCZero on the puzzle. We examine the algorithmic differences between the engines and use our observations as a basis for carefully interpreting the test results. Drawing inspiration from how humans solve chess problems, we ask whether machines can possess a form of imagination. On the theoretical side, we describe how Bellman's equation may be applied to optimize the probability of winning. To conclude, we discuss the implications of our work on artificial intelligence (AI) and artificial general intelligence (AGI), suggesting possible avenues for future research.
The Digital Synaptic Neural Substrate: A New Approach to Computational Creativity
Universiti Tenaga Nasional(马来西亚国家能源大学);University of Ljubljana(卢布尔雅那大学);Goldsmiths College, University of London(伦敦大学金史密斯学院);Jozef Stefan Institute(约瑟夫·斯特凡研究所)
研究人员一直在讨论什么在预测某些结果方面更优越:统计方法还是人脑。人工智能(AI)领域的显着技术进步一再引发这场辩论,例如解决对象和语音识别等任务,通过深度学习算法实现准确性的显着提高(Goodfellow et al. 2016),或结合各种计算智能方法,如模糊逻辑,遗传算法,基于案例的推理(Medsker 2012)。这些进步背后的隐含承诺之一是,机器有一天将能够执行复杂的任务,甚至可能取代人类执行这些任务。这引发了关于机器何时最终取代人类的新的激烈辩论(McAfee and Brynjolfsson 2017)。虽然先前的研究已经证明,人工智能在一些明确定义的任务中表现良好,如下棋,下围棋或识别图像上的物体,但人们怀疑在不久的将来能够同时解决多个任务的人工通用智能(AGI)的发展(例如,Russell and Norvig 2016)。此外,在组织环境中使用人工智能来解决复杂的业务问题的情况很少发生,解决复杂问题的人工智能应用程序主要停留在实验室环境中,而不是在实践中实施。由于通往AGI的道路仍然很长,我们认为未来几十年人类和机器之间最有可能的分工模式是混合智能。这一概念旨在利用人类智能和人工智能的互补优势,使它们能够比两者单独表现得更好(例如,2016)。
Machine Intelligence at the Edge with Learning Centric Power Allocation
Southern University of Science and Technology(南方科技大学);The University of Hong Kong(香港大学);Sun Yat-sen University(中山大学);Southern Marine Science and Engineering Guangdong Laboratory(南方海洋科学与工程广东省实验室)
Brigham Young University(杨百翰大学);Masdar Institute of Science and Technology(马斯达尔科学与技术学院);British University in Dubai(迪拜英国大学);University of Edinburgh(爱丁堡大学);Toulouse School of Economics(图卢兹经济学院);Center for Research in Management(管理研究中心);Centre National de la Recherche Scientifique(法国国家科学研究中心);Institute for Advanced Study in Toulouse(图卢兹高等研究院);University of Toulouse Capitole(图卢兹第一大学);Data61(数据61);Commonwealth Scientific and Industrial Research Organization(澳大利亚联邦科学与工业研究组织);University of California, Irvine(加州大学欧文分校);The Media Lab(媒体实验室);Massachusetts Institute of Technology(麻省理工学院)