DOCSLIB.ORG

Autonome Waffensysteme

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Autonome Waffensysteme Reinhard Grünwald Christoph Kehl Autonome Waffensysteme Endbericht zum TA-Projekt Oktober 2020 Arbeitsbericht Nr. 187 Autonome Waffensysteme Das Büro für Technikfolgen-Abschätzung beim Deutschen Bundestag (TAB) berät den Deutschen Bundestag und seine Ausschüsse in Fragen des wissen- schaftlich-technischen Wandels. Das TAB wird seit 1990 vom Institut für Tech- nikfolgenabschätzung und Systemanalyse (ITAS) des Karlsruher Instituts für Technologie (KIT) betrieben. Hierbei kooperiert es seit September 2013 mit dem IZT – Institut für Zukunftsstudien und Technologiebewertung gGmbH sowie der VDI/VDE Innovation + Technik GmbH. Reinhard Grünwald Christoph Kehl Autonome Waffensysteme Endbericht zum TA-Projekt TAB-Arbeitsbericht Nr. 187 Büro für Technikfolgen-Abschätzung beim Deutschen Bundestag Neue Schönhauser Straße 10 10178 Berlin Telefon: +49 30 28491-0 E-Mail: [email protected] Web: www.tab-beim-bundestag.de 2020 Umschlagbild: Bertrand Bouchez/unsplash.com Papier: Circleoffset Premium White Druck: Wienands Print + Medien GmbH, Bad Honnef ISSN-Print: 2364-2599 ISSN-Internet: 2364-2602 Inhalt Zusammenfassung 9 1 Einleitung 29 2 Abgrenzung des Untersuchungsgegenstands 33 2.1 Definitorische Ansätze 34 2.2 Autonom, semiautonom oder automatisiert? 35 2.3 Die Definition des US-Verteidigungsministeriums 39 2.4 Die Qualität menschlicher Kontrolle über AWS 40 3 Technische Grundlagen von Autonomie 45 3.1 Autonomie aus informationstechnologischer Sicht 45 3.2 Autonome Funktionen aktueller militärischer Systeme 46 3.2.1 Autonomie für Mobilität 46 3.2.2 Autonomie für Zielerkennung/Zielbestimmung 47 3.2.3 Autonomie für Informationsgewinnung 48 3.2.4 Autonomie für die Fähigkeit zur Zusammenarbeit 49 3.3 Künstliche Intelligenz als Schlüsseltechnologie 50 3.3.1 Maschinelles Lernen 52 3.3.2 Was können KI-Systeme heute leisten? 53 3.3.3 Begrenzungen, Schwierigkeiten und Risiken bei KI-Systemen 56 4 Verbreitung, Status und Trends unbemannter Waffensysteme 67 4.1 Überblick zu einsatzreifen unbemannten (teil)autonomen Waffensystemen 70 4.1.1 Fliegende Systeme 71 4.1.2 Bodensysteme 76 4.1.3 Systeme zu Wasser 82 4.2 Forschungs- und Entwicklungstrends 85 4.2.1 USA 86 4.2.2 Europa 93 4.2.3 Weitere Schlüsselakteure 99 5 Inhalt 5 Einsatzszenarien 101 5.1 Argumente für AWS 101 5.2 Erwartete militärische Fähigkeiten 104 5.3 Missionen/Einsatzszenarien 110 5.3.1 Erwartete Missionen 111 5.3.2 Denkbare militärische Einsatzszenarien für AWS 114 5.3.3 Zwischenfazit 119 6 Sicherheitspolitische Implikationen von AWS 125 6.1 Mehr oder weniger Kriege? 126 6.2 Veränderung der Kriegsführung 127 6.3 Destabilisierende Wirkung in Krisen 130 6.4 Auswirkungen auf regionale Stabilität 131 6.5 Auswirkungen auf das strategische Gleichgewicht 132 6.6 Rüstungsdynamiken 133 6.7 Unkontrollierte Weiterverbreitung 133 6.8 Technologische Risiken 134 7 Humanitäres Völkerrecht und autonome Waffensysteme 137 7.1 Prüfungspflicht (Artikel 36 ZP I) 138 7.2 AWS im Lichte der Prinzipien des HVR 139 8 Ethische Fragestellungen im Kontext autonomer Waffensysteme 149 8.1 AWS und die Ethik des Krieges 150 8.1.1 Die Lehre vom gerechten Krieg 151 8.1.2 Ermöglichen AWS eine ethischere Kriegsführung? 153 8.2 AWS und die Würde des Menschen 162 8.2.1 Der Begriff der Menschenwürde 162 8.2.2 Verletzt der Einsatz autonomer Waffensysteme die Menschenwürde? 164 8.3 Die Frage der Verantwortung 168 8.3.1 Rechtliche Sicht 170 8.3.2 Moralische Sicht 173 8.4 Fazit 175 6 Inhalt 9 Möglichkeiten der Rüstungskontrolle 179 9.1 Rüstungs- und Exportkontrollabkommen mit Relevanz für AWS 179 9.1.1 Rüstungskontrollverträge 180 9.1.2 Transparenz und vertrauens- und sicherheits- bildende Maßnahmen 184 9.1.3 Nichtverbreitung und Exportkontrolle 185 9.2 Die Konvention über bestimmte konventionelle Waffen 189 9.2.1 Menschliche Kontrolle über AWS 191 9.2.2 Positionen wichtiger Staaten bzw. Organisationen 194 9.2.3 Gemeinsamer Vorschlag von Deutschland und Frankreich auf der CCW GGE 207 9.2.4 Ausgangslage für die weitere Diskussion im Rahmen der CCW 209 9.3 Regulierungsansätze der präventiven Rüstungs- und Exportkontrolle 212 9.3.1 Vertrauens- und sicherheitsbildende Maßnahmen 216 9.3.2 Exportkontrolle 218 9.3.3 Verbindliche Regulierung bzw. Verbot von AWS 219 9.4 Handlungsmöglichkeiten 222 9.4.1 Die Möglichkeiten innerhalb der CCW ausschöpfen 223 9.4.2 Engagement über die CCW hinaus verbreitern 224 9.5 Fazit 228 10 Literatur 229 10.1 In Auftrag gegebene Gutachten 229 10.2 Weitere Literatur 229 11 Anhang 261 11.1 Abbildungen 261 11.2 Tabellen 261 11.3 Kästen 261 11.4 Abkürzungen 262 7 8 Zusammenfassung Robotische Waffensysteme, die ohne menschliches Zutun Ziele auswählen und bekämpfen können, waren vor nicht allzu langer Zeit ausschließlich in der Do- mäne der Science-Fiction beheimatet. Die enormen technologischen Fort- schritte, die in den letzten Jahren in den Bereichen der Robotik und der künst- lichen Intelligenz (KI) erzielt wurden, haben diese Vorstellung autonom agie- render Waffen nun an die Schwelle zur konkreten Umsetzung gerückt. Automatisierung und Autonomie werden bereits heute für eine breite Pa- lette an Funktionen bei Waffensystemen genutzt. Dazu gehören die Suche und Identifizierung potenzieller Ziele mithilfe von Sensordaten, die Zielverfolgung, Priorisierung und Bestimmung des Zeitpunkts für den Angriff auf diese Ziele sowie die Steuerung für den Zielanflug (z. B. einer Rakete oder eines Marsch- flugkörpers). Bislang erfolgen jedoch die Zielauswahl, die Angriffsentscheidung und schließlich die Freigabe des Waffeneinsatzes durch einen menschlichen Kommandeur bzw. Operator. Ein autonomes Waffensystem (AWS) wäre in der Lage, alle diese Schritte selbsttätig und ohne menschliche bzw. mit nur minimaler menschlicher Mit- wirkung durchzuführen. Dies hätte zwei entscheidende militärische Vorteile: Zum einen benötigt ein autonomes System keine Kommunikationsverbindung mit einer Basisstation, zum anderen erlaubt es schnellere Reaktionszeiten in Ge- fechtssituationen, da keine Verzögerungen durch die Laufzeiten einer Daten- übertragung und durch die Entscheidungsfindung bzw. die Reaktionszeiten ei- nes menschlichen Operators auftreten. Die Steigerung der Autonomie von Waffensystemen steht daher in allen technologisch fortgeschrittenen Ländern auf der Agenda. Sehr weit verbreitet ist die Auffassung, dass neue Anwendungen künstlicher Intelligenz im Begriff seien, sämtliche Wirtschafts- und Lebensbereiche grund- legend zu transformieren. In aktuellen verteidigungspolitischen und militäri- schen Strategiepapieren und Verlautbarungen etlicher Länder wird die Erwar- tung formuliert, dass diese Transformation auch vor dem Militärsektor nicht Halt machen werde und dass die zügige Einführung von KI-gestützten Waffen- systemen einen entscheidenden militärischen Vorsprung verspreche. Befürworter dieser Entwicklung argumentieren, dass mit AWS unter Um- ständen auch humanitäre Vorteile verbunden seien, da militärische Operationen präziser durchgeführt und somit die Zivilbevölkerung und zivile Infrastruktu- ren besser geschützt werden könnten. Kritiker äußern hingegen Bedenken, ob es ethisch vertretbar, politisch verantwortbar und (völker)rechtlich erlaubt sein könne, die Entscheidung über Leben und Tod von Menschen an Maschinen zu delegieren. Auch wären mit der Entwicklung und dem möglichen Einsatz von 9 Zusammenfassung AWS sicherheitspolitische Risiken sowie die Gefahr von Rüstungsspiralen und unkontrollierter Verbreitung potenziell riskanter Technologien verbunden. Mit dem vorliegenden Bericht wird ein breiter Analyseansatz verfolgt und eine Vielzahl von Facetten des Themas abgedeckt: die Darstellung des techno- logischen Reifegrads und der Entwicklungsperspektiven von AWS, eine Be- standsaufnahme von existierenden und in der Entwicklung befindlichen Syste- men sowie eine Analyse möglicher Einsatzszenarien und sich daraus ergebender sicherheitspolitischer Implikationen. Hinzu kommen ethische und völkerrecht- liche Fragestellungen, die eng miteinander zusammenhängen. Schließlich wer- den die aktuell vor allem im Rahmen der »Convention on Certain Conventional Weapons«1 (CCW) der Vereinten Nationen (United Nations – UN) angesiedel- ten diplomatischen Bemühungen um eine Einhegung der mit AWS möglicher- weise verbundenen Risiken beleuchtet und daraus Überlegungen zu Möglich- keiten der präventiven Rüstungskontrolle abgeleitet. Definition bzw. Abgrenzung des Untersuchungsgegenstands Die Frage der Definition von AWS birgt eine erhebliche Brisanz, da sie oftmals in einen direkten Zusammenhang mit möglichen Vereinbarungen zur Rüs- tungskontrolle gestellt wird. Dabei wird implizit oder explizit angenommen, dass die Definition von AWS gleichzeitig den Rahmen setzt, welche Systeme ggf. zu regulieren oder gar zu verbieten sind. Daher spiegelt die Haltung der Staaten und anderer Akteure in definitorischen Fragen regelmäßig deren Eigeninteres- sen und Verhandlungspositionen wider. International werden diverse Ansätze verfolgt, um AWS zu definieren und von anderen Waffensystemen abzugrenzen. Eine präzise, allgemein akzeptierte Definition existiert bis heute nicht. Ob ein Waffensystem autonom, semiauto- nom oder (hoch)automatisiert agieren kann, hängt nicht nur von seinen tech- nologischen Eigenschaften ab, sondern auch von der Komplexität der Umge- bung, in der es eingesetzt werden soll, sowie ganz entscheidend von der Qualität der Interaktion zwischen menschlichem Operator und technischem System. Die Zuordnung von Waffensystemen in die eine oder andere Kategorie ist in vielen Fällen strittig. Für die hier im Zentrum stehende Analyse
Load more

Recommended publications
  • Grid Computing for Artificial Intelligence
    0 10 Grid Computing for Artificial Intelligence Yuya Dan Matsuyama University Japan 1. Introduction This chapter is concerned in grid computing for artificial intelligence systems. In general, grid computing enables us to process a huge amount of data in any fields of discipline. Scientific computing, calculation of the accurate value of π, search for Mersenne prime numbers, analysis of protein molecular structure, weather dynamics, data analysis, business data mining, simulation are examples of application for grid computing. As is well known that supercomputers have very high performance in calculation, however, it is quite expensive to use them for a long time. On the other hand, grid computing using idle resources on the Internet may be running on the reasonable cost. Shogi is a traditional game involving two players in Japan as similar as chess, which is more complicated than chess in rule of play. Figure 1 shows the set of Shogi and initial state of 40 pieces on 9 x 9 board. According to game theory, both chess and Shogi are two-person zero-sum game with perfect information. They are not only a popular game but also a target in the research of artificial intelligence. Fig. 1. Picture of Shogi set including 40 pieces ona9x9board. There are the corresponding king, rook, knight, pawn, and other pieces. Six kinds of pieces can change their role like pawns in chess when they reach the opposite territory. 202 Advances in Grid Computing The information systems for Shogi need to process astronomical combinations of possible positions to move. It is described by Dan (4) that the grid systems for Shogi is effective in reduction of computational complexity with collaboration from computational resources on the Internet.
    [Show full text]
  • Shogi Yearbook 2015
    Shogi Yearbook 2015 SHOGI24.COM SHOGI YEARBOOK 2015 Title match games, Challenger’s tournaments, interviews with WATANABE Akira and HIROSE Akihito, tournament reports, photos, Micro Shogi, statistics, … This yearbook is a free PDF document Shogi Yearbook 2015 Content Content Content .................................................................................................................................................... 2 Just a few words ... .................................................................................................................................. 5 64. Osho .................................................................................................................................................. 6 64. Osho league ................................................................................................................................... 6 64th Osho title match ........................................................................................................................... 9 Game 1 ............................................................................................................................................. 9 Game 2 ........................................................................................................................................... 12 Game 3 ........................................................................................................................................... 15 Game 4 ..........................................................................................................................................
    [Show full text]
  • Large-Scale Optimization for Evaluation Functions with Minimax Search
    Journal of Artificial Intelligence Research 49 (2014) 527-568 Submitted 10/13; published 03/14 Large-Scale Optimization for Evaluation Functions with Minimax Search Kunihito Hoki [email protected] Department of Communication Engineering and Informatics The University of Electro-Communications Tomoyuki Kaneko [email protected] Department of Graphics and Computer Sciences The University of Tokyo Abstract This paper presents a new method, Minimax Tree Optimization (MMTO), to learn a heuristic evaluation function of a practical alpha-beta search program. The evaluation function may be a linear or non-linear combination of weighted features, and the weights are the parameters to be optimized. To control the search results so that the move de- cisions agree with the game records of human experts, a well-modeled objective function to be minimized is designed. Moreover, a numerical iterative method is used to find local minima of the objective function, and more than forty million parameters are adjusted by using a small number of hyper parameters. This method was applied to shogi, a major variant of chess in which the evaluation function must handle a larger state space than in chess. Experimental results show that the large-scale optimization of the evaluation function improves the playing strength of shogi programs, and the new method performs significantly better than other methods. Implementation of the new method in our shogi program Bonanza made substantial contributions to the program's first-place finish in the 2013 World Computer Shogi Championship. Additionally, we present preliminary evidence of broader applicability of our method to other two-player games such as chess.
    [Show full text]
  • Including ACG8, ACG9, Games in AI Research, ACG10 T/M P. 18) Version: 20 June 2007
    REFERENCE DATABASE 1 Updated till Vol. 29. No. 2 (including ACG8, ACG9, Games in AI Research, ACG10 t/m p. 18) Version: 20 June 2007 AAAI (1988). Proceedings of the AAAI Spring Symposium: Computer Game Playing. AAAI Press. Abramson, B. (1990). Expected-outcome: a general model of static evaluation. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 12, No.2, pp. 182-193. ACF (1990), American Checkers Federation. http://www.acfcheckers.com/. Adelson-Velskiy, G.M., Arlazarov, V.L., Bitman, A.R., Zhivotovsky, A.A., and Uskov, A.V. (1970). Programming a Computer to Play Chess. Russian Mathematical Surveys, Vol. 25, pp. 221-262. Adelson-Velskiy, M., Arlazarov, V.L., and Donskoy, M.V. (1975). Some Methods of Controlling the Tree Search in Chess Programs. Artificial Ingelligence, Vol. 6, No. 4, pp. 361-371. ISSN 0004-3702. Adelson-Velskiy, G.M., Arlazarov, V. and Donskoy, M. (1977). On the Structure of an Important Class of Exhaustive Problems and Methods of Search Reduction for them. Advances in Computer Chess 1 (ed. M.R.B. Clarke), pp. 1-6. Edinburgh University Press, Edinburgh. ISBN 0-85224-292-1. Adelson-Velskiy, G.M., Arlazarov, V.L. and Donskoy, M.V. (1988). Algorithms for Games. Springer-Verlag, New York, NY. ISBN 3-540-96629-3. Adleman, L. (1994). Molecular Computation of Solutions to Combinatorial Problems. Science, Vol. 266. p. 1021. American Association for the Advancement of Science, Washington. ISSN 0036-8075. Ahlswede, R. and Wegener, I. (1979). Suchprobleme. Teubner-Verlag, Stuttgart. Aichholzer, O., Aurenhammer, F., and Werner, T. (2002). Algorithmic Fun: Abalone. Technical report, Institut for Theoretical Computer Science, Graz University of Technology.
    [Show full text]
  • Improving Alphazero Using Monte-Carlo Graph Search
    Proceedings of the Thirty-First International Conference on Automated Planning and Scheduling (ICAPS 2021) Improving AlphaZero Using Monte-Carlo Graph Search Johannes Czech1, Patrick Korus1, Kristian Kersting1, 2, 3 1 Department of Computer Science, TU Darmstadt, Germany 2 Centre for Cognitive Science, TU Darmstadt, Germany 3 Hessian Center for Artificial Intelligence (hessian.AI), Darmstadt, Germany [email protected], [email protected], [email protected] Abstract A 8 AlphaZero The algorithm has been successfully applied in a 7 range of discrete domains, most notably board games. It uti- lizes a neural network that learns a value and policy func- 6 tion to guide the exploration in a Monte-Carlo Tree Search. Although many search improvements such as graph search 5 have been proposed for Monte-Carlo Tree Search in the past, 4 most of them refer to an older variant of the Upper Confi- dence bounds for Trees algorithm that does not use a pol- 3 icy for planning. We improve the search algorithm for Alp- 2 haZero by generalizing the search tree to a directed acyclic graph. This enables information flow across different sub- 1 trees and greatly reduces memory consumption. Along with a b c d e f g h Monte-Carlo Graph Search, we propose a number of further extensions, such as the inclusion of -greedy exploration, a B C revised terminal solver and the integration of domain knowl- edge as constraints. In our empirical evaluations, we use the e4 Nf3 e4 Nf3 CrazyAra engine on chess and crazyhouse as examples to Nc6 Nc6 show that these changes bring significant improvements to e5 Nc6 e5 Nc6 e5 e5 AlphaZero.
    [Show full text]
  • A General Reinforcement Learning Algorithm That Masters Chess, Shogi and Go Through Self-Play
    A general reinforcement learning algorithm that masters chess, shogi and Go through self-play David Silver,1;2∗ Thomas Hubert,1∗ Julian Schrittwieser,1∗ Ioannis Antonoglou,1;2 Matthew Lai,1 Arthur Guez,1 Marc Lanctot,1 Laurent Sifre,1 Dharshan Kumaran,1;2 Thore Graepel,1;2 Timothy Lillicrap,1 Karen Simonyan,1 Demis Hassabis1 1DeepMind, 6 Pancras Square, London N1C 4AG. 2University College London, Gower Street, London WC1E 6BT. ∗These authors contributed equally to this work. Abstract The game of chess is the longest-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. By contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go by reinforcement learning from self- play. In this paper, we generalize this approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging games. Starting from random play and given no domain knowledge except the game rules, AlphaZero convincingly defeated a world champion program in the games of chess and shogi (Japanese chess) as well as Go. The study of computer chess is as old as computer science itself. Charles Babbage, Alan Turing, Claude Shannon, and John von Neumann devised hardware, algorithms and theory to analyse and play the game of chess. Chess subsequently became a grand challenge task for a generation of artificial intelligence researchers, culminating in high-performance computer chess programs that play at a super-human level (1,2).
    [Show full text]
  • Computer Shogi 2000 Through 2004
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by DSpace at Waseda University 195 Computer Shogi 2000 through 2004 Takenobu Takizawa Since the first computer shogi program was developed by the author in 1974, thirty years has passed. During that time,shogi programming has attracted both researchers and commercial programmers and playing strength has improved steadily. Currently, the best programs have a level that is comparable to that of a very strong amateur player (about 5-dan). In the near future, a good program will beat a professional player. The basic structure of strong shogi programs is similar to that of chess programs. However, the differences between chess and shogi have led to the development of some shogi-specific methods. In this paper the author will give an overview of the history of computer shogi, summarize the most successful techniques and give some ideas for future directions of research in computer shogi. 1 . Introduction Shogi, or Japanese chess, is a two-player complete information game similar to chess. As in chess, the goal of the game is to capture the opponent’s king. However, there are a number of differences be- tween chess and shogi: the shogi board is slightly bigger than the chess board (9x9 instead of 8x8), there are different pieces that are relatively weak compared to the pieces in chess (no queens, but gold generals, sil- ver generals and lances) and the number of pieces in shogi is larger (40 instead of 32). But the most important difference between chess and shogi is the possibility to re-use captured pieces.
    [Show full text]
  • Computer Games Workshop 2007
    Computer Games Workshop 2007 Amsterdam, June 15{17, 2007 MAIN SPONSORS Preface We are pleased to present the proceedings of the Computer Games Workshop 2007, Amsterdam, June 15{17, 2007. This workshop will be held in conjunc- tion with the 12th Computer Olympiad and the 15th World Computer-Chess Championship. Although the announcement was quite late, we were pleased to receive no less than 24 contributions. After a \light" refereeing process 22 papers were accepted. We believe that they present a nice overview of state-of-the-art research in the ¯eld of computer games. The 22 accepted papers can be categorized into ¯ve groups, according to the type of games used. Chess and Chess-like Games In this group we have included two papers on Chess, one on Kriegspiel, and three on Shogi (Japanese Chess). Matej Guid and Ivan Bratko investigate in Factors A®ecting Diminishing Returns for Searching Deeper the phenomenon of diminishing returns for addi- tional search e®ort. Using the chess programs Crafty and Rybka on a large set of grandmaster games, they show that diminishing returns depend on (a) the value of positions, (b) the quality of the evaluation function, and (c) the phase of the game and the amount of material on the board. Matej Guid, Aritz P¶erez,and Ivan Bratko in How Trustworthy is Crafty's Analysis of Chess Champions? again used Crafty in an attempt at an objective assessment of the strength of chess grandmasters of di®erent times. They show that their analysis is trustworthy, and hardly depends on the strength of the chess program used, the search depth applied, or the size of the sets of positions used.
    [Show full text]
  • Report on the 13Th CSA Computer-Shogi Championship 1
    Rep ort on the th CSA ComputerShogi Championship REPORT ON THE TH CSA WORLD COMPUTERSHOGI CHAMPIONSHIP Kazusa Academic Park Kisarazu Japan May Reijer Grimbergen Department of Information Science Saga University Japan INTRODUCTION Again the Kazusa Academic Park in Kisarazu was the venue for the computer shogi championships Already the th edition and denitely one of the biggest games events in the world The number of participants had dropp ed slightly to compared to last year This may b e in part b ecause of the date of this tournament which is in the middle of one of the few ma jor holidays in Japan This year my program SPEAR and the North Korean program KCC were the only foreign entries Unfortunately Je Rollason had to pass on the tournament this year as he was in the pro cess of starting up a new company and couldnt nd the time to come to Japan or even work at his program Shotest The programmers of KCC also had decided not to show up themselves this tournament and their program was op erated bytwo Japanese op erators from the company that publishes the commercial version of the KCC program This was probably a wise decision as the relations b etween Japan and North Korea are quite strained at the momentover the ab duction of Japanese citizens by North Korea in the past and the Korean nuclear program There were even fears that there might b e demonstrations at the tournamentvenue but computer shogi turned out not to b e big enough to draw the attention necessary for having an eective demonstration Or mayb e the demonstrators couldnt
    [Show full text]
  • Shogi Yearbook 2014 [Wählen Sie Das Datum Aus]
    Shogi Yearbook 2014 [Wählen Sie das Datum aus] SHOGI24.COM SHOGI YEARBOOK 2014 Title match games, Challenger’s tournaments, interview with Chess GM Peter Heine Nielsen, tournament reports, photos, ’Following a dream’ by Karolina Styczyńska, General Shogi, statistics, … This yearbook is a free PDF document Shogi Yearbook 2014 Content Content Content .................................................................................................................................................... 2 Just a few words ... .................................................................................................................................. 5 63. Osho .................................................................................................................................................. 6 63. Osho league ................................................................................................................................... 6 63th Osho title match ........................................................................................................................... 7 Game 1 ............................................................................................................................................. 7 Game 2 ........................................................................................................................................... 10 Game 3 ..........................................................................................................................................
    [Show full text]
  • A Survey on Chess Engine Using Deep Learning Shreya B
    International Journal of Future Generation Communication and Networking Vol. 13, No. 3s, (2020), pp. 1398–1406 A Survey on Chess Engine Using Deep Learning Shreya B. Ahire1, Santosh S. Kale2, Sumit U. Mali3 Department of Computer Engineering, NBNSSOE, Ambegaon, Pune, Maharashtra, India Abstract Chess games have been avantgarde in the history of artificial intelligence. Traditional chess engines such as Stock fish and Komodo have been developed using heuristics with thousands of rules handcrafted by human chess professionals. However, in recent developments, Alpha Zero adopts a completely different strategy and replaces these handmade rules with deep neural networks. In addition, the generic Algorithm used to develop Alpha Zero knows nothing beyond the basic rules of the game, but it achieved superhuman performance in chess, go and shogi games by learning from self-playing games. In addition, you can see how the chess engine evolved over time using different technology and algorithms. In addition, alpha zero is a method of generalizing the algorithm of chess, shogi (shogi), and Go. Starting with random play, I studied how not to be given domain knowledge other than game rules, and Alpha Zero became the best in chess games in a few hours of self-learning. Finally, by observing how the use of deep learning affects the performance of the chess engine and the movement of chess games played on these two engines, Alpha Zero is summarized as compared to other traditional chess engines. Keywords: Chess Engine, Deep Learning, Artificial Intelligence, Alpha Zero. I INTRODUCTION Chess is ranked among the most popular board games and everyone plays everywhere, including city parks, living rooms, schools and well-published formal competitions.
    [Show full text]
  • Distributed Game Tree Search and Improvements – Match Between Hiroyuki Miura and Gpsshogi –
    Parameter-Free Tree Style Pipeline in Asynchronous Parallel Game-Tree Search Shu Yokoyama1, Tomoyuki Kaneko1?, and Tetsuro Tanaka2 1 Graduate School of Arts and Sciences, The University of Tokyo 2 Information Technology Center, The University of Tokyo Abstract. Asynchronous parallel game-tree search methods are effec- tive in improving playing strength by using many computers connected through relatively slow networks. In game position parallelization, the master program manages a game-tree and distributes positions in the tree to workers. Then, each worker asynchronously searches the best move and evaluation for its assigned position. We present a new method for constructing an appropriate master tree that provides more important moves with more workers on their sub-trees to improve playing strength. Our contribution introduces two advantages: (1) being parameter free in that users do not need to tune parameters through trial and error, and (2) efficiency suitable even for short-time matches, such as one second per move. We implemented our method in chess with a top-level chess program (Stockfish) and evaluated playing strength through self-plays. We confirmed that playing strength improves with up to sixty workers. Keywords: game-tree search, distributed computing, chess, parallel search 1 Introduction Parallelization of game-tree search has been extensively studied to improve game programs' playing strength, especially in chess and its variants. Several efficient methods have been developed in hardware parallelization [2, 4], in thread-level parallelization [14], and for tightly connected computers [5, 1, 12, 18]. Recently, as grid computing has become popular, new approaches utilizing computational resources placed in different locations connected through wide- area networks have been proposed.
    [Show full text]