fbpx
Wikipedia

AlphaStar (software)

January 01, 1970

For the broadcasting service, see AlphaStar (satellite broadcasting service).

AlphaStar is a computer program by DeepMind that plays the video game StarCraft II. It was unveiled to the public by name in January 2019. In a significant milestone for artificial intelligence, AlphaStar attained Grandmaster status in August 2019.

Background edit

Games created for humans are considered to have external validity as benchmarks of progress in artificial intelligence. IBM's chess-playing Deep Blue (1997) and DeepMind's AlphaGo (2016) were considered major milestones; some argue that StarCraft would also be a major milestone, due to StarCraft's "real-time play, partial observability, no single dominant strategy, complex rules that make it hard to build a fast forward model, and a particularly large and varied action space."[1] Though difficult, StarCraft may still be tractable with current technology because "its rules are known and the world is discrete with only a few types of objects".[2]

StarCraft II is a popular fast-paced online real-time strategy game by Blizzard Entertainment.[3][4]

History edit

DeepMind Technologies was founded in the UK in 2010. As early as 2011, founder Demis Hassabis called StarCraft "the next step up" after games like Go.[5] DeepMind became a Google subsidiary in 2014, after demonstrating self-learning bots with superhuman ability at a variety of Atari 2600 games. In February 2015, computer scientist Zachary Mason predicted Deepmind's research "leads to StarCraft in five or ten years".[6] In March 2016, following AlphaGo's victory over Lee Sedol, a world champion Go player, Hassabis publicly mulled building an AI for StarCraft, citing it as a strategic game with incomplete information where (unlike Go) much of the "board" is invisible.[7][8] A formal collaboration was announced at BlizzCon in November 2016, alongside a plan to release an open development environment for bots in Q1 of 2017.[9]

By 2017, DeepMind was experimenting with feeding StarCraft data into its software. In August 2017, DeepMind and Blizzard released development tools to assist in bot development, as well as data from 65,000 past games. At the time, computer scientist and StarCraft tournament manager David Churchill guessed it would take five years for a bot to beat a human, but made the caveat that AlphaGo had beaten expectations. In Wired, tech journalist Tom Simonite stated, "No one expects the robot to win anytime soon. But when it does, it will be a far greater achievement than DeepMind's conquest of Go."[10]

On 19 December 2018, DeepMind's bot defeated "a top professional player", Grzegorz "MaNa" Komincz, 5-0. DeepMind announced the bot, named "AlphaStar", on 24 January 2019. A journalist at Ars Technica and others argued that AlphaStar still had unfair advantages: "AlphaStar has the ability to make its clicks with surgical precision using an API, whereas human players are constrained by the mechanical limits of computer mice". AlphaStar also had a global view rather than being limited by the in-game camera. Furthermore, while there was a cap on the number of actions over a five-second window, AlphaStar was free to allocate its action quota unevenly across the window in order to launch superhuman bursts of activity at critical moments. DeepMind quickly retrained AlphaStar under more realistic constraints, and then lost a rematch with Komincz.[11] Starting in July 2019, the new, constrained version of AlphaStar anonymously competed against players who "opted in" on the public 1v1 European multiplayer ladder.[12] By the end of August 2019, AlphaStar had attained "grandmaster level", ranking among the top 0.2 percent of human players.[13]

Algorithms edit

Unlike AlphaZero, AlphaStar initially learns to imitate the moves of the best players in its database of human vs. human games; this step is necessary to solve what DeepMind's Dave Silver calls "the exploration problem": discovering new strategies would otherwise be like finding a "needle in a haystack". Agents then play each other and deploy reinforcement learning. These main agents also learn by playing against suboptimal "exploiter agents" whose purpose is to expose weaknesses in the main agents.[14][15]

Reactions edit

After his 5-0 defeat in December 2018, Komincz stated "I wasn't expecting the AI to be that good".[16]

Stuart Russell assessed that AlphaStar's 2018 victory required "a fair amount of problem-specific effort" and that general-purpose methods were "not quite ready for StarCraft".[2]

An article in Wired UK judged AlphaStar's new constraints, adopted for the July 2019 matches, to be "fair" this time around.[17] StarCraft pro Raza "RazerBlader" Sekha stated AlphaStar was "impressive" but had its quirks, succumbing in one game to an unorthodox army composition made up of only air units. The UK's top player, Joshua "RiSky" Hayward, expressed some disappointment, saying AlphaStar "often didn't make the most efficient, strategic decisions".[14] Pro Diego "Kelazhur" Schwimer called AlphaStar's play "unimaginably unusual; it really makes you question how much of StarCraft's diverse possibilities pro players have really explored".[13] AlphaStar's opponents often did not realize they were playing a bot.[18]

Ian Sample of The Guardian called AlphaStar a "landmark achievement" for the field of artificial intelligence.[19] Churchill stated that he had previously seen bots that master one or two elements of StarCraft, but that AlphaStar was the first that can handle the game in its entirety.[3] Gary Marcus expressed his continuing skepticism about deep learning, stating: "So far the field has struggled to take techniques like this out of the laboratory and game environments and into the real world, and I don't immediately see this result as progress in that direction".[14] AI researcher Jon Dodge was surprised by AlphaStar, stating he did not expect such a "superhuman" performance for "another couple of years"; in contrast, Churchill states "StarCraft is nowhere near being 'solved', and AlphaStar is not yet even close to playing at a world champion level".[4]

Legacy edit

DeepMind argues that insights from AlphaStar might benefit robots, self-driving cars and virtual assistants, which need to operate with "imperfectly observed information". Silver has indicated his lab "may rest at this point", rather than try to substantially improve AlphaStar.[14] Silver himself argues that "AlphaStar has become the first AI system to reach the top tier of human performance in any professionally played e-sport on the full unrestricted game under professionally approved conditions... Ever since computers cracked Go, chess and poker, the game of StarCraft has emerged, essentially by consensus from the community, as the next grand challenge for AI."[19]

Computer scientist Noel Sharkey argues, disapprovingly, that "military analysts will certainly be eyeing the successful AlphaStar real-time strategies as a clear example of the advantages of AI for battlefield planning".[19] In contrast, Silver argues: "To say that this has any kind of military use is saying no more than to say an AI for chess could be used to lead to military applications".[14]

See also edit

References edit

  1. ^ Arulkumaran, Kai; Antoine, Cully; Togelius, Julian (July 2019). "AlphaStar: An evolutionary computation perspective". Proceedings of the Genetic and Evolutionary Computation Conference Companion. pp. 314–5. arXiv:1902.01724. doi:10.1145/3319619.3321894. ISBN 9781450367486. S2CID 59604439.
  2. ^ a b Russell, Stuart (8 October 2019). "Chapter 2: Intelligence in Humans and Machines". Human Compatible: Artificial Intelligence and the Problem of Control (Hardcover ed.). Viking. ISBN 978-0-525-55861-3.
  3. ^ a b Daley, Jason (30 October 2019). "A.I. Mastered Backgammon, Chess and Go. Now It Takes On StarCraft II". Smithsonian Magazine. Retrieved 26 February 2020.
  4. ^ a b Garisto, Dan (30 October 2019). "Google AI beats top human players at strategy game StarCraft II". Nature. doi:10.1038/d41586-019-03298-6. PMID 33116320. S2CID 208116679.
  5. ^ Cheng, Jonathan (22 April 2016). "Computers That Crush Humans at Games Might Have Met Their Match: 'StarCraft'". Wall Street Journal. Retrieved 26 February 2020.
  6. ^ Gibbs, Samuel (27 January 2014). "Google buys UK artificial intelligence startup Deepmind for £400m". The Guardian. Retrieved 26 February 2020.
  7. ^ Byford, Sam (10 March 2016). "DeepMind founder Demis Hassabis on how AI will shape the future". The Verge. Retrieved 26 February 2020.
  8. ^ "Google's Artificial Intelligence Project May Start Playing Video Games". Fortune. Retrieved 26 February 2020.
  9. ^ Byford, Sam (4 November 2016). "Google DeepMind's next gaming challenge: can AI beat StarCraft II?". The Verge. Retrieved 26 February 2020.
  10. ^ "Google DeepMind AI Declares Galactic War on StarCraft". Wired. 9 August 2017. Retrieved 26 February 2020.
  11. ^ Lee, Timothy B. (30 January 2019). "An AI crushed two human pros at StarCraft—but it wasn't a fair fight". Ars Technica. Retrieved 26 February 2020.
  12. ^ Amadeo, Ron (11 July 2019). "DeepMind AI is secretly lurking on the public StarCraft II 1v1 ladder". Ars Technica. Retrieved 26 February 2020.
  13. ^ a b Statt, Nick (30 October 2019). "DeepMind's StarCraft 2 AI is now better than 99.8 percent of all human players". The Verge. Retrieved 26 February 2020.
  14. ^ a b c d e Kelion, Leo (30 October 2019). "DeepMind claims landmark moment for AI in esports". BBC News. Retrieved 26 February 2020.
  15. ^ Vinyals, Oriol; Babuschkin, Igor; Czarnecki, Wojciech M.; et al. (30 October 2019). "Grandmaster level in StarCraft II using multi-agent reinforcement learning". Nature. 575 (7782): 350–354. Bibcode:2019Natur.575..350V. doi:10.1038/s41586-019-1724-z. PMID 31666705. S2CID 204972004.
  16. ^ Heaven, Douglas (21 August 2019). "Mind meld: Artificial intelligence is improving the way humans think". New Scientist. Retrieved 26 February 2020.
  17. ^ Lee, Alex (30 October 2019). "DeepMind has finally thrashed humans at StarCraft for real". Wired UK. Retrieved 26 February 2020.
  18. ^ "DeepMind claims breakthrough as AI masters StarCraft II game". Financial Times. 2019. Retrieved 26 February 2020.
  19. ^ a b c Sample, Ian (30 October 2019). "AI becomes grandmaster in 'fiendishly complex' StarCraft II". The Guardian. Retrieved 26 February 2020.

External links edit

  • DeepMind StarCraft II Demonstration on YouTube

January 01, 1970
alphastar, software, broadcasting, service, alphastar, satellite, broadcasting, service, alphastar, computer, program, deepmind, that, plays, video, game, starcraft, unveiled, public, name, january, 2019, significant, milestone, artificial, intelligence, alpha. For the broadcasting service see AlphaStar satellite broadcasting service AlphaStar is a computer program by DeepMind that plays the video game StarCraft II It was unveiled to the public by name in January 2019 In a significant milestone for artificial intelligence AlphaStar attained Grandmaster status in August 2019 Contents 1 Background 2 History 3 Algorithms 4 Reactions 5 Legacy 6 See also 7 References 8 External linksBackground editGames created for humans are considered to have external validity as benchmarks of progress in artificial intelligence IBM s chess playing Deep Blue 1997 and DeepMind s AlphaGo 2016 were considered major milestones some argue that StarCraft would also be a major milestone due to StarCraft s real time play partial observability no single dominant strategy complex rules that make it hard to build a fast forward model and a particularly large and varied action space 1 Though difficult StarCraft may still be tractable with current technology because its rules are known and the world is discrete with only a few types of objects 2 StarCraft II is a popular fast paced online real time strategy game by Blizzard Entertainment 3 4 History editDeepMind Technologies was founded in the UK in 2010 As early as 2011 founder Demis Hassabis called StarCraft the next step up after games like Go 5 DeepMind became a Google subsidiary in 2014 after demonstrating self learning bots with superhuman ability at a variety of Atari 2600 games In February 2015 computer scientist Zachary Mason predicted Deepmind s research leads to StarCraft in five or ten years 6 In March 2016 following AlphaGo s victory over Lee Sedol a world champion Go player Hassabis publicly mulled building an AI for StarCraft citing it as a strategic game with incomplete information where unlike Go much of the board is invisible 7 8 A formal collaboration was announced at BlizzCon in November 2016 alongside a plan to release an open development environment for bots in Q1 of 2017 9 By 2017 DeepMind was experimenting with feeding StarCraft data into its software In August 2017 DeepMind and Blizzard released development tools to assist in bot development as well as data from 65 000 past games At the time computer scientist and StarCraft tournament manager David Churchill guessed it would take five years for a bot to beat a human but made the caveat that AlphaGo had beaten expectations In Wired tech journalist Tom Simonite stated No one expects the robot to win anytime soon But when it does it will be a far greater achievement than DeepMind s conquest of Go 10 On 19 December 2018 DeepMind s bot defeated a top professional player Grzegorz MaNa Komincz 5 0 DeepMind announced the bot named AlphaStar on 24 January 2019 A journalist at Ars Technica and others argued that AlphaStar still had unfair advantages AlphaStar has the ability to make its clicks with surgical precision using an API whereas human players are constrained by the mechanical limits of computer mice AlphaStar also had a global view rather than being limited by the in game camera Furthermore while there was a cap on the number of actions over a five second window AlphaStar was free to allocate its action quota unevenly across the window in order to launch superhuman bursts of activity at critical moments DeepMind quickly retrained AlphaStar under more realistic constraints and then lost a rematch with Komincz 11 Starting in July 2019 the new constrained version of AlphaStar anonymously competed against players who opted in on the public 1v1 European multiplayer ladder 12 By the end of August 2019 AlphaStar had attained grandmaster level ranking among the top 0 2 percent of human players 13 Algorithms editUnlike AlphaZero AlphaStar initially learns to imitate the moves of the best players in its database of human vs human games this step is necessary to solve what DeepMind s Dave Silver calls the exploration problem discovering new strategies would otherwise be like finding a needle in a haystack Agents then play each other and deploy reinforcement learning These main agents also learn by playing against suboptimal exploiter agents whose purpose is to expose weaknesses in the main agents 14 15 Reactions editAfter his 5 0 defeat in December 2018 Komincz stated I wasn t expecting the AI to be that good 16 Stuart Russell assessed that AlphaStar s 2018 victory required a fair amount of problem specific effort and that general purpose methods were not quite ready for StarCraft 2 An article in Wired UK judged AlphaStar s new constraints adopted for the July 2019 matches to be fair this time around 17 StarCraft pro Raza RazerBlader Sekha stated AlphaStar was impressive but had its quirks succumbing in one game to an unorthodox army composition made up of only air units The UK s top player Joshua RiSky Hayward expressed some disappointment saying AlphaStar often didn t make the most efficient strategic decisions 14 Pro Diego Kelazhur Schwimer called AlphaStar s play unimaginably unusual it really makes you question how much of StarCraft s diverse possibilities pro players have really explored 13 AlphaStar s opponents often did not realize they were playing a bot 18 Ian Sample of The Guardian called AlphaStar a landmark achievement for the field of artificial intelligence 19 Churchill stated that he had previously seen bots that master one or two elements of StarCraft but that AlphaStar was the first that can handle the game in its entirety 3 Gary Marcus expressed his continuing skepticism about deep learning stating So far the field has struggled to take techniques like this out of the laboratory and game environments and into the real world and I don t immediately see this result as progress in that direction 14 AI researcher Jon Dodge was surprised by AlphaStar stating he did not expect such a superhuman performance for another couple of years in contrast Churchill states StarCraft is nowhere near being solved and AlphaStar is not yet even close to playing at a world champion level 4 Legacy editDeepMind argues that insights from AlphaStar might benefit robots self driving cars and virtual assistants which need to operate with imperfectly observed information Silver has indicated his lab may rest at this point rather than try to substantially improve AlphaStar 14 Silver himself argues that AlphaStar has become the first AI system to reach the top tier of human performance in any professionally played e sport on the full unrestricted game under professionally approved conditions Ever since computers cracked Go chess and poker the game of StarCraft has emerged essentially by consensus from the community as the next grand challenge for AI 19 Computer scientist Noel Sharkey argues disapprovingly that military analysts will certainly be eyeing the successful AlphaStar real time strategies as a clear example of the advantages of AI for battlefield planning 19 In contrast Silver argues To say that this has any kind of military use is saying no more than to say an AI for chess could be used to lead to military applications 14 See also editAlphaZero OpenAI Five a similar bot for Dota 2 Pluribus poker bot References edit Arulkumaran Kai Antoine Cully Togelius Julian July 2019 AlphaStar An evolutionary computation perspective Proceedings of the Genetic and Evolutionary Computation Conference Companion pp 314 5 arXiv 1902 01724 doi 10 1145 3319619 3321894 ISBN 9781450367486 S2CID 59604439 a b Russell Stuart 8 October 2019 Chapter 2 Intelligence in Humans and Machines Human Compatible Artificial Intelligence and the Problem of Control Hardcover ed Viking ISBN 978 0 525 55861 3 a b Daley Jason 30 October 2019 A I Mastered Backgammon Chess and Go Now It Takes On StarCraft II Smithsonian Magazine Retrieved 26 February 2020 a b Garisto Dan 30 October 2019 Google AI beats top human players at strategy game StarCraft II Nature doi 10 1038 d41586 019 03298 6 PMID 33116320 S2CID 208116679 Cheng Jonathan 22 April 2016 Computers That Crush Humans at Games Might Have Met Their Match StarCraft Wall Street Journal Retrieved 26 February 2020 Gibbs Samuel 27 January 2014 Google buys UK artificial intelligence startup Deepmind for 400m The Guardian Retrieved 26 February 2020 Byford Sam 10 March 2016 DeepMind founder Demis Hassabis on how AI will shape the future The Verge Retrieved 26 February 2020 Google s Artificial Intelligence Project May Start Playing Video Games Fortune Retrieved 26 February 2020 Byford Sam 4 November 2016 Google DeepMind s next gaming challenge can AI beat StarCraft II The Verge Retrieved 26 February 2020 Google DeepMind AI Declares Galactic War on StarCraft Wired 9 August 2017 Retrieved 26 February 2020 Lee Timothy B 30 January 2019 An AI crushed two human pros at StarCraft but it wasn t a fair fight Ars Technica Retrieved 26 February 2020 Amadeo Ron 11 July 2019 DeepMind AI is secretly lurking on the public StarCraft II 1v1 ladder Ars Technica Retrieved 26 February 2020 a b Statt Nick 30 October 2019 DeepMind s StarCraft 2 AI is now better than 99 8 percent of all human players The Verge Retrieved 26 February 2020 a b c d e Kelion Leo 30 October 2019 DeepMind claims landmark moment for AI in esports BBC News Retrieved 26 February 2020 Vinyals Oriol Babuschkin Igor Czarnecki Wojciech M et al 30 October 2019 Grandmaster level in StarCraft II using multi agent reinforcement learning Nature 575 7782 350 354 Bibcode 2019Natur 575 350V doi 10 1038 s41586 019 1724 z PMID 31666705 S2CID 204972004 Heaven Douglas 21 August 2019 Mind meld Artificial intelligence is improving the way humans think New Scientist Retrieved 26 February 2020 Lee Alex 30 October 2019 DeepMind has finally thrashed humans at StarCraft for real Wired UK Retrieved 26 February 2020 DeepMind claims breakthrough as AI masters StarCraft II game Financial Times 2019 Retrieved 26 February 2020 a b c Sample Ian 30 October 2019 AI becomes grandmaster in fiendishly complex StarCraft II The Guardian Retrieved 26 February 2020 External links editDeepMind StarCraft II Demonstration on YouTube Retrieved from https en wikipedia org w index php title AlphaStar software amp oldid 1212030914, wikipedia, wiki, book, books, library,

article

, read, download, free, free download, mp3, video, mp4, 3gp, jpg, jpeg, gif, png, picture, music, song, movie, book, game, games.