Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!

These ads will disappear if you register on the forum


  • Please log in to reply
168 replies to this topic

#161
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

The game of chess is the most widely-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. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.


  • Zaphod likes this
Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#162
Jakob

Jakob

    Fenny-Eyed Slubber-Yuck

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 5,244 posts
  • LocationIn the Basket of Deplorables

 

Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

The game of chess is the most widely-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. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.

 

I'll be impressed when DeepMind can play quantum four-dimensional Taikyoku shogi like Trump.


Click 'show' to see quotes from great luminaries.

Spoiler

#163
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

How do you know it can't? They're not gonna say anything that chases away investors, after all!


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#164
Jakob

Jakob

    Fenny-Eyed Slubber-Yuck

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 5,244 posts
  • LocationIn the Basket of Deplorables

How do you know it can't? They're not gonna say anything that chases away investors, after all!

Nobody plays quantum 4D taikyoku shogi like Trump. His game is so great. You won't believe how great he is at quantum 4D taikyoku shogi. Believe me, he's the best player, okay? Even the failing New York Times--their ratings are so terrible, by the way--even they admit it. We have all the best games. Such beautiful games.


Click 'show' to see quotes from great luminaries.

Spoiler

#165
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

And then AlphaZero brings out its personally owned Atlas and asks "So would you like to play a game, Mr. Trump?"


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#166
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

AlphaZero AI beats champion chess program after teaching itself in four hours

AlphaZero, the game-playing AI created by Google sibling DeepMind, has beaten the world’s best chess-playing computer program, having taught itself how to play in under four hours.
The repurposed AI, which has repeatedly beaten the world’s best Go players as AlphaGo, has been generalised so that it can now learn other games. It took just four hours to learn the rules to chess before beating the world champion chess program, Stockfish 8, in a 100-game match up.

This one's a bit iffier, though. They handicapped Swordfish, and AlphaZero technically didn't train for just 4 hours.


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#167
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

Alpha Zero’s “Alien” Chess Shows the Power, and the Peculiarity, of AI

The latest AI program developed by DeepMind is not only brilliant and remarkably flexible—it’s also quite weird.
DeepMind published a paper this week describing a game-playing program it developed that proved capable of mastering chess and the Japanese game Shoju, having already mastered the game of Go.
Demis Hassabis, the founder and CEO of DeepMind and an expert chess player himself, presented further details of the system, called Alpha Zero, at an AI conference in California on Thursday. The program often made moves that would seem unthinkable to a human chess player.
“It doesn’t play like a human, and it doesn’t play like a program,” Hassabis said at the Neural Information Processing Systems (NIPS) conference in Long Beach. “It plays in a third, almost alien, way.”


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#168
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda

DeepMind Proves Its AI System's General Purposefulness

The game of chess is the most widely-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.

DeepMind's AlphaGo Zero program, by contrast, recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play.
Now, in a new paper, researchers from the Google-owned company have shown that the algorithm could be generalized that can achieve, blank slate, superhuman performance across many challenging domains. The research authors, which include Demis Hassabis have been rapidly advancing their system since their historic win at the game of Go only last year.

Note, this isn't general AI. But this is purely what I've been labeling about for a while as "less-narrow" AI or "cluster" AI or "wide" AI.


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#169
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,155 posts
  • LocationAnur Margidda


Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!





Also tagged with one or more of these keywords: DeepMind, deep learning, deep reinforcement learning, progressive neural network, artificial intelligence, AGI, differentiable neural, Google, RankBrain, artificial neural networks

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users