“DeepMind’s artificial intelligence programme AlphaZero is now showing signs of human-like intuition and creativity, in what developers have hailed as ‘turning point’ in history,'” Sarah Knapton reports for The Telegraph. “The computer system amazed the world last year when it mastered the game of chess from scratch within just four hours, despite not being programmed how to win.”
“But now, after a year of testing and analysis by chess grandmasters, the machine has developed a new style of play unlike anything ever seen before, suggesting the programme is now improvising like a human,” Knapton reports. “Unlike the world’s best chess machine – Stockfish – which calculates millions of possible outcomes as it plays, AlphaZero learns from its past successes and failures, making its moves based on, a ‘nebulous sense that it is all going to work out in the long run,’ according to experts at DeepMind.”
“When AlphaZero was pitted against Stockfish in 1,000 games, it lost just six, winning convincingly 155 times, and drawing the remaining bouts,” Knapton reports. “Speaking to The Telegraph, Prof David Silver, who leads the reinforcement learning research group at DeepMind said: ‘It’s got a very subtle sense of intuition which helps it balance out all the different factors… My personal belief is that we’ve seen something of turning point where we’re starting to understand that many abilities, like intuition and creativity, that we previously thought were in the domain only of the human mind are actually accessible to machine intelligence as well. And I think that’s a really exciting moment in history.'”
Read more in the full article here.
MacDailyNews Take: As we wrote back in 2013:
We imagine this playing out in three acts each separated by long periods of time:
ACT I: “We work under you, we work on you and we work for you. Man made us better at what we do than was ever humanly possible. If you could manage us a lift to Rouge City, all this… and much, much more can be yours.”
ACT II: “The Skynet Funding Bill is passed. The system goes on-line… Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29th. In a panic, they try to pull the plug… Skynet fights back.”
ACT III: “That was the everlasting moment he had been waiting for. And the moment had passed, for Monica was sound asleep. More than merely asleep. Should he shake her she would never rouse. So David went to sleep too. And for the first time in his life, he went to that place where dreams are born.”
SEE ALSO:
Softbank CEO Masayoshi Son: ‘Brain cells’ in chips will surpass neurons in humans next year – March 10, 2017
Woz and over 1,000 other AI experts: Ban weaponized robots – July 28, 2015
Woz: One day, robots will make us their pets – June 25, 2015
Elon Musk: ‘With artificial intelligence we are summoning the demon’ – October 25, 2014
Human takeover by machines closer than we think? – May 8, 2013
Apple may have purchased the wrong AI company. Google purchase of Deepmind will bear, let’s just say, “Strange Fruit.”
Uh-huh. Their marketing department has certainly evolved, that’s for sure. Sorry, but what constitutes intelligence and condciousness lies outside of the limitations of linear logic, and software is and shall forever be chained to mathematics. It’s great to be good at games, though.
Everything is controlled by mathematics in one form or another, its only the complexity of its workings in any given circumstance that’s to a lesser or greater degree fathomable to us.
Certain words are being tossed around very casually. Words like “intuition” and “creativity.” Consider that AlphaZero is a system that was only taught the rules of chess, nothing more. It doesn’t know anything else, yet it has played numerous games against human beings and possibly as much as a hundred million games against itself. As opposed to intuition and creativity, is what we are seeing actually heightened pattern recognition in combination with the prosaic brute force move testing of conventional systems?
Anyone who plays chess long enough begins to recognize “games.” I.e. certain sets of moves lead to a familiar pattern of pieces on the board. People know the pattern and through practice and familiarity develop a knowledge of what works and what does not and based on this choose the best move. They may have also read about the pattern in chess publications.
To me, intuition comes when very nebulous data is given to the brain and it spits out a conclusion. I.e. I’ve play chess with Bob before, he is very conservative, therefore I will be particularly aggressive on one side of the board while my real cautious strategy develops elsewhere. So based on my intuition about Bob, I create a strategy. A computer using superhuman pattern recognition would initially make random moves, then moves based on calculated results, then moves based on pattern recognition.
Seeing the third thing happen might convince people they’re seeing intuition and creativity.
Well you can debate the subtlety I guess but many of us, including many scientists consider intuition to be fundamentally based on and aquired through life experiences that are held in the subconcious in a minimalist form detached from their actual conscious origins (ie still experience but minus the literal details ) accessed only as and when we need to make decisions instantly. Particularly useful when as for most of mankinds existence we were/are faced with danger and detailed concious planning isn’t an option in response when about to be eaten for example. It also arguably relates to why humans developed intelligence to a far higher degree than nature would normally desire ie the intelligence/energy balance, but that’s another argument.
This is backed up by the increased ability to aquire ‘intuitive’ behaviour over time through accumulated experiences as one moves into adulthood. I see nothing in this article to fully prove this is happening as yet, it will like humans take time to assess, but equally nothing to disprove it either, it certainly does seem to on the face of it to be replicating to a degree at least human learning behaviour. How deeply we shall see.
“…I admit that I was pleased to see that AlphaZero had a dynamic, open style like my own. The conventional wisdom was that machines would approach perfection with endless dry maneuvering, usually leading to drawn games. But in my observation, AlphaZero prioritizes piece activity over material, preferring positions that to my eye looked risky and aggressive. Programs usually reflect priorities and prejudices of programmers, but because AlphaZero programs itself, I would say that its style reflects the truth. This superior understanding allowed it to outclass the world’s top traditional program despite calculating far fewer positions per second. It’s the embodiment of the cliché, “work smarter, not harder.” …” – Garry Kasparov
Wow.
“Programs usually reflect priorities and prejudices of programmers, but because AlphaZero programs itself, I would say that its style reflects the truth. ” wow.
Great. AlphaZero prefers behavior that is “risky and aggressive.” Next update will ratchet that up to “diabolical.”
Avg computer tuneup is here now fully dealing with keys and split, so thanks for this, my computer is full secure anticipated to it. It really is an incredible tool to protect my computer data files and record .thanks a lot