Share this post

Artificial Intelligence is Go

DeepMind, Google’s artificial intelligence project, has taken a huge step forward.  It has successfully enabled a machine to teach itself how to play the ancient Chinese game of Go.  Without any human assistance.

AlphaGo Zero is the latest program developed by DeepMind.

Programmed with the rules but no other data AlphaGo Zero learnt the nuances and tactics of the game by repeatedly playing itself.  Within 72 hours it had progressed enough to be at a level to beat the best human players in the world.

alphago logo

What has it done exactly?

AlphaGo Zero is the new version of AlphaGo.  This AI software beat the champion Go players including the World’s number one, China’s Ke Jie earlier this year.

The software analysed thousands of past games.  This allowed it to crunch the numbers and learn from human strategy.  It was able to beat Ke Jie in every game of a three match series.

But Now AlphaGo Zero has beaten the original software.  In fact it hasn’t just beaten it.  It annihilated it.  AlphaGo Zero won a hundred game series 100-0.

game of go

Why is this so special?

What makes AlphaGo Zero such a leap forward is that it taught itself the game from scratch.  In three days it was able to surpass the knowledge of Go humans had amassed in over two thousand years of playing the game.

David Silver from DeepMind told the BBC: “We’ve actually removed the constraints of human knowledge and it’s able, therefore, to create knowledge itself from first principles, from a blank slate.”

That’s impressive.  And pretty scary.

Who are DeepMind?

Founded in 2010 by Demis Hassabis, Shane Legg and Mustafa Suleyman DeepMind is a London company.  Google acquired DeepMind in 2014.  It is now part of their Alphabet group.

deepmind logo

DeepMind are one of the world’s leaders in artificial intelligence.  On their website they say: “We’re on a scientific mission to push the boundaries of AI, developing programs that can learn to solve any complex problem without needing to be taught how.

“If we’re successful, we believe this will be one of the most important and widely beneficial scientific advances ever made, increasing our capacity to understand the mysteries of the universe and to tackle some of our most pressing real-world challenges.”

What does all this mean?

DeepMind says scientific and medical research will benefit from their development of AI.  Chief Executive, Demis Hassabis, told journalists: “We’re quite excited because we think this is now good enough to make some real progress on some real problems even though we’re obviously a long way from full AI.

“I hope these kind of algorithms will be routinely working with us as scientific experts, medical experts on advancing the frontiers of science and medicine – that’s what I hope.”

So humans aren’t about to be enslaved by robots?

Conspiracy theorists may see DeepMind as part of Google’s dastardly plot to take over the world.  And AlphaGo Zero the thin end of a wedge which ends with maniacal robots.

But, no you can relax.  Real world applications of AI is still years away. As are human hating killer robots.

But a global giant like Google controlling a machine which can think for itself will invoke a sense of unease among many.

Share this post

Craig Ellyard

Token old guy in the office and lifelong Hull City fan with all the psychological issues that brings. To relax I enjoy walking my two Labradors, as well as running and cycling.

5 comments

Add yours
  1. Johnson 21 October, 2017 at 10:15

    I can imagine it would be easier to play board games with a robot since smartphones already have the logic in these games to play against any opponent. This should be implemented soon.

  2. Chris 23 October, 2017 at 12:29

    Isaac Asimov saw this coming.

    3 laws safe! The 3 laws of robotics need to be hard wired NOW but we need to understand the implications.

    Read (Not watch the film) I robot, and then google Asimovs “zeroth law” to see what I mean

  3. Ian 4 November, 2017 at 18:39

    My fear of the AI is how it`s being used by the likes of google to push agendas. An Ai is all great, but not when companies can make it do things to spy on or manipulate people. Dangerous times.

  4. Jim 13 November, 2017 at 16:25

    Every new Idea / Invention can be used for good and bad. Seeing as these AI units are capable of out witting us already who is going to be intelligent enough to be able to control it for good and will they be morally ‘upright’ enough not to leave a key in the coding for private access if the hunger for power / or financial gain is in their DNA. I’m happy to be here at the outset of this technology but would not like to live through the ‘problem’ period when things start to go wrong.

Post a new comment