INFORMATION

This website uses cookies to store information on your computer. Some of these cookies are essential to make our site work and others help us to improve by giving us some insight into how the site is being used.

For further information, see our Privacy Policy.

Continuing to use this website is acceptance of these cookies.

We are not accepting any new registrations.

The future of AI...

Any topic related to science can be discussed here.
Post Reply
Message
Author
User avatar
Alan H
Posts: 24067
Joined: July 3rd, 2007, 10:26 pm

The future of AI...

#1 Post by Alan H » December 8th, 2017, 1:22 pm

Alan Henness

There are three fundamental questions for anyone advocating Brexit:

1. What, precisely, are the significant and tangible benefits of leaving the EU?
2. What damage to the UK and its citizens is an acceptable price to pay for those benefits?
3. Which ruling of the ECJ is most persuasive of the need to leave its jurisdiction?

Lord Muck oGentry
Posts: 634
Joined: September 1st, 2007, 3:48 pm

Re: The future of AI...

#2 Post by Lord Muck oGentry » December 10th, 2017, 12:39 am

I'm not sure whether it's frightening or exciting, but it's certainly astonishing. Stockfish's Elo rating is 3391, and AlphaZero's rating on the strength of these results should, as I understand it, be about 100 points higher. To get it into context: a rating difference of 300 points predicts a score of 85% for the stronger player in a reasonably long series of games. A novice would be about 1000, a good beginner 1300-1400, a strong club player 1900, an expert or candidate master 2200, a grandmaster 2500. Only a few world champions and contenders have achieved ratings above 2800.
Stockfish towers over machines that tower over the strongest humans. And AlphaZero has just crushed Stockfish without the loss a of a single game.

Ummm... I find myself thinking of Iain M Banks' super-intelligent Ships, and hoping that this chess monster doesn't turn out to be a Torturer- class Rapid Offensive Unit.

OK, I'm frightened. :D
What we can't say, we can't say and we can't whistle it either. — Frank Ramsey

User avatar
Alan H
Posts: 24067
Joined: July 3rd, 2007, 10:26 pm

Re: The future of AI...

#3 Post by Alan H » December 10th, 2017, 1:04 am

Lord Muck oGentry wrote:
I'm not sure whether it's frightening or exciting, but it's certainly astonishing. Stockfish's Elo rating is 3391, and AlphaZero's rating on the strength of these results should, as I understand it, be about 100 points higher. To get it into context: a rating difference of 300 points predicts a score of 85% for the stronger player in a reasonably long series of games. A novice would be about 1000, a good beginner 1300-1400, a strong club player 1900, an expert or candidate master 2200, a grandmaster 2500. Only a few world champions and contenders have achieved ratings above 2800.
Stockfish towers over machines that tower over the strongest humans. And AlphaZero has just crushed Stockfish without the loss a of a single game.

Ummm... I find myself thinking of Iain M Banks' super-intelligent Ships, and hoping that this chess monster doesn't turn out to be a Torturer- class Rapid Offensive Unit.

OK, I'm frightened. :D
I for one welcome our new Skynet Overlords...
Alan Henness

There are three fundamental questions for anyone advocating Brexit:

1. What, precisely, are the significant and tangible benefits of leaving the EU?
2. What damage to the UK and its citizens is an acceptable price to pay for those benefits?
3. Which ruling of the ECJ is most persuasive of the need to leave its jurisdiction?

User avatar
animist
Posts: 6522
Joined: July 30th, 2010, 11:36 pm

Re: The future of AI...

#4 Post by animist » December 10th, 2017, 11:30 am

maybe you should have titled this thread "The future of natural intelligence, if any"

User avatar
Alan H
Posts: 24067
Joined: July 3rd, 2007, 10:26 pm

Re: The future of AI...

#5 Post by Alan H » December 10th, 2017, 3:42 pm

animist wrote:maybe you should have titled this thread "The future of natural intelligence, if any"
:laughter:
Alan Henness

There are three fundamental questions for anyone advocating Brexit:

1. What, precisely, are the significant and tangible benefits of leaving the EU?
2. What damage to the UK and its citizens is an acceptable price to pay for those benefits?
3. Which ruling of the ECJ is most persuasive of the need to leave its jurisdiction?

User avatar
Alan H
Posts: 24067
Joined: July 3rd, 2007, 10:26 pm

Re: The future of AI...

#6 Post by Alan H » February 13th, 2018, 1:47 pm

Alan Henness

There are three fundamental questions for anyone advocating Brexit:

1. What, precisely, are the significant and tangible benefits of leaving the EU?
2. What damage to the UK and its citizens is an acceptable price to pay for those benefits?
3. Which ruling of the ECJ is most persuasive of the need to leave its jurisdiction?

Post Reply