Author Topic: The Standard Theory  (Read 2102 times)

Offline Baruch

Re: The Standard Theory
« Reply #15 on: February 02, 2018, 07:30:03 PM »
You are not allowed to view links. Register or Login
I look at patterns, not numbers. Although when considering a tactical move numbers are involved, such as how many pieces I have threatening a square and how many my opponent has defending it, but mostly it's the patterns that count. When you get right down to it though, according to You are not allowed to view links. Register or Login, there exists nothing but number in the universe.

Sure.  The only component of the universe is Pythagoras, who died about 2500 years ago ;-)

All "theory of everything" I think, per SGOS recent question, are bridges too far.
πŽπŽœπŽœπŽŸπŽŒπŽ€πŽπŽŽπŽ€πŽ€πŽšπŽ€πŽŸπŽπŽœπŽœπŽŸπŽπŽ€πŽπŽ‰πŽ€πŽ€πŽšπŽ€
luu shalmaata luu balt’aata
May you be well, may you be healthy

Offline trdsf

Re: The Standard Theory
« Reply #16 on: February 04, 2018, 03:05:51 PM »
You are not allowed to view links. Register or Login
Which is why I stopped playing computers at chess.  Part of chess is the threat that causes your opponent to move, sometimes to disadvantage. 

A player can be caused to worry about a threat; a computer can't.

And there is another thing that annoys me about playing chess against a computer.  To make the computer "play equal" to me, the computer has to DELIBERATELY make some errors.  DELIBERATELY!  That's cheating.
And yet, it took until Deep Blue in 1997 for a computer to beat a reigning world champion under regular time controls.  Clearly there's more to the game than just data crunching.  Kasparov himself said some of the computer's moves suggested genuine thought and creativity, although I think that's unlikely.

Interestingly, relative to your last point, Nate Silver suggests that it was a flaw in the software that led to Deep Blue's 44th move in game 1 of the rematch, the one that blew Kasparov's mind and may have shaken him enough to cost him the match.  Silver suggests that Deep Blue simply was unable to come up with a most optimal move, and instead played a 'fail safe' move.  Kasparov thought it was so counterintuitive a play, either the machine was really thinking, or a human was intervening.  After that, he lost twice, drew thrice, and went down 3Β½-2Β½.

I was working for IBM at the time.  The intranet coverage was wonderful.
Sir Terry Pratchett, on being told about the theory that the universe is a computer simulation: "If we all get out and in again, would it start to work properly this time?"

Re: The Standard Theory
« Reply #17 on: February 04, 2018, 05:13:36 PM »
Now they have Alpha Zero, that seems to play even more intuitively sometimes.


You are not allowed to view links. Register or Login

Quote
The latest advance from DeepMind behaves in a very surprising way. Expect other AI systems to be just as odd.
You are not allowed to view links. Register or Login
"In life, unlike chess, the game continues after checkmate."
Isaac Asimov

Re: The Standard Theory
« Reply #18 on: February 04, 2018, 06:41:09 PM »
You are not allowed to view links. Register or Login
Threatening your opponent.  Watching him ("usually him") sweat on a move.  Seeing the uncertainty of his moves.  Knowing when the Knight is more dangerous than the Bishop.

Threat levels are quantifiable.   
Winner of WitchSabrinas Best Advice Award 2012


We can easily forgive a child who is afraid of the dark; the real
tragedy of life is when men are afraid of the light. -Plato

Offline trdsf

Re: The Standard Theory
« Reply #19 on: February 04, 2018, 11:43:22 PM »
You are not allowed to view links. Register or Login
Now they have Alpha Zero, that seems to play even more intuitively sometimes.


You are not allowed to view links. Register or Login
This actually shouldn't be a surprise.  There's a sort of assumption that when we talk about an artificial intelligence, we're talking about an artificial human-like intelligence (which I will abbreviate AHLI), like HAL 9000 without the murderous streak (or with it, depending on your thinking about AI).

The thing is, our sensory input is necessarily a part of what makes up our intelligence.  I don't think we'll ever have an AHLI until we can give the machine at least one of the senses we take for granted, and probably more than one.  We can have an AI without that, but we shouldn't expect it to 'think' (and I use the term with caution) like we do because it doesn't experience information input like we do.

So I'm not at all surprised that Alpha Zero doesn't play chess in a recognizable way, but still plays it extremely well.  It doesn't experience the game like we do, assuming it "experiences" the game in any manner we understand.  It wasn't taught by a grandmaster with his/her own personality and prejudices, it didn't read the literature, it doesn't have an experience of the history of the game, it has no concept of personally looking foolish, and it doesn't have an opinion about whether a move "looks weird".  It just has an analysis chain that says this works, and whether it's a theoretical novelty or not is of no relevance whatsoever.

I doubt that Alpha Zero is sentient on any level that we would recognize as sentience in a living entity.  However, it may force us to widen our perspective on what does and does not constitute intelligent behaviorβ€”in the age of SETI, this is probably a good thing.
Sir Terry Pratchett, on being told about the theory that the universe is a computer simulation: "If we all get out and in again, would it start to work properly this time?"

Offline Mr.Obvious

Re: The Standard Theory
« Reply #20 on: February 05, 2018, 04:49:08 AM »
You are not allowed to view links. Register or Login
Ask Mr.Obvious about that...  We have quite a game going.

Next time we play in one setting, with facecam. ;)
"If we have to go down, we go down together!"
- Your mum, requesting 69 last night.


Offline Cavebear

Re: The Standard Theory
« Reply #21 on: February 06, 2018, 10:53:40 PM »
You are not allowed to view links. Register or Login
And yet, it took until Deep Blue in 1997 for a computer to beat a reigning world champion under regular time controls.  Clearly there's more to the game than just data crunching.  Kasparov himself said some of the computer's moves suggested genuine thought and creativity, although I think that's unlikely.

Interestingly, relative to your last point, Nate Silver suggests that it was a flaw in the software that led to Deep Blue's 44th move in game 1 of the rematch, the one that blew Kasparov's mind and may have shaken him enough to cost him the match.  Silver suggests that Deep Blue simply was unable to come up with a most optimal move, and instead played a 'fail safe' move.  Kasparov thought it was so counterintuitive a play, either the machine was really thinking, or a human was intervening.  After that, he lost twice, drew thrice, and went down 3Β½-2Β½.

I was working for IBM at the time.  The intranet coverage was wonderful.

Well, yes.  I wasn't playing a really good computer.  I meant more that I can't beat even a weak computer program.  I depend a lot on making confusing situations.  A computer doesn't care, but a person does.
Atheist born, atheist bred.  And when I die, atheist dead!

 

SMF spam blocked by CleanTalk