|
Chess discussion continues here |
On November 17 2013 02:13 The_Red_Viper wrote:Show nested quote +On November 17 2013 01:49 Lysteria wrote:On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily? There isn't a lot of games where a computer isn't stronger than a human being, the main difference is for chess, people are/were interested to develop it. If you give money, time and manpower to produce it, any computer will destroy everyone in Starcraft/DotA/whatever the game is. What I love about chess is thinking about lines and answers by myself. If you use computers as a tool to be better and not as a mean to win games without using your own brain, I don't see why you wouldn't enjoy it. It is true, but it is way harder in those games and with the current power of pc's i don't think that would really be viable. That being said i don't understand his question either, with unlimited ressources pc's will be better than humans in everything, so yeah.
You are wrong. It would take some time and resources, but it would be perfectly viable to make an AI for SC2 or Dota that would crush any human player, and you wouldnt need a supercomputer at all, homecomputers would be able to do it just fine. And that is just when it comes to the strategy. The AI would macro perfectly and micro beyond what is possible for any human.
A Ferrari is faster than any sprinter, a hydraulic press can move alot more than any weightlifter. Ofcourse specialized software/machines will beat humans. I am just surprised somebody has a lower opinion about chess because the same goes for the game.
|
On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily?
I enjoy team games very much and tried to get into chess but motivation left me because i always knew that the most profitable moves can be calculated via computer. I really wish this was not the case.
So you don't play games or sports because there'll be always someone better than you?
Until chess is solved, and that is still some way off, human players will always be more creative than a computer. That alone is worth watching for.
|
Heia magnus! great question from the norwegian interviewer, asking if vishy could elaborate on what he meant by doing his best. I'd be annoyed too, I mean they get like 2 minutes between the game and the press conference.
|
On November 16 2013 23:46 Ethenielle wrote:
Fair question? You can't be serious.
Do you think he has a special drink prepared in case he loses 2 games? Maybe a dolphin mascot he'll take a swim with? Of fuckin course he'll do his best, there's nothing else TO do.
Jesus christ I hate sports "journalists" more than anything. Just an endless stream of dumb questions.
"How do you feel after practically winning the world championships, Magnus?"
Very hard to figure that one out.
I wholeheartedly agree with this post.
Basically the journalists asked the same question three times in a row. What makes you think he will give a different answer the third time? And what answer are you expecting really? Have you no imagination?
Just to segway a little; I think a lot of post-game interviews in SC2 tournaments suffered from this same issue. I like it a lot that many tournaments have really cut down on the post-game BS. So rarely is there anything said worth saying.
Journos and hosts should really try to make an effort and come up with some more interesting questions. Then again, I think theres too much 'journalism' and 'news making' going on anyway. Really, do we need a hall full of them at a press conference? Does every single thing have to be explicated so far?
|
On November 17 2013 02:13 The_Red_Viper wrote:
It is true, but it is way harder in those games and with the current power of pc's i don't think that would really be viable.
That being said i don't understand his question either, with unlimited ressources pc's will be better than humans in everything, so yeah.
I don't really think so, when you look at some AIs for BW/SC2 made by 2/3 people without even being paid and with their own computers, it already shows "perfect" micro, and perfect execution of build orders. They still suck most of the time strategy wise, but I'm pretty sure with enough money, it would be possible. Especially since it's theorically easier for a human to be close to an AI in chess, you don't have to have 10 thousand APM and can take way more time to use your brain. 
It's just, I don't see any professionnal group interested in programing an AI for a specific game no one will play in 20-30 years, when you have chess next to it, played for centuries, by millions.
|
On November 17 2013 02:47 Ravnemesteren wrote:Show nested quote +On November 17 2013 02:13 The_Red_Viper wrote:On November 17 2013 01:49 Lysteria wrote:On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily? There isn't a lot of games where a computer isn't stronger than a human being, the main difference is for chess, people are/were interested to develop it. If you give money, time and manpower to produce it, any computer will destroy everyone in Starcraft/DotA/whatever the game is. What I love about chess is thinking about lines and answers by myself. If you use computers as a tool to be better and not as a mean to win games without using your own brain, I don't see why you wouldn't enjoy it. It is true, but it is way harder in those games and with the current power of pc's i don't think that would really be viable. That being said i don't understand his question either, with unlimited ressources pc's will be better than humans in everything, so yeah. You are wrong. It would take some time and resources, but it would be perfectly viable to make an AI for SC2 or Dota that would crush any human player, and you wouldnt need a supercomputer at all, homecomputers would be able to do it just fine. And that is just when it comes to the strategy. The AI would macro perfectly and micro beyond what is possible for any human. A Ferrari is faster than any sprinter, a hydraulic press can move alot more than any weightlifter. Ofcourse specialized software/machines will beat humans. I am just surprised somebody has a lower opinion about chess because the same goes for the game.
I dont think that is true for something like sc2, unless you give the AI perfect information (like in chess naturally)
|
On November 17 2013 03:00 The_Red_Viper wrote:Show nested quote +On November 17 2013 02:47 Ravnemesteren wrote:On November 17 2013 02:13 The_Red_Viper wrote:On November 17 2013 01:49 Lysteria wrote:On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily? There isn't a lot of games where a computer isn't stronger than a human being, the main difference is for chess, people are/were interested to develop it. If you give money, time and manpower to produce it, any computer will destroy everyone in Starcraft/DotA/whatever the game is. What I love about chess is thinking about lines and answers by myself. If you use computers as a tool to be better and not as a mean to win games without using your own brain, I don't see why you wouldn't enjoy it. It is true, but it is way harder in those games and with the current power of pc's i don't think that would really be viable. That being said i don't understand his question either, with unlimited ressources pc's will be better than humans in everything, so yeah. You are wrong. It would take some time and resources, but it would be perfectly viable to make an AI for SC2 or Dota that would crush any human player, and you wouldnt need a supercomputer at all, homecomputers would be able to do it just fine. And that is just when it comes to the strategy. The AI would macro perfectly and micro beyond what is possible for any human. A Ferrari is faster than any sprinter, a hydraulic press can move alot more than any weightlifter. Ofcourse specialized software/machines will beat humans. I am just surprised somebody has a lower opinion about chess because the same goes for the game. I dont think that is true for something like sc2, unless you give the AI perfect information (like in chess naturally)
Yes it would. An AI would be able to multitask scouting alot better than a human, and it would be able to calculate/narrow down all possible techroutes from how much gas was left in nodes (if lets say it had overlords checking gasnodes). Having single units check every little nook and cranny, probing every part of a defence, is very taxing for a human while macroing, but for a well programmed AI it would be easy.
What I am saying, and what other people are saying, is that it would be completely viable to make such an AI if time and resources was put into it. Personally I think the gap between a proper AI in SC2 and a human would be bigger than the gap is between software and humans in chess (Because an AI in SC2 could have mechanics that would be impossible for a human to reach).
|
On November 17 2013 03:00 The_Red_Viper wrote:Show nested quote +On November 17 2013 02:47 Ravnemesteren wrote:On November 17 2013 02:13 The_Red_Viper wrote:On November 17 2013 01:49 Lysteria wrote:On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily? There isn't a lot of games where a computer isn't stronger than a human being, the main difference is for chess, people are/were interested to develop it. If you give money, time and manpower to produce it, any computer will destroy everyone in Starcraft/DotA/whatever the game is. What I love about chess is thinking about lines and answers by myself. If you use computers as a tool to be better and not as a mean to win games without using your own brain, I don't see why you wouldn't enjoy it. It is true, but it is way harder in those games and with the current power of pc's i don't think that would really be viable. That being said i don't understand his question either, with unlimited ressources pc's will be better than humans in everything, so yeah. You are wrong. It would take some time and resources, but it would be perfectly viable to make an AI for SC2 or Dota that would crush any human player, and you wouldnt need a supercomputer at all, homecomputers would be able to do it just fine. And that is just when it comes to the strategy. The AI would macro perfectly and micro beyond what is possible for any human. A Ferrari is faster than any sprinter, a hydraulic press can move alot more than any weightlifter. Ofcourse specialized software/machines will beat humans. I am just surprised somebody has a lower opinion about chess because the same goes for the game. I dont think that is true for something like sc2, unless you give the AI perfect information (like in chess naturally)
I am pretty sure the AI would have a huge advantage on micro/macro alone in macro games. So basically you just need to scout and react to cheese.
Imagine TvP with perfect ghost control (no storms), perfect split vs banes, perfect immortal all-ins in PvZ, perfect blink in PvP. Perfect split vs mines and tanks as Zerg.
|
ok i didnt respond correctly, i think its not viable to program such an AI, yes there are AI's that do ONE microtechnique perfectly, but there are so many things you would have to consider, i just think it is way harder to find algorithms for a real time game..
But i think that is enough for that, this thread is about chess the last time i checked..^^
|
These press conferences are lame anyway. For many reasons:
- Players are veery tired after each game and want just to go outside and relax instead of answering questions.
- Neither of those guys are particularily outspoken; they just shrug off whatever issue they are asked about.
- In terms of post-game analysis, neither side wants to reveal his thought process to the opponent.
- Journalists are so lame, even I could ask better questions. Like why Vishy played h4-h5 instead of just waiting.
IMHO, such conferences, if ever, should be arranged like 1 hour before games, like before big football matches.
|
On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily?
I enjoy team games very much and tried to get into chess but motivation left me because i always knew that the most profitable moves can be calculated via computer. I really wish this was not the case.
thought about it too. then again, people enjoy watching powerlifting even though a machine could lift stuff which is several magnitudes of weight heavier right? it will be interesting to see how chess will fare once it has been completely figured out, as in a perfect computer software is developed.
|
I would say that computer engines actually make chess more interesting as a whole.
It is true that some of the mystery is gone like when world championship games were analyzed for months afterwards without reaching definite conclusions.
However, nowadays everyone can use an engine at home which makes the game much more accessible and easier to learn.
Also, one can much more appreciate how good the top level players actually are.
As an exercise, go through any top GM game and try to make a move in each position.
Then compare your move with the engine evaluations.
How much does your move's evaluation deviate from the player's move?
Add these deviations up. This shows you how well you would do.
|
On November 17 2013 02:48 Serek wrote:Show nested quote +On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily?
I enjoy team games very much and tried to get into chess but motivation left me because i always knew that the most profitable moves can be calculated via computer. I really wish this was not the case. So you don't play games or sports because there'll be always someone better than you? Until chess is solved, and that is still some way off, human players will always be more creative than a computer. That alone is worth watching for.
Like Carlsen, who plays 95% computer moves? The better you are at chess, the more you will play like a computer. That's not a bad thing, since computers don't play boring chess and they don't arrive at the move choice by the same method anyway.
On November 17 2013 04:21 urboss wrote:
I would say that computer engines actually make chess more interesting as a whole.
It is true that some of the mystery is gone like when world championship games were analyzed for months afterwards without reaching definite conclusions.
However, nowadays everyone can use an engine at home which makes the game much more accessible and easier to learn.
Also, one can much more appreciate how good the top level players actually are.
As an exercise, go through any top GM game and try to make a move in each position.
Then compare your move with the engine evaluations.
How much does your move's evaluation deviate from the player's move?
Add these deviations up. This shows you how well you would do.
Watching chess with your engine is like watching Starcraft 2 with full map vision for the spectator, I think, it's quite nice and allows you to understand the objective quality of the position (together with the commentators) even beyond the understanding of the players.
--
Also, have about a dozen phd AI researchers work full-time on a Starcraft 2 AI purely with the intention to crush all human opposition and they will achieve this goal within the year.
|
Is there any way to just see the moves without all the fucking "WHAT IF HE DOES THIS" crap from the commentators? Like just a gif playthrough of each game?
|
Russian Federation823 Posts
On November 17 2013 02:48 Serek wrote:Show nested quote +On November 17 2013 01:33 kusto wrote:
I have a serious question (no trolling) to all chess people here:
How can you find passion in a game in which a computer is better than a human being? Why does this not prevent you from enjoying this, if it's demystified so easily?
I enjoy team games very much and tried to get into chess but motivation left me because i always knew that the most profitable moves can be calculated via computer. I really wish this was not the case. So you don't play games or sports because there'll be always someone better than you?
No. Read my post again.
|
|
|
Russian Federation823 Posts
You can't compare the problem with SC/BW because these are games where you don't have map hack, therefore incomplete information. Humans have an advantage because they can decide a lot easier when for example scouting in a certain manner can pay off.
Also, you can have potentially infinite decision trees (not computable with finite resources) because you don't have discrete moves but a possible move for each timestep. So a scenario where an AI makes quasi-perfect decisions im BW/SC2 is impossibly expensive to compute.
|
On November 17 2013 04:46 kusto wrote:
You can't compare the problem with SC/BW because these are games where you don't have map hack, therefore incomplete information. Humans have an advantage because they can decide a lot easier when for example scouting in a certain manner can pay off.
Also, you can have potentially infinite decision trees (not computable with finite resources) because you don't have discrete moves but a possible move for each timestep. So a scenario where an AI makes quasi-perfect decisions im BW/SC2 is impossibly expensive to compute.
Are you by chance a second year Computer Science student with no real experience with building an AI?
|
The fact that SC has fog of war doesn't matter if your goal with the AI isn't to solve the game. If all you want to do is make an AI that beats all human players, you just have to program it to do something that will beat human players, like execute some micro-intensive all-in and execute it perfectly. Compare that to how it took hundreds of millions of dollars' worth of research to get chess engines to the point they are now.
The way chess engines are designed doesn't point towards solving chess, either. They rely heavily on human-programmed "evaluation" functions to tell them how good a certain position is because it's still unfeasible to search all possibilities even 10 turns ahead. To solve chess, you'd have to either examine every single possible game or demonstrate somehow that the ones you didn't examine weren't worth examining.
|
On November 17 2013 04:46 kusto wrote:
You can't compare the problem with SC/BW because these are games where you don't have map hack, therefore incomplete information. Humans have an advantage because they can decide a lot easier when for example scouting in a certain manner can pay off.
Also, you can have potentially infinite decision trees (not computable with finite resources) because you don't have discrete moves but a possible move for each timestep. So a scenario where an AI makes quasi-perfect decisions im BW/SC2 is impossibly expensive to compute.
Good point. Your fallacy however is in the fact that you expect the computer to play the game like humans. Scouting/decision-making can be made near obsolete or at least trivial with enough APM and blink stalkers. At least that is my theory. It would probably be easy to create an AI that would defeat the mass blink stalker AI. Furthermore creating "the perfect AI" would most likely be an almost impossible task. However that is not the topic, the topic was to create one that can defeat humans.
To bring the discussion back around to chess and answer your first question simultaneously. The motivations for immersing yourself in games are usually one of two (or both). The ability to defeat the opposition (take some of the opponents pie) or the striving for continual improvement (grow my own pie). Whether or not a computer is generally better than humans is irrelevant. The invention of computers able to better evaluate positions than humans does not make the evaluation of positions for humans easier, when you do, in fact not, have a computer. Unless of course you're capable of remembering every single evaluation (not likely). All the computers are doing is make analysis of chess trivial because you can just ask the computer to do it.
|
|
|
|
|
|
EPT
