|
United States33077 Posts
Via: Yonhap News
As Google/DeepMind's artificial intelligence AlphaGo continues to roll against top Baduk pro Lee Se-dol, StarCraft has emerged as a potential future target for the AI.
Consensus greatest-of-all-time Brood War pro Flash expressed cautious confidence in a short interview on Thursday. Another Terran legend, Boxer, mirrored his sentiments in an interview with Yonhap News. Here are Boxer's quotes from the article:
- "I don't know how smart [AlphaGo] is, but even if it can win in Baduk(Go), it can't beat humans in StarCraft."
- "It would be a mistake to think artificial intelligence could beat humans in StarCraft. StarCraft is a game where situational strategy is far more important than in Baduk, so it's an area where AI cannot catch up."
- "There are many variables in StarCraft, such as scouting, obviously, as well as maps, racial balance, micro, mind-games, etc."
- "Even if countless data is inputted and studied by the AI so it has some degree of instinct, it won't reach pro level."
- "If such an offer comes in the future, I'll gladly accept."
- "Even if it has studied all of the many strategies I've used, I'll go at it with an unstoppable strategy I've prepared."
- "Competing in a few StarCraft tournaments nowadays, I felt that this is where my roots are. It's exciting just thinking about facing a machine as mankind's representative."
|
I would assume Starcraft would be even easier for the AI because mechanics are so important. Just consider how well the simple blink/micro bots work.
|
Im not sure about this. AI has the disatvantages of the points which are stated, sure why not, but an AI can also have thousands of APM and just perfectly muta micro and harass like never faced against a human before.
But it should be much harder for an AI to "understand" Starcraft on pro level than chess or GO, because of their static nature. Whereas SC has many variables and unknown factors to work with.
|
On March 13 2016 02:45 greenelve wrote: Im not sure about this. AI has the disatvantages of the points which are stated, sure why not, but an AI can also have thousands of APM and just perfectly muta micro and harass like never faced against a human before.
But it should be much harder for an AI to "understand" Starcraft on pro level than chess or GO, because of their static nature. Whereas SC has many variables and unknown factors to work with.
They would have to implement APM constraints in the AI of course, otherwise the whole experiment would be useless.
|
Automaton 2000 Micro + Alpha Ai R.I.P
|
All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited.
|
Katowice25012 Posts
I mean your duty as a SC player is to thump your chest and say "no machine will beat me!" right? It's not like anyone is going to have a reasonable discussion about this in interviews.
|
I think it's pretty cool we get these statements, might make it more possible for the event to happen, since they seem so confident.
|
|
it's not a question of "if," it's a question of when. maybe not in 5 years, maybe not in 10 years, but nothing is going to stop AI from getting better and becoming able to excel in complex tasks. they said the same thing about chess, same thing about go, same thing about lots of computerized tasks. it's cute that he thinks it's not possible, but there's no reasonable argument outside of "when will it happen"
On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited. sorry for finding science interesting!
|
Boxer is definitely Fan Hui in this scenario, not Lee Sedol.
|
On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited.
people thought Go AI beating professional players was a long way off. AFAIK deepmind's project is a generalized solution that takes only pixel data as an input.
could be far, far closer than you might think.
+ Show Spoiler +
+ Show Spoiler +
|
its possible to get wrecked by superior micro.
|
A well programmed AI would have an easy time with perfect micro and macro. Decision making would be tougher of course...
|
The same has the complete Go community said about AlphaGo and also Lee Sedol said, he will win so easy against AlphaGo. It became a train wreck... for the Go community.
Same will happen with AlphaStarcraft for Boxer, Flash, Bisu and the complete Community.
|
Canada8988 Posts
On March 13 2016 03:01 Garrl wrote:Show nested quote +On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited. people thought Go AI beating professional players was a long way off. AFAIK deepmind is a generalized solution that takes only pixel data as an input. could be far, far closer than you might think.
Well it could also be far far longer than you might think, it the 60th it was pretty clear for a lot of people including scientist, that 50 years later, there was going to be daily space travel, and lunar colony. Yet here we are in 2016 and we are still far of doing that. The time it take to solve science problem is hard to predict.
|
I think people massively underestimate what perfect mechanics does to the game It depends on the rules/limitations placed on the AI, but imagine
* Every Medivac always picking up units about to be hit by a stalker and immediately dropping it for the next shot * Marines that always maintain their range advantage on roaches * Tanks that always target the banelings first * Marines that always perfect split v banelings (you can find that online already) * Weak units that always rotate out of the front line * Medivacs healing the most important target in range, rather than the closest * Perfect charges vs tank lines (single units charging ahead of the main attack * ...to name a very few basic micro tricks
And while all this happens, perfect macro? Humans overestimate themselves . Computers won't even need "good" strategy to beat humans, just a large number of difficult to handle micro tricks and beastly macro. The "AI" that will need to be added is just to stop the computer glitching out against weird tricks (e.g. somehow tricking the AI into permanent retreat based on units trying to find perfect range.
Edit: Humans are actually at an advantage in Chess and Go, because they are put under far less real time pressure. Also - I realise I'm talking about SC2 here, not BW, but a lot would apply. I think it's a real shame that we can't have AI competitions in SC2,
|
On March 13 2016 02:56 brickrd wrote:it's not a question of "if," it's a question of when. maybe not in 5 years, maybe not in 10 years, but nothing is going to stop AI from getting better and becoming able to excel in complex tasks. they said the same thing about chess, same thing about go, same thing about lots of computerized tasks. it's cute that he thinks it's not possible, but there's no reasonable argument outside of "when will it happen" Show nested quote +On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited. sorry for finding science interesting!
maybe there will come a day when instead of watching WCS between 2 humans we'll see a WCS between 2 AI teams that build AIs to play the game 
On March 13 2016 03:01 Garrl wrote:Show nested quote +On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited. people thought Go AI beating professional players was a long way off. AFAIK deepmind's project is a generalized solution that takes only pixel data as an input. could be far, far closer than you might think. + Show Spoiler +https://www.youtube.com/watch?v=ePv0Fs9cGgU
Atari 2600 Space Invaders where bullets disappear and reappear in mid-air because the system can only handle a few independently moving objects at any one time?
its a pretty bad example.. the arcade version would be a much truer test.
|
On March 13 2016 03:16 Nakajin wrote:Show nested quote +On March 13 2016 03:01 Garrl wrote:On March 13 2016 02:50 Musicus wrote: All this talk about something that might or might not happen in 5 to 10 years. When the challenge is out and the date is set, I will get excited. people thought Go AI beating professional players was a long way off. AFAIK deepmind is a generalized solution that takes only pixel data as an input. could be far, far closer than you might think. Well it could also be far far longer than you might think, it the 60th it was pretty clear for a lot of people including scientist, that 50 years later, there was going to be daily space travel, and lunar colony. Yet here we are in 2016 and we are still far of doing that. The time it take to solve science problem is hard to predict. lunar colonies have nothing to do with adaptive AI winning at complex competitive games... the fact that an AI beat the go champion earlier than expected has everything to do with it...
just because we don't know for certain that doesn't mean that we shouldn't use relevant reference points to make predictions...
|
As a programmer, I'm actually quite skeptical that Boxer would be able to beat an intelligent AI given that a team of scientists are given enough time to develop one. It's inevitable.
|
|
|
|