If they don't cap the apm at 300 or so, this means nothing. I am not impressed with super speed, only decision making matters. Deepmind, you can do better.
AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
211–220 of 459 posts
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#212Earlier quoted context omitted.
Another advantage was that the AI is just reading the game state through an API, it doesn't have to look on the screen. The game can be difficult to watch from a pro's perspective since they have to constantly click around the map to see what's happening, but the AI has perfect knowledge of everything it is capable of seeing, all without having to physically move a mouse to click on the screen.
Yeah, put an eye tracker in a pro and you'll see that the eyes are constantly changing the focus point, if you can watch the entire scene with the same precision without the need to focus on it you're already at a nice advantage. As an aside, a few pro gamers prefer to play on windowed mode for exactly this reason.
Just supporting this. I remember the uproar when 800x600 resolution was removed as an option in SC2 around 2012[1].
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#213AlphaStar interacted with the StarCraft game engine directly via its raw interface, meaning that it could observe the attributes of its own and its opponent’s visible units on the map directly, without having to move the camera - effectively playing with a zoomed out view of the game Additionally, and subsequent to the matches, we developed a second version of AlphaStar. Like human players, this version of AlphaStar…
yeah I was disappointed to discover it worked this way. don't get me wrong, it's a major accomplishment in AI regardless, but it's a significant advantage and it would be easier for me to appreciate the AI's skill if I didn't have to keep reminding myself that it can see the whole map at once. it's such an information advantage.
You'll likely be happy to hear that this has been (is being) addressed.
I watched the live broadcast of this announcement where they did a recap of all 10 previous matches (against TLO and Mana) and they talked about this concern. During today's announcement they presented a new model that could not see the whole map and had to use the camera movement to focus properly. The deepmind team said it took somewhat longer to train but they were able to achieve the same levels of performance according to their metrics and play-testing against previous version.
However...
They did a live match vs LiquidMana (6th match against Mana) against the latest version (with camera movement) and LiquidMana won! LiquidMana was able to repeatedly do hit-and-run immortal drop harassment in AlphaStar's base, forcing it to bring troops back to defend its base, causing it to fall behind in production and supply over time and ultimately lose a major battle.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#214Earlier quoted context omitted.
The game is difficult to watch, but does anyone honestly believe that an AI is going to have a difficult time parsing the scene if it is trained to do so? That to me just seems like a question of resources. We're pretty good at image recognition and segmentation now, and that's without the unlimited amounts of training data one could generate when using a controlled game environment with a limited range of possible a…
But on the other hand, an AI that beats humans using brute force in a game where it makes a ton of difference isn't much fair too.
The funny thing is that once we're talking about the real world, which will come, that incentive actually reverses.
At that point the name of the game will be maximizing the advantage the body/infrastructure provides the AI, not minimizing it.
Weird.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#215Earlier quoted context omitted.
> How many of those 500 actions are actually useful? Exactly, a human doing 500 APM during intense moments is going to be way different than an AI bursting 1000 APM with pixel-precision during the most crucial moment in a game. TLO spent a ton of time at >1000 APM and walked his army directly into enemy shots all the time. MaNa had much better control at ~400 APM. So APM is really irrelevant to control - for humans.…
Such high actions per minute does not seem fun to me, and possibly a repetitive strain injury waiting to happen.
> a repetitive strain injury waiting to happen.
Yes, I have one from it and wasn't even playing that high (I averaged less than 100 apm). I understand that it's a common problem.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#216Earlier quoted context omitted.
> Even if you limit the AI to max human APM, it's still going to dominate in these micro-heavy battles because it's going to make every one of its actions count. right, and we saw that with the incredible precision with stalker blink micro. There are many ways you could make it more comparable to humans. They have already tried that by even giving it an APM. > You can't fault the AI for winning at the game because of…
I just can't help but feel that nothing AI does will ever be good enough according to this mindset, i.e. true "intelligence" is by definition things that computers cannot do. Beating the world champion in Chess was, at one point, considered an impossible achievement for computers. Now it's considered so routine it doesn't even count as AI according to many. And in a few months when AlphaStar is beating top human play…
> I just can't help but feel that nothing AI does will ever be good enough
It can be good enough in a certain problem space, such as chess. But unlike chess or go, which are purely mental games, Starcraft has large physical component (vision, APM, reaction time). It can make it hard to determine when it has “mastered” this RTS. Like you said, it may be a few more months (years?) before AlphaStar can master Starcraft on “mental” level. The physical component is trivial for a computer, so mastering that is not much of a milestone.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#217Impressive, but there's a couple of things I'd like to see them try one day.
Plug AlphaStar into a robot that physically interacts with a keyboard and mouse to control the game. This "robot" should only have what's relevant for playing the game and emulates a human i.e. a camera that looks at a screen (this is the only knowledge it has of the game), and two arms & hands with five digits that control the mouse and keyboard. Then limit its APM to the best a human can realistically do.
The other thing I want to see them address is the virtual training time. 200 years of StarCraft is insane. LiquidMana has been playing for ~20 years and of course he hasn't played the game 24/7. Lets pretend he has played StarCraft like it's a full-time job since he was 5; 8 hours a day, 5 days a week, for 20 years. That's ~42,000 hours of playing StarCraft.
Develop an A.I. that is only trained for that many hours of virtual game time.
If they can create an A.I. with those requirements, that can defeat top-level players, I will be completely blown away.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#218Earlier quoted context omitted.
This is how it responds to cannon rush. : ) https://www.youtube.com/watch?v=vYdWQjTWTFM
Isn't the point of a cannon rush to build the first cannons where they can't be seen?
Surprisingly not. The trick is usually to build pylons (or other cannons) such that they protect the cannons from being attacked by probes. Building them out of sight is usually too slow as a rush.
Still, he didn't do that either.
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#219Earlier quoted context omitted.
> While they have similar APM to SC2 pros Wasn't the APM closer to half that of the pros? https://storage.googleapis.com/deepmind-live-cms/images/SCII...
This is super deceiving and I'm kind of upset they posted this image, knowing it would mislead people not familiar with the game. The AI sits around during lulls at During the fights, the critical moments in when MaNa would top out at ~600 humanly inaccurate APM (this is 10 inputs per second), the AI would jump up to over 1000 - we don't know exactly what it was doing, but it was presumably pixel-precise. Meanwhile t…
Do you know why TLO's APM is sometimes so large? Did he actually peak at 2000, or is he using a repeater or something like that?
Re: AlphaStar: Mastering the Real-Time Strategy Game StarCraft II
#220Earlier quoted context omitted.
Another way to think about it is that a human brain is mostly doing transfer-learning, on top of a 99%-baked deep net that was wired up during foetal development from our DNA, where that DNA-persisted model has "seen" hundreds of millions of years of training data. Humans don't have to learn to process, recognize, and classify objects in visual sense-data, for example. We can do that from the moment we're born, becau…
That's not how any of this works. We do not have "millions of years" of information encoded into DNA. DNA doesn't store that much data. In fact, it's about 1.6 gigabytes only! And most of that information is basically a ruleset for growing proteins which become our body. All the stuff we've learned about games and so on have come from our current lifetime. I don't have caveman memory for how to fight a tiger.
I said "deep net" for a reason. A DNN model almost always turns out to be far, far smaller than the training data that was used to create it.
For one example: any smartphone's face-recognition feature. Each such feature is a DNN which took millions of hours of face data to train... but the resultant model fits on an ASIC.
Our DNA doesn't directly encode such a model, but it encodes a particular morphogenic chemical gradient, and set of proteins, that go together to make specialized neural "organs" (like your substantia nigra, or your basal ganglia, or your superchiasmatic nucleus, etc.) which manage to serve the same function to your brain that access to a pre-trained "black box" DNN model would serve an untrained NN in achieving transfer learning.