Humans Compete Against OpenAI FIVE Bots in Dota 2 Showdown

cageymaru

cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
22,423
Former professional and celebrity Dota 2 human players were pitted against OpenAI FIVE bots in a best of three match. The OpenAI FIVE bots were trained utilizing machine learning where they taught themselves to play while constrained by the rules of the game. The hero pool for the match was limited to 18 heroes as the machine learning algorithm had only trained with those combinations.

The humans were absolutely destroyed in the first match as the OpenAI FIVE bots routed them with precise long range sniper shots, slows, and silences. The humans fared much better in the second match as they used tricks to confuse the bots, but a couple of simple mistakes during team fights allowed the bots to raze the base of the humans. Simple mistakes during a normal match might get punished, but it would take a lot of them to determine a winner. Playing against bots and one mistake could cause end of the match. In the third match, the Twitch chat audience chose the worst combination of heroes for the bots to play and the A.I. predicted a 2.9% chance to win. The humans ran over them with fire in their eyes and pure vengeance in their hearts.

OpenAI Five plays 180 years worth of games against itself every day, learning via self-play. It trains using a scaled-up version of Proximal Policy Optimization running on 256 GPUs and 128,000 CPU cores -- a larger-scale version of the system we built to play the much-simpler solo variant of the game last year. Using a separate LSTM for each hero and no human data, it learns recognizable strategies.
 
The human brain will never process at the speed of A.I. In the end humans must resort to tactics that seem illogical and implausible.... Or lose.
 
Nukester said:
The human brain will never process at the speed of A.I. In the end humans must resort to tactics that seem illogical and implausible.... Or lose.
Click to expand...
I can tell you that if a real life sniper A.I. can shoot 25% as the OpenAI FIVE sniper in Dota 2, humans will have a hard time hiding from A.I. hunters in a war. Of course if the military bots are dumb to new situations as the A.I. exhibited at certain points during the match, then we have a good chance to adapt and survive.
 
Nukester said:
The human brain will never process at the speed of A.I. In the end humans must resort to tactics that seem illogical and implausible.... Or lose.
Click to expand...

This is only true for legacy AI. Deep Learning can handle unexpected situations much better. Going into the future with AGI humans will be the ones who cannot keep up with new tactics the AI can come up with.
 
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former. ~ AE
 
Nukester said:
The human brain will never process at the speed of A.I. In the end humans must resort to tactics that seem illogical and implausible.... Or lose.
Click to expand...

So....like the Star Trek original series episode when they had to "short circuit" Henry Mudd's robots with illogical nonsense and bad acting?
 
hAs ScIeNcE gOne ToO FFaaRRRrr? no seriously kill it with fire, though shall not make a computer in the likeliness of a man.
 
Nukester said:
The human brain will never process at the speed of A.I. In the end humans must resort to tactics that seem illogical and implausible.... Or lose.
Click to expand...
For extremely simple, rudimentary calculation this may hold true. But the overall computational power of the human brain is still far greater than even the worlds fastest super computer. The Sunway Taihulight has a processing power of 93 Petaflops. The human brain is estimated to have a processing power of over an exaflop.
 
You must log in or register to reply here.
Back
Top