Dota Talk : The AlphaGo of Dota 2 - OpenAI

in #dota27 years ago

 Just yesterday, the 11th of August 2017, on the main stage of the annual grand event of Dota 2 - The International 2017, a legendary player Dendi played a never-before-seen match against an unexpected "player" in a matchup of 1 versus 1. 


pic: Dendi, a legendary player


Sitting in front of Dendi, isn't exactly a person, but it is a portable drive being inserted to a computer. Yeap, most probably you have guessed it, it was the Dota2 AI. Right before the matchup, the host of the event "Machine“ talked about how is it possible for a bot to so-called even stand a chance to beat Dendi?


Because clearly, everybody knows, how "noob" or weak it is a Dota 2 Bot. Most people get eventually tired or bored playing with Bot after a few games, as they do not really pose any real challenges. However, this Dota 2 Bot with the name "OpenAI" emerged victorous against Dendi is just around fifteen minutes, averaging 7 minutes per game.



It was  a best of three, and the first to get two games will win the matchup. The first match fell short for Dendi in just 10 minutes, and the second match was even more one-sided where Dendi conceded defeat after getting killed just once in the game. "This dude is scary," that's what exactly Dendi said during the game.


This OpenAI artificial intelligence bot is being developed by Elon Musk's research team with the code name OpenAI.



pic: Elon Musk retweeted the news


The official statement given by OpenAI explained that this AI attain the level of understanding the game truly by self-training. Where, it would make copy of itself, and do lots of training with itself. They are making progresses slowly but certainly, and the CTO of OpenAI, Greg Brockman said, "the training only took about 2 weeks, and the AI is already beating several top players in game."

Prior to this event, the "AlphaGo” already beaten humans in the complex go chess game. But right now, in the more complex game of MOBA or multiplayer online battle arena game, which includes much more strategies and actions real time, the AI had indeed made a huge step forward and the step is huge.


With the rapid development of AI and even self-learning ability, will AI finally and eventually becomes  a threat for humans? Maybe, the time will tell.


 -End-  

Sort:  

Congratulations @plokmi! You have completed some achievement on Steemit and have been rewarded with new badge(s) :

Award for the number of upvotes received

Click on any badge to view your own Board of Honor on SteemitBoard.
For more information about SteemitBoard, click here

If you no longer want to receive notifications, reply to this comment with the word STOP

By upvoting this notification, you can help all Steemit users. Learn how here!

will AI finally and eventually becomes a threat for humans?

What's the rush? Ha-ha!

I do have to wonder why we would want to build an AI that could beat the best humans in military simulations. Shall we give them they keys to our drones, next?

Coin Marketplace

STEEM 0.35
TRX 0.12
JST 0.040
BTC 70597.89
ETH 3559.60
USDT 1.00
SBD 4.77