Remember that QWOP game from years ago? Well, Deepmind's AI just mastered it. | 還記得遊戲QWOP嗎?Deepmind's AI 繼圍棋後再一次傷害人類的尊嚴
Hey guys, still remember the infamous flash game called "QWOP" from years ago? If not, here is some GIFs to recall
還記得一隻很難的Flash遊戲QWOP嗎?有點印像?來看以下的GIF 來回憶一下!
![](https://steemitimages.com/0x0/http://25.media.tumblr.com/tumblr_m2cbmaVfGS1qk9qb8o1_500.gif)
Or some really awesome cosplay...
或一些非常棒的cosplay
![](https://steemitimages.com/0x0/https://media.giphy.com/media/1ecijSSplLBSw/giphy.gif)
![](https://steemitimages.com/0x0/https://media.giphy.com/media/GecUzBPU9tGik/giphy.gif)
Want to go through the struggle again?
想重溫一下當年與鍵盤的激情?按以下連接吧!
How far you can go? 5 meters? Not bad.
But you will never be as good as Deepmind now. Why?
你能走多遠? 5米? 不錯。有進步空間!
但是 Deepmind 不打算給你有任何進軍專業的希望!(給人活路好嗎?)
![](https://steemitimages.com/0x0/https://media.giphy.com/media/BmTGkPAs7OzDy/giphy.gif)
Deepmind, a Google's artificial intelligence company which developed the AI Go player - AlphaGo , has just developed another AI which managed to teach itself to walk, jump and climb without any superintendence.
Programmers at Deepmind were incentivized the AI model to go from point A to point B with no prior instructions. The AI was never shown what walking is like. However, the AI learned from trying and failing and finally came up with a way to finish the mission-seems-to-be-impossible task. As you can see from the simulation, it walked awkwardly but it did a good job of going from one point to another even if there is obstacle ahead.
開發人工智能 AlphaGo 的 Google 子公司 DeepMind,近日公開了另一個有關AI模型的片段。在沒有任何監督下,AI模型自行學習走路,甚至做出跳、爬等動作。特別的是,程式人員只設定了AI模型要到達的地點,但沒有教它們走路等技巧(輸入分析數據)。
然而,AI從失敗和嘗試當中最終學到了如何完成似乎是不可能完成的任務。從模擬影片中可以看出,儘管它走起來很笨拙,但是即使有障礙,也能從一個點走到另一個點。
Here is the video
My take-away from the simulation video
It may sound quite silly, but I honestly inspired by this awkwardly walking AI. It shows me that even it seems like a mission impossible, give it a try and learn from failures. (And given enough time of course) You may be surprised by what you have learned.
儘管聽起來可能很愚蠢,但我真的被這笨拙的AI啟發了。 即使看起來像一個任務是不可能的,試一試從失敗中學習。你可能會驚訝於你所學到的東西。
希望你喜歡以上資訊。 歡迎給讚和在下面討論你跟QWOP的過去 :) 謝謝!
你也可以到我個人頁面查看我以前的帖子
https://steemit.com/@nuagnorab
对啊,以前有玩过,好好玩
想當年在網上跟朋友比,看誰能走更遠!這遊戲非常難,就像是重新用鍵盤學走路!沒想到現在AI又輕鬆制霸了人類。
AI简直太逆天了,以后我们人类的很多工作都要被AI替代了,看来我得考虑一下失业后的问题了。
可不是嗎,Deepmind 一點都不給人留活路呀。。。
沒有玩過這個遊戲,所以看不出有趣的地方。
這遊戲就像是重新用鍵盤學走路。只利用4鍵,分別是"Q", "W", "O", "P" 來控制遊戲里的運動員向前行。文內有遊戲連結,要給你那2個可愛的小女孩玩玩嗎?(金魚叔叔腔)
恐怕她們會只掛着笑那被操控的人的姿勢而忙了為什麼要玩
想當年成功用膝頭磨到100米
高手!可惜了
Congratulations! This post has been upvoted from the communal account, @minnowsupport, by nuagnorab from the Minnow Support Project. It's a witness project run by aggroed, ausbitbank, teamsteem, theprophet0, and someguy123. The goal is to help Steemit grow by supporting Minnows and creating a social network. Please find us in the Peace, Abundance, and Liberty Network (PALnet) Discord Channel. It's a completely public and open space to all members of the Steemit community who voluntarily choose to be there.
If you like what we're doing please upvote this comment so we can continue to build the community account that's supporting all members.
This post got a
3.45
% upvote thanks to @nuagnorab - Hail Eris !QWOP is pretty simple if you get really coordinated.
*(i can go on forever actually. xD)
Wow @yukimaru! You are doing great at this game! I can barely pass the starting line. :P
When I first heard about this Deepmind's AI taught itself to walk I immediately think of QWOP.
I just played it again and just took a screenie. For a simple yet useful tip, focus on not pressing Q. Feel the rhythm of w then o & p. You'll get the drift sometime :)
@nuagnorab - I scared with a 1.9% upvote from @randowhale you lost a lot of money. You can find more about the value of randowhale upvotes at my blog where I track series of randowhale upvotes and their value. At the moment I scare all around and below 4% is a loss. 1.9% is a big loss