2014年6月22日 星期日

霍金 -- 人工智慧


霍金:很快人類就鬥不過人工智慧了

著名物理學家史蒂芬·霍金(Stephen Hawking)在 2014 615日出席了美國 HBO 頻道的「Last Week Tonight」節目,並與主持人約翰·奧利弗(John Oliver)展開了深刻而有意義的對話(至少霍金的談話是深刻而有意義的),並認為機器人可能「比我們聰明」。

當奧利弗問及霍金他最希望人們能夠理解什麼事情時,霍金回答說:「虛時間(imaginary time)。」

他說道:「 虛時間就像是宇宙空間中的另一個方向,這是我的作品中還未被科幻小說家使用過的一則假說。」

說實話,儘管所有的科幻小說只是在科學家的作品中加入了血和性等元素,那為什麼沒有科幻小說作家在虛時間的基礎上創作故事呢?霍金說道:「他們不理解虛時間。」

當然,筆者也不能假裝理解什麼是虛時間。無論如何,它是一種與時間有關的東西,與每天侵蝕我們的時間朝著不同的方向運行。

不過,奧利弗最希望瞭解的是人工智慧。與眾多人工製成品一樣,人工智慧對人類而言也可能是有害,甚至是致命的。

霍金對此非常肯定:「 人工智慧在並不遙遠的未來可能會成為一個真正的危險。」

但肯定不是在 Google 那群好男孩的掌握之下嗎?霍金認為,這或許將與 Google 那些好男孩的想法無關。因為一般的機器人可以「進行設計改進它自己,使得它們自己比我們所有人都更聰明。」

奧利弗繼續問道:「能​​夠與機器人戰鬥,難道我不應該感到興奮嗎?」

霍金給出的回答很乾脆:「你會輸掉。」

奧利弗開始擔心之前根本不是霍金在與他談話,而有可能是一個充滿智慧的電腦在跟他問答。霍金(或假裝霍金的機器人)回答說:「你是個白痴」。

但這難道不是人類的本質嗎?儘管我們認為我們什麼都知道,至少當我們對自己誠實時,我們最清楚的一件事就是——我們都是傻瓜。我們所不知道的要遠遠超過我們所做到的。

事實上,考慮到世界上可能有很多平行宇宙,奧利弗懷疑是不是在某個平行宇宙裡,他可能會比霍金聰明。

「是的,」霍金回答道,「而且還有一個宇宙,在那裡你很有趣。」
 

Stephen Hawking: AI could be a 'real danger'

In a hilarious appearance on John Oliver's HBO show, the famous physicist muses that robots could "outsmart us."

I don't want to frighten you. I'll leave it to Stephen Hawking to do that.

On June 15, 2014, the famed physicist made an appearance on HBO's "Last Week Tonight," and had a deep and meaningful conversation with host John Oliver. Well, at least Hawking's side of the conversation was deep and meaningful.

Oliver asked him what was the one thing he most wanted people to grasp. Hawking replied: "Imaginary time."

"Imaginary time is like another direction in space," he said. "It's the one bit of my work science fiction writers haven't used."

Honestly, I though that all sci-fi was written merely by adding blood, gore, and sex to the work of scientists. So why haven't sci-fi writers built stories around imaginary time? "They don't understand it," he said.

I can't pretend to understand it either. However, it's something to do with time that runs in a different direction to the time that gnaws at us every day.

Most importantly, however, Oliver wanted to know about artificial intelligence. Like so many artificial things, it carries with it the idea that it could be noxious or even deadly.

Hawking was very reassuring: "Artificial intelligence could be a real danger in the not too distant future."

Oh, but surely not in the hands of the nice boys from Google?

Hawking, though, believes that it might be irrelevant what the nice boys from Google think. For your average robot could simply "design improvements to itself and outsmart us all."

Oliver, channeling his inner 9-year-old, asked: "But why should I not be excited about fighting a robot?"

Hawking offered a very scientific response: "You would lose."

Oliver began to worry that Hawking wasn't talking to him at all. Instead, this could be a computer spouting wisdoms. Hawking (or the robot pretending to be Hawking) replied: "You're an idiot."

But isn't this the essence of humanity? For all that we believe we know, the one thing we know best -- at least when we're honest with ourselves -- is that we're idiots. What we don't know far exceeds what we do.

Indeed, Oliver wondered whether, given that there may be many parallel universes, there might be one where he is smarter than Hawking.

"Yes," replied the physicist. "And also a universe where you're funny."

 

沒有留言:

張貼留言