霍金等呼籲提防人工智能副作用

Dozens of scientists, entrepreneurs and investors involved in the field of artificial intelligence, including Stephen Hawking and Elon Musk, have signed an open letter warning that greater focus is needed on its safety and social benefits.

數十位科學家、企業家及與人工智能領域有關的投資者聯名發出了一封公開信，警告人們必須更多地注意人工智能(AI)的安全性及其社會效益。參加聯署的人中包括了科學家史蒂芬•霍金(Stephen Hawking)及企業家埃倫•馬斯克(Elon Musk)。

The letter and an accompanying paper from the Future of Life Institute, which suggests research priorities for “robust and beneficial” artificial intelligence, come amid growing nervousness about the impact on jobs or even humanity’s long-term survival from machines whose intelligence and capabilities could exceed those of the people who created them.

這封發自生命未來研究所(Future of Life Institute，簡稱FLI)的公開信還附帶了一篇論文，其中建議應優先研究“強大而有益”的人工智能。目前，人們日益擔心機器的智力和能力可能會超過創造它們的人類，從而影響到人類的就業，甚至影響到人類的長期生存。

“Because of the great potential of AI, it is important to research how to reap its benefits while avoiding potential pitfalls,” the FLI’s letter says. “Our AI systems must do what we want them to do.”

這封FLI的公開信表示：“由於人工智能的巨大潛力，開展如何在規避其潛在陷阱的同時獲取其好處的研究十分重要。我們的人工智能系統，必須按照我們的意願工作。”

The FLI was founded last year by volunteers including Jaan Tallinn, a co-founder of Skype, to stimulate research into “optimistic visions of the future” and to “mitigate existential risks facing humanity”, with a focus on those arising from the development of human-level artificial intelligence.

FLI去年由包括Skype聯合創始人讓•塔林(Jaan Tallinn)在內的志願者創立。成立該研究所的目的一方面是爲了促進對“未來樂觀圖景”的研究，一方面則是爲了“降低人類面臨的現存風險”。這其中，在開發與人類相當的人工智能技術過程中出現的那些風險，將是該所關注的重點。

Mr Musk, the co-founder of SpaceX and Tesla, who sits on the FLI’s scientific advisory board alongside actor Morgan Freeman and cosmologist Stephen Hawking, has said that he believes uncontrolled artificial intelligence is “potentially more dangerous than nukes”.

SpaceX和特斯拉(Tesla)的共同創始人馬斯克、著名演員摩根•弗里曼(Morgan Freeman)以及宇宙學家史蒂芬•霍金都是FLI科學顧問委員會的委員。馬斯克表示，他相信不受控制的人工智能“可能比核武器更危險”。

Other signatories to the FLI’s letter include Luke Muehlhauser, executive director of Machine Intelligence Research Institute, Frank Wilczek, professor of physics at the Massachusetts Institute of Technology and a Nobel laureate, and the entrepreneurs behind artificial intelligence companies DeepMind and Vicarious, as well as several employees at Google, IBM and Microsoft.

這封FLI公開信上的其他聯署人還包括機器智能研究所(Machine Intelligence Research Institute)的執行主任呂克•米爾豪澤(Luke Muehlhauser)，麻省理工學院(MIT)物理學教授、諾貝爾獎得主弗蘭克•維爾切克(Frank Wilczek)，人工智能企業DeepMind和Vicarious的幕後主管，以及幾名谷歌(Google)、IBM和微軟(Microsoft)的員工。

Rather than fear-mongering, the letter is careful to highlight both the positive and negative effects of artificial intelligence.

這封信並不以一封兜售恐懼心理爲目的。與此相反，它十分謹慎地同時強調了人工智能的積極面和消極面。

“There is now a broad consensus that AI research is progressing steadily, and that its impact on society is likely to increase,” the letter reads. “The potential benefits are huge, since everything that civilisation has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools AI may provide, but the eradication of disease and poverty are not unfathomable.”

信中寫道：“如今存在的一個廣泛共識是，人工智能研究正在穩步進展之中，它對社會的影響也很可能會逐漸增大。人類文明所能提供的一切都是人類智慧的結晶。這種智慧被人工智能可能提供的工具放大後，我們能做到什麼是我們無法想象的，不過那樣的話根除疾病和貧困將不再是遙不可及的。從這個意義上說，人工智能有巨大的潛在好處。”

Benefits from artificial intelligence research that are already coming into use include speech and image recognition, and self-driving vehicles. Some in Silicon Valley have estimated that more than 150 start-ups are working on artificial intelligence today.

目前，人工智能研究的部分好處已經成爲現實，其中包括語音識別和圖像識別，以及自動駕駛的汽車。在硅谷，部分人估計如今從事人工智能業務的初創企業超過了150家。

As the field draws in more investment and entrepreneurs and companies such as Google eye huge rewards from creating computers that can think for themselves, the FLI warns that greater focus on the social ramifications would be “timely”, drawing not only on computer science but economics, law and IT security.

人工智能正吸引越來越多的投資，許多創業家和谷歌等企業都在盼望着能通過建立會自主思考的電腦，獲得鉅額回報。對於這種局面，FLI警告說，人們或許應“及時”將更多注意力集中在人工智能的社會後果上，不僅要從計算機科學的角度開展研究，還要從經濟、法律及信息安全的角度開展研究。