个人资料
正文

杞人人工智能担忧政治的天

(2018-04-21 04:26:12) 下一个

人工智能专家:习近平读我的书让人“既高兴又害怕”

2018年4月21日 06:27

军事、医疗、司法、管理、交通…… 人工智能正在各行各业改变人们的生活。有一本既被科技从业者认可,也受到普通读者欢迎的科普读物《终极算法》(The Master Algorithm)讲述了人工智能正在如何改变社交、科学、商业甚至战争。

这本由华盛顿大学计算机学教授佩德罗•多明戈斯(Pedro Domingos)撰写的书籍有不少著名的“粉丝”,包括微软公司的创始人比尔·盖茨(Bill Gates)和谷歌公司前首席执行官艾瑞克·施密特(Eric Schmidt)。今年初,《终极算法》的读者群又增加了一个人:中国领导人习近平。

在习近平新年献词的视频中,他身后的书架上除了马克思列宁主义著作和一些经济学、文学作品外,还有关于科技和人工智能的书籍,包括多明戈斯的《终极算法》。

日前,多明戈斯在接受德国媒体《明镜周刊》(Der Spiegel)的访问时表示,习近平读他的书让他感到“既高兴又害怕”(both exciting and scary)。

人工智能与数字极权主义

去年7月,中国国务院公布了《新一代人工智能发展规划》,计划在2030年把中国建设成为“世界主要的人工智能创新中心”,打造1500亿美元的本土产业。

中国三大科技公司百度、阿里巴巴和腾讯都在人工智能领域投入大量资金。此外,还有许多专注于某一领域的人工智能公司,如去年刚刚完成C轮4.6亿美元融资的旷世科技(Face ++),专注于图像识别和深度学习;还有A+轮融资4000万美元的深鉴科技(Deephi)专注于芯片。

人工智能在中国已经被应用于司法和社会管理。据商业媒体Jumpstart报道,中国内蒙古的一座监狱已经将旷世科技的人脸识别技术用于管理监狱,提升监狱安全和管理效率。

同时,公路、机场、车站甚至居民小区都大量安装摄像头。警方从这些摄像头中获取大量的数据,不仅可以应对突发事件,更能做到提前预测人群可能在哪里聚集

多明戈斯表示,世界上不同地方的人对人工智能有不同的预判。美国加州硅谷的人从自由主义的角度对人工智能保持乐观,而欧洲人则更多保持悲观情绪,认为人工智能的发展难以为人所控制。

至于中国和俄罗斯对人工智能的态度,多明戈斯表示:“不幸的事,(中国和俄国)更多看到的是人工智能在威权而不是自由主义方面的潜能。习近平和普京会问自己:我们能拿这个科技做什么呢?”

多明戈斯认为,欧洲与中国相比在人工智能领域已经掉队了。

多明戈斯说,未来人们可能会生活在一个中国主宰的世界,“未来世界可能名义上不受中国控制,但实际上却是,因为虚拟世界被中国统治。”

德国墨卡托中国研究中心的创始人、政治学教授韩博天(Sebastian Heilmann)在其新著《红天鹅:中国非常规决策过程》(Red Swan:How Unorthodox Policy-Making Facilitated China’s Rise)中也表达过类似的担忧。

在与香港端传媒的专访中,韩博天表示:“(人工智能)契合共产党的科层制度与它的组织原则:他们利用数字技术实施全面监控、控制、引导、审查等,新技术对共产党来说是完美的控制工具。”

韩博天把这种依靠大数据提高极权效率的做法称为“数字列宁主义”。他认为,应该担心的是中国的科技巨头和政府的方向完全一致,大公司和政府对数据都非常渴望,它们之间已经形成一种联盟。

对于人工智能应用的担忧,多明戈斯表示,最终决定权并不在人工智能手里,而是在控制人工智能的人手里。“谁会操纵重要的算法?是我们,还是习近平?这才是问题所在。”

 

This AI expert found it “both exciting and scary” that Xi Jinping reads his book

Xi Jinping’s bookshelf includes not only classics on communism but also works on artificial intelligence, as TV viewers spotted during his new year’s speech this year. One of the books that helps the Chinese president understand AI is The Master Algorithm, a 2015 bestseller by Pedro Domingos.

In a recent interview with German magazine Der Spiegel, Domingos, who teaches computer science at the University of Washington, said that when he saw his book on Xi’s bookshelf, he found it “both exciting and scary.” The machine-learning expert explained:

Exciting because China is developing rapidly, and there are all sorts of ways the Chinese and the rest of the world can benefit from AI. Scary because this is an authoritarian government, going full tilt on using AI to control their population. In fact, what we are seeing now is just the beginning. Like any technology, AI gives you the power to do good and evil. So far, we have been focusing on the power to do good, and I think it is enormous. But the power to do evil is there, too.

China has vowed to become an AI powerhouse, with the goal of making the domestic industry worth $150 billion by 2030. The country is already home to some of the world’s most valuable AI startups, such as Face++ and SenseTime, whose facial recognition technologies have been enlisted by police to build surveillance networks nationwide that can identify people quickly. The far western region of Xinjiang, where most of the nation’s Uighur Muslim minority lives, is turning into a laboratory for testing high-end spying technologies.

The Master Algorithm is an introduction to machine learning and how it relates to everyday life. The answer to all the learning problems of AI technology, according to Domingos, is an ultimate “master” algorithm that gives itself feedback to develop endlessly. The book is recommended by Bill Gates as a must-read on AI.

Domingos said autocrats like Xi and Russia’s Vladimir Putin are interested in AI because they “unfortunately see the authoritarian and less the libertarian potential.” China’s advantage in the AI arms race, he said, is the huge data pool that fuels machine learning. In addition, the government and big corporations in China are willing to “help each other with little compunction,” as Domingos predicted leading AI companies will become more nationally controlled during future retrenchments.

China’s big three tech giants—Baidu, Alibaba, and Tencent—are all betting big on AI, focusing on areas ranging from self-driving cars to health care. According to a new cybersecurity law, companies must store all of the data they generate from China inside the nation’s borders. And there have been incidents where tech firms, both foreign and domestic, have been asked to hand over personal data to the Chinese government.

At the end of the day, Domingos said, it’s not AI making the decisions but those who control it. “Who will be steering the major algorithms? Is it us—or is it Xi Jinping? That’s the question.”

“We could end up in a world that China may not formally control, but they effectively do because they rule the cyberworld,” he said.

[ 打印 ]
阅读 ()评论 (0)
评论
目前还没有任何评论
登录后才可评论.