网络时代的公地悲剧

It is hard to keep up with the stream of scandals, big and small, involving social networks such as Facebook and Twitter. From unwittingly aiding Russian efforts to subvert elections to finding themselves exploited by extremists and pornographers, they are constantly in trouble.

如今人们很难跟得上涉及Facebook和推特(Twitter)等社交网络的大大小小的丑闻。从不经意间帮助俄罗斯破坏选举，到发现自己被极端主义者和色情作品制作者利用，社交网络陷入一个又一个麻烦。

The latest is YouTube failing to stop videos of children being commented on by paedophiles, while letting advertisements appear alongside them. Only months after Alphabet’s video platform faced an advertiser boycott over extremist videos and had to apologise humbly, companies such as Diageo and Mars are again removing ads.

最新丑闻是YouTube没能制止恋童癖者评论儿童视频，还让广告出现在这些评论旁边。就在这家Alphabet旗下视频平台因极端主义视频而遭到广告商抵制、不得不谦卑地道歉几个月后，帝亚吉欧(Diageo)和玛氏(Mars)等公司再度撤下了广告。

Each scandal produces fresh calls for networks to be treated like publishers of news, who are responsible for everything that appears under their names. Each one forces them further to tighten their “community standards” and hire more content checkers. By next year, Facebook intends to employ 20,000 people in “community operations”, its censorship division.

每次丑闻都引发了新呼吁，要求把社交网络当做新闻出版商那样对待。新闻出版商对于以其名义发表的所有内容负责。每次丑闻都强迫社交网络收紧其“社区标准”，并聘用更多的内容审核员。到明年，Facebook打算在其审查部门“社区运营”雇用2万名员工。

Tempting as it is for publications that have lost much of their digital advertising to internet giants to believe they should be treated as exact equivalents, it is flawed: Facebook is not just a newspaper with 2.1bn readers. But being a platform does not absolve them of responsibility. The opposite, in fact — it makes their burden heavier.

对于被互联网巨头抢走很多数字广告的出版公司而言，让平台受到同等对待是一个诱人的想法。但这是有问题的：Facebook不仅仅是一份拥有21亿读者的报纸。但是，起到一个平台的作用并不免除社交媒体的责任。事实恰恰相反，这一点加重了它们的担子。

A better way to think of Russian political ads, extremist videos, fake news and all the rest is as the polluters of common resources, albeit ones that are privately owned. The term for this is the tragedy of the commons. Open ecosystems that are openly shared by entire communities tend to get despoiled.

对于俄罗斯政治广告、极端主义视频和假新闻之类，更好的思考方式是把它们视为公共资源的污染者，尽管这些资源是私有的。描述这种状况的术语是“公地悲剧”。由整个社区公开共享的开放生态系统，往往会遭到糟蹋。

Garrett Hardin, the US ecologist and philosopher who coined the phrase in 1968, warned that “the inherent logic of the commons remorselessly generates tragedy”, adding gloomily that, “Ruin is the destination toward which all men rush, each pursuing his own best interest in a society that believes in the freedom of the commons.”

在1968年发明这一短语的美国生态学家、哲学家加勒特•哈丁(Garrett Hardin)警告称，“公地的内在逻辑无情地生成悲剧”。他沮丧地补充道，“毁灭是所有人冲向的目的地，在信奉公地自由的社会里，每个人都追逐着自己的最大利益。”

His prime example was the overgrazing of common land, when the number of farmers and shepherds seeking to use the resource of free feed for animals becomes too high. He also cited companies polluting the environment with sewage, chemical and other waste rather than cleaning up their own mess. Rational self-interest led to the commons becoming barren or dirty.

他的主要例子是公地被过度放牧，原因是太多农牧业者寻求利用免费饲料资源来饲养动物。他也提到了企业任由自己排放的污水、化学物和其他废弃物污染环境，而不去清理自己造成的烂摊子。理性自利导致公地变得贫瘠或肮脏。

Here lies the threat to social networks. They set themselves up as commons, offering open access to hundreds of millions to publish “user-generated content” and share photos with others. That in turn produced a network effect: people needed to use Facebook or others to communicate.

这也是社交网络面临的威胁所在。它们自命为公地，向数亿人开放平台，让他们发布“用户生成内容”，并与其他人分享照片。这进而造成了一种网络效应：人们需要使用Facebook或其他平台进行沟通。

But they attract bad actors as well — people and organisations who exploit free resources for money or perverted motives. These are polluters of the digital commons and with them come over-grazers: people guilty of lesser sins such as shouting loudly to gain attention or attacking others.

但它们也吸引了危险分子——那些出于金钱或变态动机而滥用免费资源的个人和组织。这些是数字公地的污染者，伴随着他们的还有过度放牧者：过错较轻的人，比如大声喊叫以吸引注意力，或者攻击别人。

As Hardin noted, this is inevitable. The digital commons fosters great communal benefits that go beyond being a publisher in the traditional sense. The fact that YouTube is open and free allows all kinds of creativity to flourish in ways that are not enabled by the entertainment industry. The tragedy is that it also empowers pornographers and propagandists for terror.

正如哈丁所指出的，这是不可避免的。数字公地营造了巨大的社区效益，超出传统意义上的出版商范畴。YouTube既开放又免费的事实，让各种创意得以迸发，这是娱乐业做不到的。悲剧在于，它也为色情作品制作者和恐怖主义宣传者提供了便利。

So when Mark Zuckerberg, Facebook’s founder, denounced Russia’s fake news factory — “What they did is wrong and we’re not going to stand for it” — he sounded like the police chief in Casablanca who professes to be shocked that gambling is going on in a casino. Mr Zuckerberg’s mission of “bringing us all together as a global community” is laudable but it invites trouble.

所以，当Facebook创始人马克•扎克伯格(Mark Zuckerberg)谴责俄罗斯的假新闻加工厂(“他们的行为是错的，我们将不会容忍这类行为”)时，他听起来像是卡萨布兰卡的警察局长，承认对赌场正在进行的赌博感到震惊。扎克伯格的使命宣言(“让我们所有人汇聚在一个全球社区”)值得赞赏，但也招来了麻烦。

Hardin was a pessimist about commons, arguing that there was no technical solution and that the only remedy was “mutual coercion, mutually agreed upon by the majority”. The equivalent for Facebook, Twitter and YouTube would be to become much more like publishers, imposing tight rules about entry and behaviour rather than their current openness.

当年哈丁对公地感到悲观。他辩称，对此并没有技术解决方案，唯一的补救措施是“大多数人同意的集体胁迫”。对于Facebook、Twitter和YouTube，这种性质的补救将是让它们变得更像出版商，在进入和行为方面实行严格规则，而不是像目前这样开放。

They resist this partly because it would bring stricter legal liability and partly because they want to remain as commons. But every time a scandal occurs, they have to reinforce their editorial defences and come closer to the kind of content monitoring that would change their nature.

它们抵制这么做，一方面因为这会带来更严苛的法律责任，另一方面因为它们想维持公地的状态。但每次爆发丑闻后，它们都不得不强化自己的内容防御，向着改变它们性质的那种内容监控走近一步。

It would cross the dividing line if they reviewed everything before allowing it to be published, rather than removing offensive material when alerted. Defying Hardin, they aspire to a technical solution: using artificial intelligence to identify copyright infringements and worse before their users or other organisations flag them for review.

如果它们不是在得到警告后才删除冒犯性材料，而是在所有内容发布之前都要对其进行审核，那将跨越分界线。跟哈丁意见相左的是，它们有志拿出一项技术解决方案：在用户或其他组织提醒它们进行审核之前，使用人工智能来识别侵犯版权或更恶劣的内容。

More than 75 per cent of extremist videos taken down by YouTube are identified by algorithms, while Facebook now finds automatically 99 per cent of the Isis and al-Qaeda material it removes. It is like having an automated fence around a territory to sort exploiters from legitimate entrants.

如今，被YouTube下线的极端主义视频有75%以上是通过算法识别出来的，而被Facebook清除的伊斯兰国(ISIS)和基地组织(al-Qaeda)材料中，有99%是自动查找出来的。这就像围绕一块领土建起一道自动化栅栏，把合法进入者跟作恶者区分开来。

Machines cannot solve everything, though. If they could exclude all miscreants, the commons would turn into something else. The vision of an unfettered community is alluring but utopias are always vulnerable.

不过，机器无法解决所有事情。如果机器能够阻挡所有坏人，那么公地将变成其他事物。一个不受限制的社区的愿景是诱人的，但乌托邦永远是脆弱的。

john.gapper@ft.com 译者/何黎