10 Ways the Internet (As We Know It) Will Die

We often think of the Internet as a platform for unfettered global communication, where information flows freely, innovators can launch new applications at will, and everyone can have a voice. But it’s unlikely that our children’s Internet will look anything like what we have now.

How might the Internet as we know it die? Here are 10 possibilities.

1. Someone subverts the Domain Name Service. The Internet relies on DNS. But if someone broke — or worse, subverted — the fundamental way in which we find web sites, we wouldn’t trust URLs any more. Phishing would be easy. Own the DNS and you own the Internet.
2. Zombie networks attack! Untold numbers of enslaved PCs are waiting to do the bidding of shadowy hackers. Matt Sergeant of MessageLabs puts the size of the Storm botnet at between five and 10 million machines (though others peg the size of the network at much less.) Today, bots fill our inboxes with spam. But in the past, they’ve been used to take out companies and countries and to blackmail sites. In the end, it’s an arms race in which only one side has to play by the rules.
3. Massive physical infrastructure failure. If an accident involving a couple of cables in the Mediterranean can make the Internet unusable for hundreds of millions, imagine what an intentional attack could do.
4. Death by a thousand fragments: Ever since Usenet, people have been grouping together with those who think like them. In his book “The Big Switch,” Nicholas Carr cites one study that claimed more than 90 percent of the links originating within either the conservative or liberal community stay within that community. Some link referral tools can even be configured to keep visitors on sites with the same world view. The end result? Islands of like-minded people, increasingly sure there is only one right answer and that they’re in sole possession of it. And an end to the dreams of a global community envisioned by the Internet’s creators.
5. A really good virus breaks the routers. The Internet’s self-healing mechanisms rely on the Border Gateway Protocol, or BGP. But what if someone gets inside the routers? In a 2006 NANOG presentation, Cisco looked at claims of vulnerability and concluded that “the most damaging attacks are caused by the deliberate misconfiguration of a trusted router.” Corrupt BGP, and you not only stop the Internet from forwarding traffic, you interfere with our ability to get to the routers and fix them.
6. Updates break how updates work. Most software these days is designed to patch itself and remain current. But sometimes the process of automated upgrades triggers its own problems. On Aug. 16, 2007, Skype went down in what the company claimed was a side effect of a massive automatic update to Windows. It’s only a matter of time before an update makes a fundamental piece of software, like a networking stack, unable to update itself, cutting off millions and requiring manual intervention.
7. The Net stops being neutral. If the carriers start to charge us for access to sites the way cable companies charge for premium television, pretty soon you’ll have a “Google fee” on your monthly bill. This already happens with many mobile phones that feature the services of Facebook and YouTube. It’s perhaps the most insidious death, because it would signal the end of innovation — no one would be able to launch the next Skype, Twitter or YouTube without the tacit approval of carriers.
8. The lawyers get involved. The Internet has been an experiment in free speech. That may be coming to an end. Unable to go after the sites themselves, lawyers go after the hosters and registrars. That’s how Swiss banking group Julius Baer took whistleblower Wikileaks off the air. And once there’s precedent, others are sure to follow. The recording industry is already wondering if it can go after carriers for enabling copyright infringement. This is the irony of Net Neutrality: When telcos start treating different bytes differently, they aren’t “common carriers” and may be liable for what they transmit, including illegal content. So they’ll comply.
9. Walled gardens: Many countries already restrict how the Internet is used. China’s firewall — which includes 30,000 people tasked with finding improper users — is a good example. But the Internet is a tool for social change and revolution that could threaten any government. Imagine, for example, a U.S. Congress that outlaws online pornography and blocks known adult sites (which accounted for 18.8 percent of all web visits in 2004, according to Hitwise, although the U.S. government says that figure is actually a mere 1 percent.) Instead of a global Internet, we’d have a return to localized standards of decency imposed by legislators. It’d be like “Dirty Dancing” all over again.
10. Humans take themselves out: As Discover Magazine pointed out years ago, we’ve got plenty of ways to do ourselves in, from nukes to plagues to sucking ourselves into a black hole of our own making. And what’s an Internet without users?

The Internet has already morphed from its initial aspirations of open academia to a commercial platform controlled by corporations and carriers. In many ways, the time between the start of ARPAnet in 1969 and the end of Netscape this past February is just a brief period in history that the Facebook generation won’t miss.

10 招干掉互联网

－直以来，我们都想当然地认为互联网是自由交流的全球平台：信息自由传播，随意发布新的应用程序，每个人都可以发出自己的声音，而且未来的互联网将与现在截然不同。不过，假如互联网消失了呢？

－－怎么可能？！

－－以下列出 10 种可能性，我们心里很明白，如果它们中任意一个成为现实，足以灭掉互联网。

1. DNS （域名解析服务）被颠覆了。互联网的正常运行全靠 DNS。如果有人破坏——或更糟，颠覆了－－网站访问的基石，人们将会不再相信什么 URL 网址，网络欺诈也将变得如同儿戏，因为：占有 DNS 者，占有互联网。


2. 僵尸网络攻击！无数“肉鸡”电脑已被控制，躲在阴暗潮湿不通风角落的黑客只需一声令下，即可发起攻击。MessageLabs 的 Matt Sergeant 估计僵尸网络病毒控制了 5-10 百万的计算机（但有人认为这个数字没这么大）。现在，这些僵尸机器人让我们的收件箱充满了垃圾邮件，而过去，它们只是用来发泄一下不满情绪或者勒索恐吓一下网站。但最终，这将演变为一场军备竞争，只不过其中一方不受任何规则的约束。


3. 物理设备发生大规模故障。如果弄断几根地中海的电缆，可以让成千上万的用户无网可上。如果有人蓄意搞破坏，后果将不堪设想啊！


4. 肢解为 N 个独立网络。从古老的 Usenet 至今，人们一直进行着“物以类聚”的分组行为。Nicholas Carr 在《The Big Switch》一书中指出：一项研究声称一个社区，不管保守或自由，其 90% 的链接都没能离开该社区。甚至，对某些参考链接工具进行配置以后，可以使访问者一直停留在具有相同观点的网站上。这样结果会怎样？一个个独立的信息孤岛，人们分享的都是相同的观点和思想，并渐渐确信他们拥有唯一的、正确的答案。互联网的祖先们早有预见：这将是全球化社区之梦的破灭。


5. 出现一个真正意义上的病毒，攻破路由器。互联网的自我修复机制靠的是边界网关协议，即 BGP。但是，如果有人入侵路由器内部，会发生什么事情呢？ 2006 年，在 NANOG 的一份演示文档中，Cisco 对这些不堪一击的弱点进行了探讨，断定“最具毁灭性的破坏来自对托管路由器进行有意的误配置。”如果 BGP 崩溃，不仅无法阻止互联网流量过载，还将失去登录路由器并将其修复的能力。

6. 更新导致崩溃。今日的大多数软件都需不断更新，并保持为最新版本。但有时，自动更新流程自身会引发故障。2007 年 8 月 16 日，Skype 无法使用。公司声称是 Windows 大量的自动更新造成了该故障。总有一天，更新包作为软件重要的一环，如网络堆栈，将造成故障，并且由于无法对其自身进行更新升级，从而导致大量用户无法操作或使用并需要人工介入修复。

7. 网络不再免费。如果运营商开始收取网站访问费用，正如有线电视公司收取相应的费用一样，你将很快发现每月账单上多了一项 “Google 月租”。很多提供 Facebook 和 YouTube 移动服务服务的公司，已经开始收取这种费用。这种死亡方式可能是最不易发现的，因为它预示着创新的末日——没有运营商的同意或默认，无人能推出下一个 Skype、Twitter 或 YouTube。

8. 律师介入。互联网作为自由言论的实验场将走到尽头。由于无法起诉网站本身，律师将追逐着网站拥有者和注册用户。正是通过这种手段，为瑞士的银行组织工作的律师 Julius Baer 对机密曝光网站 Wikileaks 进行起诉。一旦有了先例，将有更多人使用这种手段。唱片公司已经开始考虑起诉运营商违反版权。这是对网络中立性的讽刺：当电信商开始以不同的方式对待不同的字节，他们就不再是“公共运营商”，并且对传输数数据，包括非法内容，负法律责任。所以，他们将最终选择顺从。

9. 封闭城门。很多国家已经开始限制互联网的访问。中国的 GFW－－雇佣 3 万人查找他们认为不正当的网站－－是一个绝好的例子。我们拥有的不再是全球互通的互联网，而是当地立法者制定的道德标准。

10. 人类将自己干掉。Discover Magazine 在多年前已指出，我们拥有大量工具可以进行自我毁灭，从核武、炭疽病毒到人类自制的黑洞。没有人类的互联网是什么样？

互联网的最初灵感来自对学术自由和开放的向往，而现在它已成为由企业和运营商控制的商业平台。从 1969 年 ARPAnet 诞生到去年二月 Netscape 关闭，对于 Facebook 一代，这段历史也许只是昙花一现罢了。