The odd thing about the reaction to this book is that almost everybody is most interested in the fact that I included a YouTube URL in the book. Such a tiny thing, but in the moment I thought: okay, I’ll be really adventurous. I didn’t know it would be the thing people really paid attention to. Maybe it was a mistake. I think it was a bit of nostalgic postmodernism. In the way that people paint a photorealistic painting of a street sign. “Look at this! Look at this sequence of letters! Think about the fact that it takes you to a human voice singing in a Paris hotel room. Look at that, and be happy.” So, at that point, I was just sitting there, thinking: “Well, I really want people to listen to Stephen Fearing. I really would like that. If my book could do one thing, it would be that people would actually be guided to listen to Stephen Fearing.” And of course the worst possible way to tell them to go, I guess, is to give them a dead YouTube link, because they’re going to make a typo. The best way is to type “Paris hotel Fearing” or something. So I kind of blew it.
Which is a reminder of what a lousy technology, from the user’s point of view, the URL is—though oddly, it was created in order to help users: a URL, with its readable domain name, like www.google.com, merely points to what your computer and your network think of as the real target, the IP address, like http://126.96.36.199/. For human beings, the former is easier to remember than the latter. (The slashes come from UNIX file-path syntax: Clearly Tim Berners-Lee and the other architects of the Web were thinking of pages within a domain as being like files within a directory.) Those alphabetic URLs worked pretty well, for those used to using UNIX anyway, until websites started proliferating, which led inevitably to URLs getting longer and longer—and then, equally inevitably, the rise of URL-shortening services.
But these URL shorteners weren’t intended to make locations more readable—shortened URLs are typically gibberish—but to keep them from taking over emails and other messages in which they were included. And anyway, within the last decade more and more people have been giving up on even a basic understanding of URLs, trusting Google results instead. Google’s highlighted blue links—in your own language!—offer a layer of additional, simplified comprehensibility above the layer of alphabetic comprehensibility that Berners-Lee created to cover the basically incomprehensible layer of the IP address.
This doesn’t always work: thus the ridiculous scene a few years ago when ReadWriteWeb ran a story called “Facebook Wants To Be Your One True Login” which temporarily became the first result if you Googled “facebook login.” The result: people clicking on that link and becoming outraged that Facebook was not allowing them to log in. This little event gives us a lot to reflect on: Should we think first about how dimwitted people must be to fail to notice that a ReadWriteWeb article was not Facebook? Or should we consider that Facebook had developed such a habit of changing its appearance that users weren’t that surprised to find a dramatically new design?
I dunno. But what interests me about the story is this: There are many users who are so resistant to or uncomfortable with URLs that typing “facebook login” into a Google search box is more comfortable than typing “Facebook.com” into a browser’s address bar.
So we should probably be thinking about what the next steps are in the evolution of the IP address. The creation of the address itself was the first stage; the creation of the URL the second; the preferential use of the search engine the third. I suppose the immediate future largely involves using services like Apple’s Siri to search by voice, but for ambitious technologists, that’s just a stopgap. The very idea of the IP address assumes an intelligent agent who has some idea of what she wants and some way to find it—and that’s an assumption tech companies don’t want to make any more. As Phil Libin of Evernote commented, “By the time you search, something’s already failed.” The ultimate layer of abstraction away from the IP address is a world when intelligent agency, the human questioner, has disappeared altogether. We won’t need URLs then.