Do We Want an Erasable Internet?

The Wall street journal

 HIGH DEFINITION

By FARHAD MANJOO

Dec. 22, 2013 4:46 p.m. ET

Snapchat's headquarters in Los Angeles. The service lets users send messages and photos that disappear in a few seconds. Bloomberg News

This is going to sound silly, but I think Snapchat was the most important technology of 2013.

It sounds silly because Snapchat is just an app. What's more, it's an app used primarily by teens and college students, and wasn't I telling you just a few weeks ago that young people aren't good predictors of tech success?

Then there's the question of whether Snapchat is useful at all. Snapchat sends so-called ephemeral messages, photos and captions that disappear a few seconds after the recipient opens them. Self-destructing photos sound like a recipe for mischief. When people first hear about Snapchat, they likely picture acts by a certain disgraced former congressman.

These are all good points. It's true that we don't know much about how and why people are using Snapchat, and I'm not sure its popularity with teens will translate into broader, long-term success. Snapchat, like all new services, is still more likely to fail than prosper.

Yet, even if it fails, Snapchat will have been one of the most fascinating services to hit the Internet in years. To me, the app's exploding popularity suggests that society is yearning for a new way to think about data. Snapchat is one of the first mainstream services to show us that our photos and texts don't need to stick around forever: that erasing all the digital effluvia generated by our phones and computers can be just as popular a concept as saving it.

If the Snapchat model takes off—if other sites and services began to promote the idea of erasability as a competitive feature—the Internet would look very different from the Internet of today. It would be a more private network, one without the constant worry of every ill-considered picture or thought being held up for ridicule by the whole world, forever. But it also might be a less useful Internet, a network on which you couldn't look up an old photo every time you felt nostalgic, or where computers wouldn't always feed you suggestions based on your history, since your history wouldn't be complete.

Do we want to live on that Erasable Internet, the Snapchat Internet, instead of the Internet built by Facebook and Google ?

That question is exactly why Snapchat is so important, because before Snapchat, the Erasable Internet wasn't an option. The Forever Internet seemed the only way. Now, with users, investors, and engineers rushing to ephemeral-data apps created in Snapchat's image, forever-ness isn't assumed.

Now, I understand Snapchat isn't always ephemeral. The app can be hacked to save a person's photos before they expire, and it's possible for law-enforcement authorities to compel the company to release photos stored in its servers. But, crucially, Snapchat automatically deletes photos from its servers after a user has opened them. This is unusual in the tech industry.

In Silicon Valley, data is inviolable; it is close to a religious belief that you can never have too much data, and deleting is often an afterthought. Note that for the first two years of its life, Google's Gmail didn't even have a delete button; the company believed everyone would want to store all email forever. And it wasn't until 2012 that Facebook figured out how to erase photos from its servers after a user pressed delete.

Meanwhile, the entire premise of the field of "big data" is that the more data you have, the more useful your product becomes, because the analysis of lots of user data can itself lead to new insights and features.

It's not all hype, either. There is a good chance you love some of the many tech products that could only have come about because tech companies saved and analyzed your data.

Among these are Google's spell checker, translation service, traffic maps and search suggestions; Amazon.com's media and product recommendations; and Facebook's personalized News Feed and friend suggestions.

Plus, there's all the data that allows for the targeted advertising that pays for free sites. Without saving and analyzing our information, it's hard to see how ad-based Internet giants could exist.

Yet the costs of the Forever Internet sometime seem just as towering as the benefits. Caches of data about you residing out on servers and devices you don't control are a honey pot for the National Security Agency, hackers and your aggrieved ex.

They can create trouble with potential employers, with college admissions officials and with strangers who make snap-judgments about you based on nothing more than an ill-considered, long-forgotten, out-of-context tweet.

In real life, we're used to shifting our personalities from moment to moment and place to place, depending on who we are with and how we feel. Online, because everything is saved and searchable, you've got to constantly police yourself, to create a single, stultifying profile that restricts spontaneous self-expression. Is this really the way we want to live?

Personally, I'm OK with these costs. I love that I can go back years in my Gmail archive and read the emails my wife and I sent to one another when we first met. If we'd courted on Snapchat, that record would have been lost. I also look forward to future nostalgia—to logging into Facebook in 2023 and looking back at family photos from Thanksgiving 2013. That's why, despite finding it fascinating, I rarely use Snapchat. Deletion scares me.

But I can see how for others, Snapchat provides a sense of liberation from the constraints of a permanent record. Of course, we might not have to pick. Big Snapchat-like growth could mean that we'll have a Forever Internet and an Erasable Internet living side by side. Some users mainly will choose apps and services that save our data by default, while others will choose instant deletion. A lot will choose both, depending.

And that is probably how it should be. No one said the Internet had to be forever. This year, Snapchat proved it.

Write to Farhad Manjoo at farhad.manjoo@wsj.com