Artificial Forgetting

Here’s one way to help create the “right to be forgotten”: Have your online utterances vanish after a period of time.

Clive Thompson
May 15, 2014 · 5 min read

How can you make personal info vanish online?

This week, the European Court of Justice ruled that Google has to honor citizens’ “right to be forgotten”. In situations where the request is reasonable, a search engine must remove links to outdated personal info online. This decision acknowledged what we all know, which is that Google isn’t so much a search engine as a mammoth reputation-management system. The case was brought by a Spanish man who hated the fact that Google linked to 16-year-old newspaper stories about his old debts. The average person probably doesn’t have news articles to worry about, but there are tons of other things to fret over — particularly social-media postings that could be ripped out of context years later.

But allowing people to juke search-engine results opens up several cans of worms, legal and cultural. To my mind, the biggest problem is that this new right will likely be used — and abused — mostly by the wealthy and powerful. Exercising legal rights requires money and legal firepower; they have it, and the rest of us mostly don’t.

But what if we dealt with the “forgetting” problem somewhere else? What if, instead of focusing on the search-engine side — the point of retrieval — we looked at the point of publishing?

What if more publishing platforms were designed to delete personal info after it gets old?

This is “artificial forgetting,” and while Snapchat uses it for person-to-person communications, it’s also been used on the web. Back in 2009, one of my favorite sites for sharing files was a startup called Drop.io. It was quick and easy to use — you uploaded a file, picked the folks with whom to share it, and boom: done. You didn’t even need to register an account.

But here’s the thing: Each time you uploaded a file, Drop.io asked you when you wanted that file automatically deleted. The default was a year, but you could pick a much shorter time — a month, a single day — or a ruleset like “after two people access it”. Or you could pick “never”, ensuring the file would be online forever.

The upload screen looked like this:

“Expire drop after”: At the old Drop.io service, you uploaded a file — then picked a date in the future when it would vanish.

It was a system for sharing personal info, but the default was deletion. It recognized that the stuff you’re putting online may have a best-before date, a point after which you don’t want it in the aether. This design had a powerful effect on behavior. In 2009, Drop.io founder Sam Lessin told me that two-thirds of the tens of million of files that people had uploaded to the service in the previous year and a half were gone. He called them “little wormholes that pop into and out of existence for a specific purpose.”

This is precisely the opposite of how every other publishing service we use today works. Facebook, Twitter, microblogging sites, photo sites — they all encourage us to put things online and leave them there until the sun explodes. For most of them, particularly social networks, this design is propelled by marketing logic; the more data they can amass on us, the better they can target ads.

But what if more publishing tools behaved like Drop.io? Our world of online personal information would be much more transient. There’d be a lot less personal stuff for Google to crawl in the first place.

There’s plenty of evidence that people would enjoy this sort of transience, particularly for the sort of funny, witty, racy interpersonal communications that are likely to land us in hot water when ripped out of context years later by a search engine. Several teenagers have told me how they hack transience into “permanent” social networks — such as by posting a photo on Instagram, tagging some friends so they show up and comment, having a quick conversation in the discussion fields, then deleting the photo a few minutes later. I’ve met some whose Instagram accounts appear to consist of a single photo, but in reality they’ve posted hundreds of times; it’s just they’ve deleted every single interaction, because they don’t want to leave the trail. Similarly, danah boyd has discovered youths who use the same post-and-delete technique on Facebook, or who “resign” their accounts when they’re not actively using them, so they’re invisible to passersby.

Obviously, artificial forgetting is only one technique to help people manage personal info online. And it’s very imperfect. To list just a few of the most obvious limitations: Digital material is hard to control once it’s online. (Behold the power of cut-and-paste! As users of the supposedly transient Snapchat have realized.) We can’t always trust whether a for-profit firm actually deletes stuff when we tell them to. Meanwhile, the NSA probably has a copy of everything posted to the Internet, no matter how “transiently” it’s posted. Plus, we’re often terrible judges of whether today’s personal info will look silly years later, so it’s not even clear we’d be intelligent users of artificial forgetting.

More to the point, the material the EU citizen sued over wasn’t published on a social system, but in a newspaper. And I don’t want newspapers auto-deleting their stories after a period of time — not voluntarily, and certainly not at the behest of a court order. One of the central cultural roles of journalism is to build a record of what’s happened in a society.

So yes, artificial forgetting isn’t a panacea. But it’s an intriguing idea, and one worth pondering more. Maybe we’d enjoy tools that encourage us not just to save things, but to get rid of them.

(Coda: Drop.io was great, but ironically — or appropriately? — it was itself deleted, when the company was phagocytosed by Facebook. This is another of the weird paradoxes of online personal info. We might worry that for-profit online services, with their stores of our personal info, will be haunting us like ghosts into our dotage. But they can be surprisingly fragile. Wake up one day poof, a service you used for months or years — Magnolia, Pownce — can vanish without a trace. Thus the flip side of the folks who worry about “the right to be forgotten” are the ones who fret about “link rot” and the propensity of material online to vanish over time, via corporate entropy. It’ll be interesting to see what happens years hence when, as is eventual, Facebook is shut down.)

(The lovely photo at the top of this piece is by Jason Eppink.)

A Pandaemonium Revolver Collection. Season 2 stars @anildash @alanalevinson @ftrain @hipstercrite @itsthebrandi @jamielaurenkeiles @vijithassar @yungrama @zeynep. Season 1 available on DVD shortly.

Thanks to Kate Lee

    Clive Thompson

    Written by

    Journalist for Wired, NYT mag. Musician. Writing a book about “how programmers think”. (Coders: Ping me about it!) www.clivethompson.net, email is clive@ there.

    The Message

    A Pandaemonium Revolver Collection. Season 2 stars @anildash @alanalevinson @ftrain @hipstercrite @itsthebrandi @jamielaurenkeiles @vijithassar @yungrama @zeynep. Season 1 available on DVD shortly.

    Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
    Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
    Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade