Steven Johnson just wrote an interesting article in the NYTimes that is quite relevant to my latest commentary on Vannevar Bush, the Memex and trails.

Offloading Your Memories
By STEVEN JOHNSON

Published: December 14, 2003

Can your computer be enlisted to help you remember all the details of your life? In a way, it’s already starting to do that. Your e-mail files contain a good portion of your personal communication, and your calendar software has a record of every dentist appointment and staff meeting you’ve had in the last few years. But while it’s easy to track down an address that your colleague e-mailed you six months ago, it’s a bit more challenging to reconstruct a joke your friend told you during a phone call sometime in the late 90’s. So why not take matters a step further and record everything? Now that most of our information streams are built out of zeros and ones, it’s vastly easier to capture all those bits for posterity: every phone call, every passing conversation, every book you read or face you see — the totality of information that flows through a human life.

In the past year, a handful of separate research projects have surfaced, all sharing the same goal: offloading our memories to machines. Sunil Vemuri, a graduate student at the M.I.T. Media Lab, has spent the last two years wedded to a voice-activated microphone that makes a digital recording of every conversation he has and then transmits it to a computer, where it is cataloged and permanently stored.

Advertisement

Similarly, Gordon Bell, a researcher for Microsoft, has created a unified database — he calls it MyLifeBits — that contains the whole package: e-mail, Web pages, the audio from his phone calls, the text of faxes he receives, home movies and so on. All the data in Bell’s system are integrated into a single collection that would allow you — if MyLifeBits becomes commercially available — to sort through life in a variety of ways. You could review everything chronologically, of course, but you could also easily pull together every phone call, fax, photo and e-mail message involving your Aunt Marge. Or you could build more complicated queries, like looking back over 20 years and determining which friends you communicated with most frequently year by year.

In effect, what Bell and others are creating is TiVo for real life. Whatever flows through your perceptual systems can be rewound and queued up for viewing at a later date. Arguing with the barista over whether you specified soy milk in your latte? Trying to determine who actually came up with that brilliant money-saving idea in the staff meeting six months ago? No problem — just go to the tape!

A skeptic might object that 99 percent of the information captured by these personal archives is useless, filed away on the digital equivalent of a cobwebbed library stack. And the skeptic would be right: think of all those spam messages and telemarketing calls you would record for future historians to be annoyed at all over again. But such an extreme ratio of noise to signal is a problem only if storing the data has real costs associated with it and if your ability to find what you’re looking for decreases as more junk is added to the database.

The cost side of the equation is the easy part. Bell estimates that if you were to capture a relatively healthy daily diet of information — 100 Web pages, 8 hours of audio, 100 e-mail messages, one-tenth of a book, 10 photos, 5 scanned pages — it would take you five years to fill up an 80-gigabyte drive that now sells for approximately $100. By the time you maxed out that drive, you would be able to buy a drive with more than 10 times as much capacity for the same price, giving you 60 years of storage, competitively priced at one penny a day.

Of course, if you can’t find what you’re looking for, the dream of total recall becomes little more than fantasy. Are our current search tools up to the task of scouring 60 years’ worth of spoken conversations for every reference to French philosophy or Yankees second basemen? In a word, no. But five years ago we were all complaining that we could never find what we were looking for on the Web. Now we have Google. There’s no reason to suspect that a comparable needle-searching revolution won’t happen to our own private haystacks.

If these personal archives do become commonplace, the outcome, after decades of dark prophecies about the end of privacy, will be a curious one: you will turn out to be the one recording your every move, not the National Security Agency or Equifax or John Ashcroft — a surveillance society of one. We have met Big Brother, and he is us.