The Microsoft researchers obsessively record e-mails, photos, videos, phone calls, health records, financial transactions, Web site visits, and everything else they can in an attempt to electronically compensate for the fallibility of their own biological memories. Before you recoil at the prospect of letting your own life become this digitally augmented, though, consider that it will be whether you want it or not.
"Total Recall," which goes on sale Thursday, is a reasonable and general introduction to the idea that you might want to digitally record much of your life. Chiefly, the book exists to encourage people to take the plunge, but it's also got practical advice about doing so, some warnings about unpleasant possible side effects, and even 10 related start-ups Bell says he'd invest in.
The book is a useful, accessible work from people who've already examined the issue in technical detail. I believe the pair overpromise on the near-term benefits of e-memory and understate some of its difficulties, but they don't try to pretend it'll be universally beneficial.
Bell and Gemmell have real chops in the area. Bell, who started his recording effort in 1998, designed seminal computers for Digital Equipment Corp., established a supercomputing prize in his own name, co-founded the Computer History Museum, and joined Microsoft Research in 1995. Gemmell works on next-generation search and personal information storage, and with Bell and coworker Roger Leuder helped create MyLifeBits software to try to give people a handle on their electronic records. They're the kind of folks who can be found with a Microsoft SenseCam slung around their necks to capture images, sound, and other data as they live their lives.
"We regard it as a memory supplement or surrogate. We're trying to offload biomemories," Bell said. Though plenty of people like to share their lives digitally, the authors see their effort as very private.
So why should the average person listen to them? Because they're right about one thing: all this information likely will be recorded one way or the other as sensors proliferate and digital storage gets ever cheaper and more capacious. It's best to have your own copy--and it's likely it will in fact prove useful.
The authors have high hopes: "higher productivity, more vitality and longer life spans, deeper and wider knowledge of our world and ways to accomplish things in it."
Still, it's a hard sell.
I doubt most people will embrace what Gemmell and Bell call e-memory with much enthusiasm. Most folks see filing as a necessary evil at best, and e-memory dramatically increases the amount necessary. What file system should you pick for your photos, GPS location logs, energy consumption measurements, and blood pressure records? What backup strategy? What privacy settings? What computer hardware, software, and online services?
I've been stewing over some of these questions for a while--forced the issue--but my conversation with the authors still was thought-provoking.
History by some definitions began when people started recording things, and the Total Recall era of Bell and Gemmell has the potential to bring as radical a shift in human behavior as the arrival of books.
If this shift is to come to pass, a key factor will be extracting something useful out of all the data you collect. "The root of the problem is to get the computer to first record it and store it and somehow be able to act on it," Bell said.
The obvious mechanism is search, of course, but Gemmell and Bell also advocate wall-hanging screen savers that constantly replay parts of your life in the background. Another idea of theirs got me thinking about things more holistically.
Specifically, there's the idea of correlation: collecting data in one area could help you find or use data in another. For example, a log of your position kept through a GPS system in your phone could help you retrieve a message you knew you sent while on a particular trip. Or in a more sci-fi scenario, your logs of movies and books read, cross-referenced with your heart rate and blood pressure measurements, could help you decide what sorts of movie to watch.
Then there's the idea of prediction. A sufficiently smart system processing your collected data could not just remind you of simple things such as birthdays, but also anticipate more complicated advice--time to get a check-up or spend more time with friends, perhaps.
Here's where I see storm clouds on the e-memory horizon, though.
First, if it's easy for me to store everything and perform sophisticated correlation analysis, it's easy for scheming governments or profit-hungry corporations to do the same. With the arrival of online behavior monitoring and omnipresent surveillance cameras, it's harder to find a corner of the world to claim as your own.
One answer from Gemmell and Bell: "Big Brother, meet Little Brother." Your personal records can give you an alibi proving your innocence in some matter. But that's only OK as long as you happen to live someplace where the rule of law has some teeth and the authorities aren't abusing their power.
And if everybody walks around with recording devices running around the clock, the government may be the least of your concerns for privacy. Politicians today are still adapting to the YouTube era when every moment is on the record, but we're headed toward an era when it applies to us all.
Perhaps that will deter crime on the assumption that people behave better when they know they're being watched. But perhaps also some white lies and imperfect memories help us all get along. Can we really handle the unvarnished truth about our friends and coworkers, much less ourselves?
The authors believe it's time to adapt to a world in which truth and honesty are harder to overlook, and that social protocols will adapt perforce. For those moments when you want some Nixonian plausible deniability, perhaps protocols will emerge to hold a conversation that's logged only in people's neurons.
Bell and Gemmell expect new conventions for asking others' permission to record. But if data capture is as easy, ubiquitous, and unobtrusive as "Total Recall" suggests, it looks like it would be wiser to assume everything is being recorded already.
From a technological standpoint, I worry about data being fragmented. When Bell started his work, he was a fan of centralized data stored on your own hard drives. But now we have the cloud, too--online sites not just for backup but increasingly the central repository of your data held in some service.
Companies such as Yahoo, Facebook, and Google have plenty of incentives to set up services that house precious data such as photos, communications with friends, and documents--but they may not want to share that openly with you or some company that offers e-memory services on your behalf. And even if they do, there are practical barriers aplenty.
Perfectly preserved ones and zeros can nevertheless decay with age, locked in obsolete formats. The researchers advocate periodic backup into "golden" formats such as PDF or JPEG whose widespread use will protect against obsolescence. Ironically, Bell himself illustrates the importance of the problem: the Microsoft Money software he touts for tracking his own financial transactions is being phased out.
Data ownership is another complicated matter. As Bell and Gemmell observe, parents today already electronically monitor their children closely and exhaustively document their lives. I wonder if or when the resulting information should become the child's property.
And in the most sci-fi section of the book, the authors ponder digital immortality, the possibility of conversing with a digital avatar representing the deceased. "How I wish I had even a tenth of my grandfather's life," Bell says--but how much of one's life should belong to one's heirs?
The authors and I aren't the only ones to have pondered the ideas of e-memory and extrapolated into the future. They cite various science fiction books that have bearing on their ideas, for example.
But it was politician and presidential progenitor Vannevar Bush whose "memex" provided a template for the e-memory work in a 1945 essay, in The Atlantic Monthly, "As We May Think." His idea: "A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory."
Even though I'm not convinced everybody will eagerly embrace e-memory, I am convinced that it figures centrally in our lives. I have seen the power of Gmail's archive and my own digital photo archive. And as it gets easier to do record everything, we will find a way to make it useful.
Many years ago, I made T-shirt presents for my family that used the highest-resolution scans I could make of some graphics. At the time, I was astounded at how much space each image took up--20MB or more. Over years of hard drive tidying efforts, I considered deleting the files many times, but they had some sentimental value and I put off the decision each time.
Now I'm glad I hung onto them. The T-shirts have all worn out, but I still have the artwork--and 20MB is a piddling amount of storage space. For comparison, each of the thousands of photos I've taken with my present camera today is closer to 30MB.
At the same time, I find myself e-mailing myself notes, recording voice memos, keeping my GPS logs of excursions, and taking snapshots of the signs and maps at trailheads. Without even trying hard, I'm assembling my personal life's digital equivalent of everything from Library of Alexandria to the trash heaps archaeologists delight in excavating.
I'm a ways out there on the nerd spectrum, too. But with ordinary folks carrying camera-enabled smartphones, e-mailing their way through work, and socializing through status updates, brace yourself for e-memory to arrive in your life.