Total Recall: How the E-Memory Revolution Will Change Everything [Roughcut]

On his website, The Technium, Kevin Kelly (of Wired and Whole Earth fame)
writes about "What Technology Wants."
Here's what IT wants - "Everything, Everywhere, All the Time."

In IT's strive toward omniscience, it's clear that the next key piece is
Total Recall of all personal, individual memories.
Gordon Bell and Jim Gemmell lay out precisely how and why that will happen.

I've been in the memory business for over 40 years:
first as a student of neurobiology at MIT, then as an AI researcher at Stanford,
and finally as a physician. (Search "Bob Blum" for my essays on
machine consciousness and other Big Questions.)

I had heard of Gordon Bell for decades, but had never met him
until recently when I heard Gordon and Jim present this work
at the (Xerox) PARC Forum. (That video is now on the PARC Forum archive).
That prompted me to buy the book.

Despite being age 75, Gordon is a lively, energetic spirit
who readily deflected my public query/position ,
"don't neuroscientists consider forgetting to be crucial
as a means of increasing memory relevance?"
(My concern then and still is on maintaining high signal to noise ratio -
quieting the mind to achieve the zen of pure signal.)

Young Jim Gemmell is also bright and engaging.
Although I'm guessing that Jim contributed half of the leg work,
the book is presented as a first person account of Gordon's 75 year life.
The work is a delightful combination of the future of personal data capture
as well as a recounting of their experiences with MyLifeBits, a system implementation.
That work was presented in Scientific American in March 2007 online - qv.)

Gordon has ridden the ascendancy of IT from prebirth
(the family business was Bell Electric), through his student days at MIT,
through his years as principal architect of the DEC System PDP and VAX computers
(that ruled the IT world for a decade), and finally to his current position
as eminence grise at Microsoft.

This book was a great stroll down memory lane.
I love visiting the Computer History Museum in Mountain View, California.
I had forgotten that this wonderful museum was made possible by Gordon Bell.
He describes his efforts at collecting the oral histories that went into the museum,
and how much easier it would have been with Total Recall.
Imagine having every conversation of Charles Babbage, Thomas Watson,
John Von Neumann, and Alan Turing.)

Now, here's their main point - as you live your life, COLLECT EVERYTHING:
every visual field, every conversation, every location, every accessible bodily function -
not merely every email and web page. They describe a panoply of present and future sensors
that will perhaps make it effortless: micro video cams, physiologic monitors, gps,etc.

Ok, I've got 40 physical file drawers (I'm a fellow packrat),
but that proposition raises hackles even with me as it must with every reader.
Really? Isn't that endlessly time-consuming and distracting?
(They say "no" - I say "maybe.") "Let the system do all the work," they say,
silently collecting all you see, hear, do, and are.
Then, at least, it's all potentially available, if and when you want to retrieve it.
(Record your every moment from birth, then these e-memories will be available to
your great grandchildren, your biographer, and your therapist.)

Their goal is identical to Google's:
index and make readily available every single experience.
(They describe MIT researcher Deb Roy's 24 by 7 real-time video capture
of every instant in the life of his new born baby for three years:
bizarre perhaps but priceless for students of language.)

Health, work, education, travel, personal life are all grist for Total Recall.
The health care piece especially resonated with me, since I had developed
a system at Stanford that discovered medical knowledge from stored personal data.
They correctly describe the huge importance that will accrue to each of us
as we gather data on our day to day health and habits.
Refrigerator to Gordon - "What? No vegetables? Just Ice Cream?!"
Computer to Bob - "You've sat too long. Time to get up and bike."

Gordon describes his several episodes of severe coronary disease including cardiac arrest.
Heart disease, like most diseases, is brought on one day at a time over a lifetime.
The remedy is self observation and performance monitoring. All our habits need to be recorded.
We are our own best doctors. Not only is the data personally essential,
but it's a great epidemiologic resource. When a close friend died of cancer,
it was tragic that the causative factors had not been captured.
(We are probably swimming in a sea of unidentified carcinogens.)

One thinks of the many potential hazards of collecting data this intimate:
identity theft, denial of insurance coverage, blackmail, and cognitive clutter.
They discuss all of the many pitfalls and present some novel solutions,
eg the self-destructing Swiss Data Bank. They do NOT advocate making your data public.

I had expected (and got) a thorough discussion of their experiences with
personal data capture. But, what a pleasant, upside surprise was the engaging story of
their use by Gordon as he built the Computer Museum, dealt with his heart attacks,
dealt with a poignant incident (the mysterious disappearance/death of his boss,
superstar computer scientist Jim Gray), his deals with dozens of start-ups and
entrepreneurial projects, or his on-going effort at age 75 to build an immortal version
of himself that may be able to grow and learn after his body has been recycled!
(Want to get rich? The book is filled with great ideas for entrepreneurs - untapped gold!)

Yes. 99% of life is banal, but, as they say,
"the palest ink is better than the best memory,"so RECORD IT ALL.
I award the book a mere 4 stars, only because it is not "War and Peace."
But, if you work with computers and are interested in the future, this book is for you.

"Everything, Everywhere, All the time" is inevitable, at least for some of us.

There are many branches of the tree of accelerating technology. Bell and Gemmell's book focuses on the current and future effects of information retention and, more importantly, information retrieval. They range from practical advice -- make scanned bills PDF searchable -- to future scenarios where so much information has been retained that we can "talk" with our long dead great grandparents (via artificial bots made smart by massive knowledge of the subject). The book is somewhat happhazardly organized but I gave it a five star anyway because of the insights and the fact that Bell went through the process of recording his life, using prototype software. Someone who has "done it" speaks with more authority than an armchair quarterback. After reading the book, I thought ... of course -- explained in the context of massive increases in storage, networking and computing power, it all makes sense. Bell and Gemmell are relatively conservative in their predictions. They touch on some of the security issues but do not dwell on them. It is probably just as well, since the trend to increasing storage of events is inevitable and security will just have to be worked out. How many business meetings have I attended where ten people have been introduced in about ten seconds? A universal recall device would come in pretty handy.

Bill Yarberry, Houston, Texas

Remember when everyone at concerts held up their lighters? Now it's cell phones -- taking pictures. This is the basic problem with "Total Recall": We are far better at capturing moments than we are at preserving them. Heck, we're better at preserving them than we are at organizing them so that later we can find what we want. This is truer now that it's all-digital than it was when it was all-shoebox. Clearly, our intentions are better than our methods.

Like you, I'm being continually supplied with free software to help me find all the photos on my machine. But when the pile is big enough (and mine, like yours, certainly is), no amount of brute force searching is more useful than simple organizing. When will software be able to do that for me?

Therefore, the real proposition of "Total Recall" is brute force combined with AI. Gordon Bell says we have the technology now to save everything, so the real challenge is developing truly helpful AI. His ideas are completely dependent on this happening in the near term. Others, myself included, are far more skeptical. And without useful AI, "Total Recall" is just another impractical utopian ideal.

I'm sure Gordon Bell is a better engineer than I am, and I know he's a better entrepreneur. But as to his ability to see the future, I have my doubts. It's great to imagine an ideal, however impractical -- unless it gets in the way of practical, incremental improvements. By diverting effort and resources to his fantasy, "Total Recall" may be more roadblock than highway to tomorrow.