Microsoft Word Kurzweil, Ray The Singularity Is Near doc

The Longevity of Information 
"The horror of that moment," the King went on, "I shall never, never forget it!" "You will, though," the 
Queen said, "if you don't make a memorandum of it." 
—L
EWIS 
C
ARROLL
,
T
HROUGH THE 
L
OOKING
-G
LASS
The only things you can be sure of, so the saying goes, are death and taxes—but don't be too sure about death. 
—J
OSEPH 
S
TROUT
,
NEUROSCIENTIST
I do not know sire, but whatever they will turn out to be I am sure you will tax them. 
—M
ICHAEL 
F
ARADAY
,
RESPONDING TO A QUESTION FROM THE 
B
RITISH 
E
XCHEQUER AS TO 
WHAT PRACTICAL USE COULD BE MADE OF HIS DEMONSTRATION OF ELECTROMAGNETISM
Do not go gentle into that good night, ... 
Rage, rage against the dying of the light. 
—D
YLAN 
T
HOMAS
The opportunity to translate our lives, our history, our thoughts, and our skills into information raises the issue of how 
long information lasts. I've always revered knowledge and gathered information of all kinds as a child, an inclination I 
shared with my father. 
By way of background, my father was one of those people who liked to store all the images and sounds that 
documented his life. Upon his untimely death at the age of fifty-eight in 1970, I inherited his archives, which I treasure 
to this day. I have my father's 1938 doctoral dissertation from the University of Vienna, which contains his unique 
insights into the contributions of Brahms to our musical vocabulary. There are albums of neatly arranged newspaper 
clippings of his acclaimed musical concerts as a teenager in the hills of Austria. There are urgent letters to and from 
the American music patron who sponsored his flight from Hitler, just before Kristallnacht and related historical 
developments in Europe in the late 1930s made such escape impossible. These items are among dozens of aging boxes 
containing a myriad of remembrances, including photographs, musical recordings on vinyl and magnetic tape, personal 
letters, and even old bills. 
I also inherited his penchant for preserving the records of one's life, so along with my father's boxes I have several 
hundred boxes of my own papers and files. My father's productivity, assisted only by the technology of his manual 
typewriter and carbon paper, cannot compare with my own prolificacy, aided and abetted by computers and high-
speed printers that can reproduce my thoughts in all kinds of permutations. 
Tucked away in my own boxes are also various forms of digital media: punch cards, paper-tape reels, and digital 
magnetic tapes and disks of various sizes and formats. I often wonder just how accessible this information remains. 
Ironically the ease of approaching this information is inversely proportional to the level of advancement of the 
technology used to create it. Most straightforward are the paper documents, which although showing signs of age are 

---

eminently readable. Only slightly more challenging are the vinyl records and analog tape recordings. Although some 
basic equipment is required, it is not difficult to find or use. The punch cards are somewhat more challenging, but it's 
still possible to find punch-card readers, and the formats are uncomplicated. 
By far the most demanding information to retrieve is that contained on the digital disks and tapes. Consider the 
challenges involved. For each medium I have to figure out exactly which disk or tape drive was used, whether an IBM 
1620 circa 1960 or a Data General Nova I circa 1973.Then, once I've assembled the requisite equipment, there are 
layers of software to deal with: the appropriate operating system, disk information drivers, and application programs. 
And, when I run into the inevitable scores of problems inherent in each layer of hardware and software, just whom am 
I going to call for assistance? It's hard enough getting contemporary systems to work, let alone systems for which the 
help desks were disbanded decades ago (if they ever existed). Even at the Computer History Museum most of the 
devices on display stopped functioning many years ago.
41
Assuming I do prevail against all of these obstacles, I have to account for the fact that the actual magnetic data on 
the disks has probably decayed and that the old computers would still generate mostly error messages.
42
But is the 
information gone? The answer is, Not entirely. Even though the magnetic spots may no longer be readable by the 
original equipment, the faded regions could be enhanced by suitably sensitive equipment, via methods that are 
analogous to the image enhancement often applied to the pages of old books when they are scanned. The information 
is still there, although very difficult to get at. With enough devotion and historical research, one might actually retrieve 
it. If we had reason to believe that one of these disks contained secrets of enormous value, we would probably succeed 
in recovering the information. 
But mere nostalgia is unlikely to be sufficient to motivate anyone to undertake this formidable task. I will say that 
because I did largely anticipate this dilemma, I did make paper printouts of most of these old files. But keeping all our 
information on paper is not the answer, as hard-copy archives present their own set of problems. Although I can 
readily read even a century-old paper manuscript if I'm holding it in my hand, finding a desired document from among 
thousands of only modestly organized file folders can be a frustrating and time-consuming task. It can take an entire 
afternoon to locate the right folder, not to mention the risk of straining one's back from moving dozens of heavy file 
boxes. Using microfilm or microfiche may alleviate some of the difficulty, but the matter of locating the right 
document remains. 
I have dreamed of taking these hundreds of thousands of records and scanning them into a massive personal 
database, which would allow me to utilize powerful contemporary search-and-retrieve methods on them. I even have a 
name for this venture—DAISI (Document and Image Storage Invention)—and have been accumulating ideas for it for 
many years. Computer pioneer Gordon Bell (former chief engineer of Digital Equipment Corporation), DARPA 
(Defense Advanced Research Projects Agency), and the Long Now Foundation are also working on systems to address 
this challenge.
43
DAISI will involve the rather daunting task of scanning and patiently cataloging all these documents. But the real 
challenge to my dream of DAISI is surprisingly deep: how can I possibly select appropriate hardware and software 
layers that will give me the assurance that my archives will be viable and accessible decades from now? 
Of course my own archival needs are only a microcosm of the exponentially expanding knowledge base that 
human civilization is accumulating. It is this shared species-wide knowledge base that distinguishes us from other 
animals. Other animals communicate, but they don't accumulate an evolving and growing base of knowledge to pass 
down to the next generation. Since we are writing our precious heritage in what medical informatics expert Bryan 
Bergeron calls "disappearing ink," our civilization's legacy would appear to be at great risk.
44
The danger appears to be 
growing exponentially along with the growth of our knowledge bases. The problem is further exacerbated by the 
accelerating speed with which we adopt new standards in the many layers of hardware and software we employ to 
store information. 
There is another valuable repository of information stored in our brains. Our memories and skills, although they 
may appear to be fleeting, do represent information, coded in vast patterns of neurotransmitter concentrations, 
interneuronal connections, and other relevant neural details. This information is the most precious of all, which is one 

---

reason death is so tragic. As we have discussed, we will ultimately be able to access, permanently archive, as well as 
understand the thousands of trillions of bytes of information we have tucked away in each of our brains. 
Copying our minds to other mediums raises a number of philosophical issues, which I will discuss in the next 
chapter—for example, "Is that really me or rather someone else who just happens to have mastered all my thoughts 
and knowledge?" Regardless of how we resolve these issues, the idea of capturing the information and information 
processes in our brains seems to imply that we (or at least entities that act very much like we do) could "live forever." 
But is that really the implication? 
For eons the longevity of our mental software has been inexorably linked to the survival of our biological 
hardware. Being able to capture and reinstantiate all the details of our information processes would indeed separate 
these two aspects of our mortality. But as we have seen, software itself does not necessarily survive forever, and there 
are formidable obstacles to its enduring very long at all. 
So whether information represents one man's sentimental archive, the accumulating knowledge base of the 
human-machine civilization, or the mind files stored in our brains, what can we conclude about the ultimate longevity 
of software? The answer is simply this: