Microsoft Word Kurzweil, Ray The Singularity Is Near doc

This is not because humans will have become what we think of as machines today, but 
rather machines will have progressed to be like humans and beyond.
Technology will be the metaphorical 
opposable thumb that enables our next step in evolution. Progress (further increases in order) will then be based 
on thinking processes that occur at the speed of light rather than in very slow electrochemical reactions. Each 
stage of evolution builds on the fruits of the last stage, so the rate of progress of an evolutionary process 
increases at least exponentially over time. Over time, the "order" of the information embedded in the 
evolutionary process (the measure of how well the information fits a purpose, which in evolution is survival) 
increases. 
•
An evolutionary process is not a closed system; evolution draws upon the chaos in the larger system in which it 
takes place for its options for diversity. Because evolution also builds on its own increasing order, in an 
evolutionary process order increases exponentially. 
•
A correlate of the above observation is that the "returns" of an evolutionary process (such as the speed, 
efficiency, cost-effectiveness, or overall "power" of a process) also increase at least exponentially over time. We 
see this in Moore's Law, in which each new generation of computer chip (which now appears approximately 
every two years) provides twice as many components per unit cost, each of which operates substantially faster 
(because of the smaller distances required for the electrons to travel within and between them and other factors). 
As I illustrate below, this exponential growth in the power and price-performance of information-based 
technologies is not limited to computers but is true for essentially all information technologies and includes 
human knowledge, measured many different ways. It is also important to note that the term "information 

---

technology" is encompassing an increasingly broad class of phenomena and will ultimately include the full range 
of economic activity and cultural endeavor. 
•
In another positive-feedback loop, the more effective a particular evolutionary process becomes—for example, 
the higher the capacity and cost-effectiveness that computation attains—the greater the amount of resources that 
are deployed toward the further progress of that process. This results in a second level of exponential growth; 
that is, the rate of exponential growth—the exponent—itself grows exponentially. For example, as seen in the 
figure on p. 67, "Moore's Law: The Fifth Paradigm," it took three years to double the price-performance of 
computation at the beginning of the twentieth century and two years in the middle of the century. It is now 
doubling about once per year. Not only is each chip doubling in power each year for the same unit cost, but the 
number of chips being manufactured is also growing exponentially; thus, computer research budgets have grown 
dramatically over the decades. 
•
Biological evolution is one such evolutionary process. Indeed, it is the quintessential evolutionary process. 
Because it took place in a completely open system (as opposed to the artificial constraints in an evolutionary 
algorithm), many levels of the system evolved at the same time. Not only does the information contained in a 
species' genes progress toward greater order, but the overall system implementing the evolutionary process itself 
evolves in this way. For example, the number of chromosomes and the sequence of genes on the chromosomes 
have also evolved over time. As another example, evolution has developed ways to protect genetic information 
from excessive defects (although a small amount of mutation is allowed, since this is a beneficial mechanism for 
ongoing evolutionary improvement). One primary means of achieving this is the repetition of genetic 
information on paired chromosomes. This guarantees that, even if a gene on one chromosome is damaged, its 
corresponding gene is likely to be correct and effective. Even the unpaired male Y chromosome has devised 
means of backing up its information by repeating it on the Y chromosome itself.
12
Only about 2 percent of the 
genome codes for proteins.
13
The rest of the genetic information has evolved elaborate means to control when 
and how the protein-coding genes express themselves (produce proteins) in a process we are only beginning to 
understand. Thus, the process of evolution, such as the allowed rate of mutation, has itself evolved over time. 
•
Technological evolution is another such evolutionary process. Indeed, the emergence of the first technology-
creating species resulted in the new evolutionary process of technology, which makes technological evolution an 
outgrowth of—and a continuation of—biological evolution. 
Homo sapiens
evolved over the course of a few 
hundred thousand years, and early stages of humanoid-created technology (such as the wheel, fire, and stone 
tools) progressed barely faster, requiring tens of thousands of years to evolve and be widely deployed. A half 
millennium ago, the product of a paradigm shift such as the printing press took about a century to be widely 
deployed. Today, the products of major paradigm shifts, such as cell phones and the World Wide Web, are 
widely adopted in only a few years' time. 
•
A specific paradigm (a method or approach to solving a problem; for example, shrinking transistors on an 
integrated circuit as a way to make more powerful computers) generates exponential growth until its potential is 
exhausted. When this happens, a paradigm shift occurs, which enables exponential growth to continue.