Microsoft Word Kurzweil, Ray The Singularity Is Near doc

particles in the universe, with a theoretical maximum 
capacity of about 10
90
cps. In other words a universe-scale computer would be able to compute at 10
90
cps.
98
To arrive 
at those estimates, Lloyd took the observed density of matter—about one hydrogen atom per cubic meter—and from 
this figure computed the total energy in the universe. Dividing this energy figure by the Planck constant, he got about 
10
90
cps. The universe is about 10
17
seconds old, so in round numbers there have been a maximum of about 10
107
calculations in it thus far. With each particle able to store about 10
10
bits in all of its degrees of freedom (including its 
position, trajectory, spin, and so on), the state of the universe represents about 10
90
bits of information at each point in 
time. 
We do not need to contemplate devoting all of the mass and energy of the universe to computation. If we were to 
apply 0.01 percent, that would still leave 99.99 percent of the mass and energy unmodified, but would still result in a 
potential of about 10
86
cps. Based on our current understanding, we can only approximate these orders of magnitude. 
Intelligence at anything close to these levels will be so vast that it will be able to perform these engineering feats with 
enough care so as not to disrupt whatever natural processes it considers important to preserve. 
The Holographic Universe.
Another perspective on the maximum information storage and processing capability of 
the universe comes from a speculative recent theory of the nature of information. According to the "holographic 
universe" theory the universe is actually a two-dimensional array of information written on its surface, so its 
conventional three-dimensional appearance is an illusion.
99
In essence, the universe, according to this theory, is a giant 
hologram. 
The information is written at a very fine scale, governed by the Planck constant. So the maximum amount of 
information in the universe is its surface area divided by the square of the Planck constant, which comes to about 10
120
bits. There does not appear to be enough matter in the universe to encode this much information, so the limits of the 
holographic universe may be higher than what is actually feasible. In any event the order of magnitude of the number 
of orders of magnitudes of these various estimates is in the same range. The number of bits that a universe reorganized 
for useful computation will be able to store is 10 raised to a power somewhere between 80 and 120. 
Again, our engineering, even that of our vastly evolved future selves; will probably fall short of these maximums. 
In chapter 2 I showed how we progressed from 10
–5
to 10
8
cps per thousand dollars during the twentieth century. 
Based on a continuation of the smooth, doubly exponential growth that we saw in the twentieth century, I projected 
that we would achieve about 10
60
cps per thousand dollars by 2100. If we estimate a modest trillion dollars devoted to 
computation, that's a total of about 10
69
cps by the end of this century. This can be achieved with the matter and energy 
in our solar system. 
To get to around 10
90
cps requires expanding through the rest of the universe. Continuing the double-exponential 
growth curve shows that we can saturate the universe with our intelligence well before the end of the twenty-second 
century, provided that we are not limited by the speed of light. Even if the up-to-thirty additional powers of ten 
suggested by the holographic-universe theory are borne out, we still reach saturation by the end of the twenty-second 
century. 

---

Again, if it is at all possible to circumvent the speed-of-light limitation, the vast intelligence we will have with 
solar system-scale intelligence will be able to design and implement the requisite engineering to do so. If I had to place 
a bet, I would put my money on the conjecture that circumventing the speed of light is possible and that we will be 
able to do this within the next couple of hundred years. But that is speculation on my part, as we do not yet understand 
these issues sufficiently to make a more definitive statement. If the speed of light is an immutable barrier, and no 
shortcuts through wormholes exist that can be exploited, it will take billions of years, not hundreds, to saturate the 
universe with our intelligence, and we will be limited to our light cone within the universe. In either event the 
exponential growth of computation will hit a wall during the twenty-second century. (But what a wall!) 
This large difference in timespans—hundreds of years versus billions of years (to saturate the universe with our 
intelligence)—demonstrates why the issue of circumventing the speed of light will become so important. It will 
become a primary preoccupation of the vast intelligence of our civilization in the twenty-second century. That is why I 
believe that if wormholes or other circumventing means are feasible, we will be highly motivated to find and exploit 
them. 
If it is possible to engineer new universes and establish contact with them, this would provide yet further means 
for an intelligent civilization to continue its expansion. Gardner's view is that the influence of an intelligent civilization 
in creating a new universe lies in setting the physical laws and constants of the baby universe. But the vast intelligence 
of such a civilization may figure out ways to expand its own intelligence into a new universe more directly. The idea 
of spreading our intelligence beyond this universe is, of course, speculative, as none of the multiverse theories allows 
for communication from one universe to another, except for passing on basic laws and constants. 
Even if we are limited to the one universe we already know about, saturating its matter and energy with 
intelligence is our ultimate fate. What kind of universe will that be? Well, just wait and see. 
M
OLLY 
2004: