Microsoft Word Kurzweil, Ray The Singularity Is Near doc

 
Okay, now run that stealthy scenario by me again—you know, the one where the bad nanobots spread 
quietly through the biomass to get themselves into position but don't actually expand to noticeably destroy 
anything until they're spread around the globe. 
R
AY
:
 
Well, the nanobots would spread at very low concentrations, say one carbon atom per 10
15
 in the biomass, so 
they would be seeded throughout the biomass. Thus, the speed of physical spread of the destructive nanobots 
would not be a limiting factor when they subsequently replicate in place. If they skipped the stealth phase and 

---

expanded instead from a single point, the spreading nanodisease would be noticed, and the spread around the 
world would be relatively slow. 
M
OLLY 
2004:
 
So how are we going to protect ourselves from that? By the time they start phase two, we've got only 
about ninety minutes, or much less if you want to avoid enormous damage. 
R
AY
:
 
Because of the nature of exponential growth, the bulk of the damage gets done in the last few minutes, but your 
point is well taken. Under any scenario, we won't have a chance without a nanotechnology immune system. 
Obviously, we can't wait until the beginning of a ninety-minute cycle of destruction to begin thinking about 
creating one. Such a system would be very comparable to our human immune system. How long would a 
biological human circa 2004 last without one? 
M
OLLY 
2004:
 
Not long, I suppose. How does this nano-immune system pick up these bad nanobots if they're only one 
in a thousand trillion? 
R
AY
:
 
We have the same issue with our biological immune system. Detection of even a single foreign protein triggers 
rapid action by biological antibody factories, so the immune system is there in force by the time a pathogen 
achieves a near critical level. We'll need a similar capability for the nanoimmune system. 
C
HARLES 
D
ARWIN
:
 Now tell me, do the immune-system nanobots have the ability to replicate? 
R
AY
:
 
They would need to be able to do this; otherwise they would not be able to keep pace with the replicating 
pathogenic nanobots. There have been proposals to seed the biomass with protective immune-system nanobots 
at a particular concentration, but as soon as the bad nanobots significantly exceeded this fixed concentration 
the immune system would lose. Robert Freitas proposes nonreplicating nanofactories able to turn out 
additional protective nanorobots when needed. I think this is likely to deal with threats for a while, but 
ultimately the defensive system will need the ability to replicate its immune capabilities in place to keep pace 
with emerging threats. 
C
HARLES
:
 So aren't the immune-system nanobots entirely equivalent to the phase one malevolent nanobots? I mean 
seeding the biomass is the first phase of the stealth scenario. 
R
AY
:
 
But the immune-system nanobots are programmed to protect us, not destroy us. 
C
HARLES
:
 I understand that software can be modified. 
R
AY
:
 
Hacked, you mean? 
C
HARLES
:
 Yes, exactly. So if the immune-system software is modified by a hacker to simply turn on its self-replication