[extropy-chat] Re: An Error in Nanosystems!

Chris Phoenix cphoenix at CRNano.org
Thu Apr 8 16:51:33 UTC 2004

Hal Finney writes:
 > The point is that the guy can make mistakes, just like anybody else.

This is news?  No one claims he's infallible.  This conversation is 
starting to get surreal.

The real news is that, after more than a decade, no one has found an 
error in Nanosystems that appears to signficiantly reduce the 
possibility of diamondoid nanomachines working as Drexler described.

Jeff Soreff told me about an actual math error: Drexler allegedly 
double-counts degrees of freedom in computing the number of possible 
structures with a certain surface.  But Jeff said that this didn't make 
a significant difference.

Richard Jones pointed out to me that Drexler hadn't explicitly 
considered the possibility of surface atoms wiggling sideways and then 
snapping back, causing friction.  I think Drexler considered it 
implicitly, but this might count as an error.  Fortunately, 
superlubricity has now been observed in at least two rather different 

But these are the only two I know of.

 >> If the whole bearing were
 >> relatively prime, it would wobble as the region of alignment traveled
 >> around the axis.
 > I don't know about that: see figure 10.19, which has 6 inside and
 > 11 outside, relatively prime.

Huh.  I guess I was wrong about Drexler being worried about wobble at 
the time he wrote Nanosystems.  But I still think I remember it from 
somewhere.  Doesn't really matter.


I think your analysis is wrong.  You're assuming that the H will be able 
to "pop" to a low-energy state when it is in a high-energy position. 
But to fit between two other H's, its bond would have to be 
significantly twisted.  So the energy of that configuration would be 
significantly higher than the low-energy value.  In fact, I bet it's not 
even a minimum.

As I understand it, the way to figure it is to relax the structure with 
two opposing H's, and compute the energy for the "pop" by subtracting 
the the relaxed structure's energy from the opposed structure's energy. 
  Then compare that energy to thermal energy to see the expected 
transition time.  Then compare that to the velocity (it's actually not 
quite that simple) to see how often it will be dragged over the "pop" 
vs. going over on its own.

And after all that, if you get a number we don't like, we're allowed to 
redesign the bearing to remove H's from the interface.

 > No, I didn't even know there was a hardcover version.  I bought the
 > book shortly after it came out and I have a first printing, I believe
 > (evidenced by the "10 9 8 7 6 5 4 3 2 1" on the bottom of the copyright
 > page).

refers to a hardcover edition.

My edition says "10 9 8 7 6 5 4 3".

 > I would hate to think that Drexler is hiding the fact that his book had
 > errors in order to promote that absurd and counterproductive myth that
 > it is perfect!

Fortunately, you don't have to think that.  The myth is a strawman, 
obviously ridiculous.  I would hate to think that you're so eager to 
believe Drexler is both stupid and dishonest...

 >  For an important and foundational book like Nanosystems,
 > not providing errata is a lapse which should be corrected.

IMHO, it depends on whether there are any errata worth noting.  Despite 
my earlier joking, I actually think the typo I found is more significant 
than the "relatively prime" thing.  But I found that less than a year 
ago.  Anyway, I'll ask him about errata lists.


Chris Phoenix                                  cphoenix at CRNano.org
Director of Research
Center for Responsible Nanotechnology          http://CRNano.org

More information about the extropy-chat mailing list