> The New York Times
> April 11, 1995
> Science Times, pp. C1, C10.
>
>
> A Vat of DNA May Become Fast Computer Of the Future
>
Uh, uh, uh, this isn't exactly news to people in the field, John. Go read
Eric Drexler's _Engines of Creation_ to get a look at some of the wilder
possibilities of molecular-scale design. Or read _The Diamond Age_ (I
reviewed it here.) By the way--this is engineering based on placing
molecular-scale elements _in space_--there is therefore a connection with
architectural design.
>
> A DNA computing system would bear no resemblance to a
> conventional computer, raising the question of what a
> computer is. "To some, a computer is a physical device in
> the real world," Dr. Adleman said. But, he added, DNA
> computers have made him think that instead, "being a
> computer is something that we externally impose on an
> object." That means that DNA might not be the only new type
> of computer. "There might be a lot of computers out there,
> and I suspect there are," he said.
>
This is news, oh, from the 1930s or so--go look up the history of
cybernetics and general systems theory, sometime. All these analytical
methods were orginally devised to study living and social systems; the
surprise was the extensive deployment of electronic systems.
>
> Dr. Lipton and his graduate students, Dan Boneh and
> Christopher Dunworth, used some of the same ideas of
> encoding information in DNA to propose an attack on the
> data encryption standard system, or D.E.S. This system was
> invented by t he National Security Agency for Government
> use and is widely used by banks and other companies. Ever
> since the system was invented, computer scientists have
> considered one of the ultimate challenges to be to find a
> clever way to break it. The convention is to set ground
> rules. You know the message that was sent and the encrypted
> message that was received. Now, can you find the key -- the
> sequence of logical operations used to scramble data?
>
DES has already been cracked by (I think) a special-purpose massively
parallel system built at Bell Labs--or was it a bunch of internetworked
systems? It was a Bell Labs, anyway. Same approach, different technology.
DES is not especially strong, anyway--it was deliberately weakened by the
NSA (see cryptoanarcy, passim.)
I'm glad you posted this, but, well--damnit, I wish the Times would do its
homework better!
R.
> April 11, 1995
> Science Times, pp. C1, C10.
>
>
> A Vat of DNA May Become Fast Computer Of the Future
>
Uh, uh, uh, this isn't exactly news to people in the field, John. Go read
Eric Drexler's _Engines of Creation_ to get a look at some of the wilder
possibilities of molecular-scale design. Or read _The Diamond Age_ (I
reviewed it here.) By the way--this is engineering based on placing
molecular-scale elements _in space_--there is therefore a connection with
architectural design.
>
> A DNA computing system would bear no resemblance to a
> conventional computer, raising the question of what a
> computer is. "To some, a computer is a physical device in
> the real world," Dr. Adleman said. But, he added, DNA
> computers have made him think that instead, "being a
> computer is something that we externally impose on an
> object." That means that DNA might not be the only new type
> of computer. "There might be a lot of computers out there,
> and I suspect there are," he said.
>
This is news, oh, from the 1930s or so--go look up the history of
cybernetics and general systems theory, sometime. All these analytical
methods were orginally devised to study living and social systems; the
surprise was the extensive deployment of electronic systems.
>
> Dr. Lipton and his graduate students, Dan Boneh and
> Christopher Dunworth, used some of the same ideas of
> encoding information in DNA to propose an attack on the
> data encryption standard system, or D.E.S. This system was
> invented by t he National Security Agency for Government
> use and is widely used by banks and other companies. Ever
> since the system was invented, computer scientists have
> considered one of the ultimate challenges to be to find a
> clever way to break it. The convention is to set ground
> rules. You know the message that was sent and the encrypted
> message that was received. Now, can you find the key -- the
> sequence of logical operations used to scramble data?
>
DES has already been cracked by (I think) a special-purpose massively
parallel system built at Bell Labs--or was it a bunch of internetworked
systems? It was a Bell Labs, anyway. Same approach, different technology.
DES is not especially strong, anyway--it was deliberately weakened by the
NSA (see cryptoanarcy, passim.)
I'm glad you posted this, but, well--damnit, I wish the Times would do its
homework better!
R.