From: Peter C. McCluskey (pcm@rahul.net)
Date: Thu Dec 05 2002 - 14:16:54 MST
mez@apexnano.com (Ramez Naam) writes:
>My problem with only having the degree of control that virus authors
>have over their creations is that computer viruses get out of control
>all the time. There are numerous cases, starting with the original
>internet worm, of viruses that were written as toys and then ended up
>getting loose from their authors and causing a large amount of
>trouble.
>
>With an assembler - a totally novel and extremely powerful type of
>self-replicator - I would want to make very very certain that it would
>not escape into the environment before I ever built one. That means
>modeling and simulating its behavior. Otherwise how do you know that
>the defenses you've built to keep it contained are sufficient?
How many computer viruses have escaped to run on a totally different
type of operating system from the one(s) they were designed to run on?
The natural environment is more complex and hostile to crude replicators
than a Microsoft OS is to simple program written for an Eniac. Would a
person who wrote a virus for an Eniac or an IBM 360 have been reckless
if he failed to simulate whether it could escape to run on a natural
information processing system such as a brain?
I am talking mainly about a time period when there will be a small
number of specialized labs or factories with a carefully purified
feedstock that is needed to support assemblers that are as fragile as
the first stage of most new technologies. I expect that designing an
assembler to deal with thousands of different molecules it will encounter
in the wild without crashing is a lot harder than designing one that will
work when exposed to tens of types of molecules.
If that's not enough, we can add an additional requirement that the
instructions needed to replicate be sent via an external signal and never
have a complete set of instructions stored in the assembler.
What risks would quantum-level simulations deal with that the precautions
I've mentioned won't deal with?
I'm not trying to claim that we can forsee an assembler on many desktops
being safe (the equivalent to having Microsoft OS's readily available for
replicators to spread on).
-- ------------------------------------------------------------------------------ Peter McCluskey | Free Jon Johansen! http://www.rahul.net/pcm |
This archive was generated by hypermail 2.1.5 : Wed Jan 15 2003 - 17:58:35 MST