From: Gordon Worley (firstname.lastname@example.org)
Date: Wed Jun 27 2001 - 08:33:40 MDT
At 8:43 PM -0700 6/26/01, gabriel C wrote:
>>Eletromagnetic radiation escape. This one is pretty simple. The AI
>>sends out signals from the processor that, even though very weak by
>>the time they reach other electronics, provide an avenue to replicate
>>outside the jail. The way around this is to just TEMPEST shield.
>I'm ignorant of the tempest stuff. Can it shield against
>SI-generated Gamma-ray signaling? Cosmic rays? I'm sure that an SI
>could figure out some way to generate these, though they may fall
>into the province of "magic".
I'm not sure if it would really be possible to get a chip to generate
those kind of EM waves. AFAIK, TEMPEST shielding just stops
radiation normally generated by computers (since the goal is to keep
anyone from picking it up to figure out what your computer is doing).
Maybe someone with more knowledge about chips could fill us in.
I should note, while these attacks are possible, I do not consider
them the most likely.
Well, Eliezer, I'm not sure what you're thinking exactly, but maybe
you're thinking of putting messages in the source code? I wouldn't
think so, though, since we've talked about that before on list. Hmm,
assuming no magic at all (not even the little bit that my escapes
need), and no outputs (by this I'm thinking can't write anything
outside of verself and this is because of code outside of ver that
works at a low enough level to be uncrackable or dodgable), code is
about all the AI's got. Store it in memory and then when humans look
at ver memeory take them over? Maybe there's something fundamental
of which I and the list need to be reminded?
-- Gordon Worley http://www.crosswinds.net/~orla_redbird/ mailto:email@example.com PGP Fingerprint: C462 FA84 B811 3501 9010 20D2 6EF3 77F7 BBD3 B003
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:36 MDT