From: Eugene Leitl (eugene.leitl@lrz.uni-muenchen.de)
Date: Mon Mar 13 2000 - 17:48:05 PST
Kragen Sitaker writes:
> Here's my fear. In the next year or two, someone is going to write a
> worm that penetrates more than a million computers on the Internet.
> They will be quiet about it. They will use this access to embed their
> worm into widely-used operating systems.
[...]
> I considered doing it myself --- because whoever does this will also be
Who hasn't? But you seem to be pretty optimistic about the
difficulties you're going to encounter. Even code camouflaging/
steganographic network traffic tricks and largely dormant code (hardly
any activity, if ever) will show up on packet sniffers, if you're
contemplating 10^6 infected machines.
> in an unsurpassed position to close the ubiquitous security holes that
> infest the world's computers. I decided I couldn't trust myself with
> the power.
It would be much more interesting to not pull the dragon's tail until
2020, or so, and then use the power of Moore and way more nodes
(surely up by then) to find ways to mutate machine code (and FPGAs, or
their successors) robustly (i.e. find a good mutation function by GA
means, in an autofeedback process), then switch to use this to make
the global network intelligent.
Even if the network doesn't transcend, this is going to be a bit more
hassle than a just another worm.
This archive was generated by hypermail 2b29 : Mon Mar 13 2000 - 17:51:37 PST