[extropy-chat] The emergence of AI

[Eugen Leitl]

On Fri, Dec 03, 2004 at 11:11:29PM -0500, Keith Henson wrote:

> If an AI was smart enough to write a virus to take over machines and its 
> intelligence was proportional to "cortical area" (processor cycles) and it 
> "wanted" to get smart fast . . . . .

http://www.icir.org/vern/papers/cdc-usenix-sec02/
http://www.lurhq.com/scanrand.html

Of could you'd do a stealthy mapping first, building a database of individual
hosts and vulnerabilities, before going production. If you 0wn the routers,
segmenting the network becomes more difficult (it would also depend on future
network topology, I'm having a hunch it will be a far more meshed tree).

I don't think the similiarity between spikes and packets is superficial. (You
can actually package spike trains into UDP streams, thus approaching
theoretical maximum utilization, especially on lagged links). 

Networking is doing actually very good (10 GBit/s throughput approaching
pricing where GBit was a few years ago), but the CPUs are hard pressed to
just barely be able to drink from the that hydrant. On network periphery the
problem is the opposite, especially on dialup.

They would be pretty slow, today.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20041204/0b483f1c/attachment.bin>