Re: Cognitive Systems Conference, CCortex, and alternative friendliness aproaches

From: Eugen Leitl (eugen@leitl.org)
Date: Thu Jul 08 2004 - 02:06:56 MDT


On Wed, Jul 07, 2004 at 04:25:40PM -0500, Bill Hibbard wrote:

> I discuss this issue in my book, Super-Intelligent Machines,
> and other writings. It seems to me that efforts to control
> nuclear, biological and chemical weapons are a healthy
> precedent for controlling AI weapons. I agree that the debate

A healthy precedent for AI containment would be the state of the art in
IT security... which is not so healthy. There is no unpenetrable system out
there, and that's counting known vulnerabilities. A professional group right
now could 0wn some 80-90% of all devices on the Internet.

Right now we have single-installation crossection bandwidths
of some 100 TBit/s and hosting locations with 10^5 systems, nevermind
some 10^9 machines on the global network. SOHO will be at 10 GBit Ethernet in a
few years, and local wireless at >1 GBit/s.

The hardware and networking advances, IT security is arguably regressing.

> will be polluted by lots of lunatics, so it is important to
> educate the public and politicians about the issues.

-- 
Eugen* Leitl leitl
______________________________________________________________
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




This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:22:42 MST