[ExI] Holy cow!

[John Clark]

On Sun, Apr 12, 2026 at 4:47 PM Adrian Tymes via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

>

> * > Perfect cybersecurity for any sufficiently complex system is not
> practical.*


*Pretty good security is possible but perfect cybersecurity is not only
impractical, it is impossible even in theory, unless your computer is so
primitive it can only do first order logic and thus can't even perform
simple arithmetic… and then it would be a computer of very limited
usefulness.  *


>

>
>
> * > However, there is a world of difference between "an attack is possible
> at all" and "an attack is likely enough to seriously worry about", let
> alone "this particular attack will definitely happen any time now". John
> confuses these three. *


*Apparently I'm not the only one who is "confused" about this, so are the
US  Treasury Secretary and the Federal Reserve Chairman, because last
Tuesday they had an emergency meeting with the leaders of the nation's
largest banks and briefed them on the cyberthreats that are now possible**
due to recent developments in Artificial Intelligence, in particular cyber
threats that are now possible due to Anthropic's Mythos and AIs like it. *

*Or maybe, and I know this sounds crazy but ... maybe you're the one who is
confused, not me. *


> *> The ability of AI to discover this many vulnerabilities does not by
> itself move us out of the first category, no matter how much John insists
> otherwise.*
>

*Anthropic made a point of mentioning that Mythos had not only
found thousands of zero day software vulnerabilities Mythos Head also found
computer scripps that could actually exploit them. In the past even when a
human found a zero day error it would take days or weeks to write such a
script, it took Mythos about half an hour.  *


*John K Clark*


>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260413/1a9d3501/attachment.htm>