From: Vladimir Nesov (robotact@gmail.com)
Date: Thu Apr 24 2008 - 19:20:57 MDT
On Fri, Apr 25, 2008 at 4:10 AM, Matt Mahoney <matmahoney@yahoo.com> wrote:
>
> There is a difference between saying "X will destroy humanity" and "we should
> not do X". The correct line of reasoning is "humans do not want to destroy
> humanity, therefore they will desire not to do X".
>
(It's all rather obvious, but I find the separation of decision
process from environment on which it acts clarifying.)
It looks like you are in some kind of determinist confusion. Choice of
action can be seen as a process implemented by a subsystem that
observes its context and has little influence on it during
observation. When choice is done, subsystem modifies its environment
(or not). When environment is being influenced all the time, you can
see this observing subsystem as a meta-level that observes the process
of modifying the environment, but doesn't itself modify it.
When people talk about "we shouldn't do X", they act in the role of
such decision-making process, modeling the default state of the
environment as not affected by their decision-making activity and then
implementing the decision if necessary. If they don't want humanity to
be destroyed, they will decide to modify the environment is a way that
prevents X from being done. It may be implemented in any appropriate
way, and placing a desire not to do X won't necessarily be feasible or
desirable...
-- Vladimir Nesov robotact@gmail.com
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT