From: William Pearson (email@example.com)
Date: Thu May 01 2008 - 12:30:26 MDT
2008/5/1 Matt Mahoney <firstname.lastname@example.org>:
> --- William Pearson <email@example.com> wrote:
> > > You are assuming that if your brain were augmented with silicon,
> > > that the
> > > carbon part would be in control. Information flows both ways.
> > True, but not all information is equally able to alter the state of
> > the systems (brains and computers)
> From an information theoretic point of view, if one bit of input
> information does not alter the state of the receiver by one bit of
> conditional algorithmic complexity, then communication was wasted.
Lots of communication is wasted, not all of it is relevant to what you
are trying to do, as the agents you are communicating with aren't
perfectly aligned with your goals so may be transmitting useless or
I'm assuming both human and computer are likely to be memory
saturated, that is they would need to forget a bit of information in
order to remember a new bit. The policy of which bits to remember and
which bits to forget (new or old) can be controlled by other bits of
information from the environment. The channel that these bits come in
on can be said to be a privileged channel. If this channel completely
comes from a human, the computer could become part of that human,
dependent upon its initial programming.
> It is not that having machines control your behavior is fundamentally
> bad. It's that humans believe they have free choice and don't want it
> taken away. A brain interface doesn't have to force you to do X
> instead of Y. It just gives you information so that you believe X is a
> better choice.
Nothing has free choice, however not all information channels have an
equal affect on the state of the systems involved. I think it possible
to create computer systems that have a symbiotic relationship, but a
much less ability to change a human than the human does it.
Just don't get your robot to be your sexual partner/best friend, else
it might get more control over you. ;)
This archive was generated by hypermail 2.1.5 : Fri May 24 2013 - 04:01:08 MDT