Geddes’s final hurrah: Going out from SL4 ‘guns blazing’ to try to prove Objective Morality

From: Marc Geddes (marc_geddes@yahoo.co.nz)
Date: Fri Aug 05 2005 - 03:30:52 MDT


‘guns blazing’, intellectual attack 1

Is ‘Volition’ the foundation of mind? No, because one
can ask what is it about mind that allows us to reason
and think about ‘Volition’ in the first place?
Clearly there are fundamental laws of cognition that
determine how thought is correlated with the physical
substrate (i.e. brain) on which thought is enacted.
But these ‘functional laws’ (which determine how
thought maps to physical processes) must be the same
for all sentients and are hence objective. This shows
that there is an objective core to cognition which
goes beyond mere volition. Further, the *process of
cognition itself* must be of more importance than
volition, because we couldn’t reason about volition or
act upon volition without it. For instance: when
helping someone fulfill their desires (altruism), it
is not only the desires of others that we value (or
place utility on), but also *the cognitive process by
which we reason and take action* - because without
this cognitive process we would not be able to reason
about or act on volition in the first place.

This shows that the real foundation of value judgments
is not volition, but *self actualization*, the ability
to reason about (be aware of) and act upon our true
nature (‘true nature’ being the fundamental laws of
cognition that determine how thoughts map to the
physical substrate of our brains). But these
fundamental laws of cognitive science which determine
how thoughts map to brain state are objective in
nature. This proves objective morality.

‘guns blazing’, intellectual attack 2

A real intelligence has to do more than simply form
models about how the world works (i.e. make
predictions). Models about the world have to be
integrated with a *value system*, i.e. utilities, in
order to make *projections* (project consequences of a
range of possible actions and *morally assess* these
consequences). But as mentioned, these evaluations
(Value judgments) about possible outcomes involve
*integrating* the prediction system (which enables the
AI to model the world) with the system for reasoning
about volition (which enables the AI to classify and
make choices about what agents *ought* to do).

The process of integration itself (linking volition to
prediction) is not volitional in nature. This proves
a level of cognition beyond the volitional level.
Further, in order for predictions to be properly
integrated with choices, there has to be ordered (non
random) rules to connect choices to prediction. But
such objective rules would constitute an objective
morality. This proves objective morality.

‘guns blazing’, intellectual attack 3

Wei Dai defined intelligence to be ‘the ability to
solve problems with limited computational resources’.
An AI cannot in general simply ‘brute force’ solutions
to problems because it’s resources (time and memory)
are limited. Any AI has to use *heuristics* to narrow
down which lines of argument to pursue. But
heuristics are based on *value judgments*. For
instance even hard scientists routinely talk about
about makes ‘an elegant proof’, ‘a beautiful argument’
etc. This proves that even for something as
scientific as mathematical problem solving, values
(utilities) are involved. So ‘a little of morality’
has to be mixed in with raw IQ for there to be any
workable intelligence in the first place. This proves
that intelligence is not totally separate from
morality.

Further, the more complex the problem (i.e. the more
possible elements involved), the more important
heuristics become. This is because as the complexity
of a problem increases, there is a combinatorial
explosion in ‘possible lines of argument’ and more and
more heuristics are needed to in order to cut down
possible lines of argument enough to make the problem
intractable.

But as was mentioned above, heuristics imply values
judgments. This proves as problems become more and
more complex, value judgments become more and more
important.

Imagine a knob for ‘turning up IQ’. The higher the
IQ, the more complex the kinds of arguments a sentient
can make. But as shown above, the more complex the
argument, the more important heuristics become and
hence the more value judgments are needed. Past a
certain IQ threshold, *value judgments dominate
problem solving ability* and hence intelligence starts
to become more and more indistinguishable from
morality (value judgments). So past a certain
threshold, intelligence implies morality. This proves
objective morality.

‘guns blazing’, intellectual attack 4

Conrey’s recent mathematical proof of the Free Will
Theorem.

“In mid-2004, John Conway and Simon Kochen of
Princeton University proved the Free-will Theorem.
This theorem states "If there exist experimenters with
(some) free will, then elementary particles also have
(some) free will."

Link:
http://www.cs.auckland.ac.nz/~jas/one/freewill-theorem.html

Conrey and Kochen have mathematically proved that if
the concept of ‘Volition’ is meaningful at the level
of description of macroscopic agents like humans, *it
is also meaningful even at the microscopic level*.
This proves that ‘Volition’ is *not* simply a
macroscopic, emergent property, but is in fact a
*fundamental* (exists even at the microscopic level)
property of reality.

If Volition is a fundamental property of reality
(possessed even by inanimate objects), then *so is
morality*. Morality after all, is simply a corollary
of volition. If microscopic particles have free will
(as Conrey mathematically proved), they also have
volition and if they have volition , even these
inanimate particles can be thought of as moral agents.
 This proves objective morality.

‘guns blazing’. Combo intellectual attack combining
1,2,3,4

By (1) it was established that an AI has to do more
than simply make predictions (form models about the
world). It also has to reason about what moral agents
*ought* to do (i.e. make predictions) and then
*integrate* the two kinds of reasoning in order to be
able to make value judgments about the consequences of
possible courses of action. But the process of
integrating the prediction system with the volitional
reasoning system cannot itself be part of either
system. This proves a level of cognition beyond the
volitional level.

By (2) it was established that Volition cannot form
the foundation for value judgments because there is
something even more important than volition: that is
the ability to *be aware of the very concept of
volition and reason about it in the first place* i.e.
the cognitive process which enables us to be
self-aware and act in the world. Cognition itself
depends on objective laws from cognitive sciences
which determine how thoughts map to brain states.
Going back to (1), the process of integrating
predictions with value judgments is not itself a part
of the volitional level, but must depend of the
fundamental objective laws of cognitive science which
determine how thoughts map to brain states. But an
objective mapping between a prediction system and a
value system proves objective morality… ability to
form correct value judgments is objectively correlated
to ability to make prediction.

Further support is obtained by (4). Conrey’s ‘Free
Will Theorem’ mathematically proved that the concept
of free will is meaningful even at the microscopic
level:

http://www.cs.auckland.ac.nz/~jas/one/freewill-theorem.html

  But if free will is meaningful even at the
microscopic level, then so is the concept of Volition
and the concept of morality as corollary. So even
inanimate objects (and as Conrey showed, even
microscopic particles) are moral agents. But since
all brains are ultimately composed of microscopic
particles, and since morality is meaningful for *all*
particles (as Conrey showed), this proves that
objective moral laws are built into the operations of
*all* sentient brains, in the sense that these laws
are needed for cognition in the first place.

The case for objective morality clinched by the above
combined with (3). A real intelligence has to solve
problems with limited computational resources. This
requires intelligent heuristics to cut down possible
lines of argument. But heuristics involve value
judgments. This proves that you can’t have
intelligence without some degree of *value judgment*.
But by the previous arguments, the nature of the value
judgments required for intelligent heuristics must be
objective, since the very process of cognition itself
requires brains, brains require particles and by
Conrey mathematical proof, even particles on a
‘microscopic amount of moral principle’.

It was further pointed out that the more complex an
intellectual argument, the more important heuristics
become, due to a combinatorial explosion of ‘possible
lines of argument’, which require ever more
intelligent heuristics to maintain tractability. This
proves that increasing IQ (ability to deal with ever
more complex problems), requires ever better value
judgments, and as shown by Conrey mathematical proof,
the nature of value judgments is objective in nature.
This proves that at a high enough IQ, intelligence
starts to become indistinguishable from mortality.

The case for objective morality is made.

---
Please vist my website:
http://www.riemannai.org
Science, Sci-Fi and Philosophy
---
THE BRAIN is wider than the sky,  
  For, put them side by side,  
The one the other will include  
  With ease, and you beside. 
-Emily Dickinson
'The brain is wider than the sky'
http://www.bartleby.com/113/1126.html
Send instant messages to your online friends http://au.messenger.yahoo.com 


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:51 MDT