Copy Link
Add to Bookmark
Report
AIList Digest Volume 6 Issue 094
AIList Digest Monday, 9 May 1988 Volume 6 : Issue 94
Today's Topics:
Philosophy - Free Will, Self-Awareness
----------------------------------------------------------------------
Date: 5 May 88 14:40:56 GMT
From: planting@speedy.cs.wisc.edu (W. Harry Plantinga)
Subject: Free Will & Mcdermott's argument
In article <28437@yale-celray.yale.UUCP> dvm@yale.UUCP (Drew Mcdermott) writes:
>. . . any system that is sophisticated enough to model situations that its own
>physical realization takes part in must flag the symbol describing that
>realization as a singularity with respect to causality.
> . . . robots with this kind of structure simply can't help but think of
>themselves as immune from causality in this sense.
>Even after a reasoner has become sophisticated about physical
>causality, his model of situations involving himself continue to have this
>feature. That's why the idea of free will is so compelling.
>I would rather not phrase the conclusion as "People don't really have
>free will," but rather as "Free will has turned out to be possession of
>this kind of causal modeler." So people and some mammals really do have
>free will. It's just not as mysterious as one might think.
>
> -- Drew McDermott
Taking this as an argument that people don't have free will in the
common sense, let's see what we have. Is this a fair restatement
of the argument?
(1) Systems that reason about causality and that reason about
themselves have a singularity. They must consider themselves
exempt from causality.
(2) Therefore people are subject to causality and do not have
free will (in the common sense).
As it stands, this is not a sound argument. Clearly there are a couple
of unwritten assumtions here. Perhaps these:
(1.1) People are "systems that reason about causality" (assuming
materialism)
(1.2) The feeling of free will is the same thing as the feeling
of not being able to reason about causality with respect to self.
If we accept (1), (1.1), and (1.2) we still can't conclude that people
don't have free will. The best result that can be argued is
(2') If people didn't have free will, they would still feel that
they did.
Note that this is very much distinct from (2), and even this argument
is based on some highly disputable premisses. (1.1) is surely not
commonly agreed upon outside of AI, and (1.2) is also dubious. My free
will doesn't feel like the simple inability to figure out what I am
going to do in the future by reasoning about causality. My free will
feels like the ability to choose among alternatives no matter what my
reason, emotion, conscience etc. say is more "reasonable."
As an argument that people don't have free will in the common sense,
this would only be convincing to someone who holds (1.1) and (1.2) and
that (2') implies (2), i.e. someone who already thinks people don't
have free will.
Harry Plantinga
planting@cs.wisc.edu
------------------------------
Date: 3 May 88 08:45:49 GMT
From: mcvax!ukc!reading!onion!henry!jadwa@uunet.uu.net (James
Anderson)
Subject: Re: Free Will & Self Awareness
In article 1380 of comp.ai: bwk@mitre-bedford.ARPA (Barry W.
Kort) says:
> Suppose I were able to inculcate a Value System into silicon.
> And in the event of a tie among competing choices, I use a
> random mechanism to force a decision. Would the behavior of
> my system be very much different from a sentient being with
> free will?
Well ...
I take, "free will" to mean that an agent can choose what action
to take despite the choices that other agents might make and
despite unchosen events in the world.
There are a number of corollaries to this definition.
1) I am a strong minded person, so I often exercise free will,
but you can deny me free will, say, by killing me.
2a) You might exercise your free will by making an oath with
yourself never to deny me free will, say, by never applying
irresistible force to me.
2b) Making an oath does not deny your own free will. You can
chose to break the oath.
3a) Random events in the choice mechanism deny free will to the
extent that they prevent the agent from determining the
outcome of a decision.
3b) I know of no way to discover, by observing the behaviour of an
agent, that exception (3a) applies to it. So the answer to
your question is "yes": I can not tell apart the behaviour of
a random system and one with free will. A free-will system
could, for example, chose to behave randomly.
4) An event which is not of any agents choosing might deny me
free will. The event may be random, as in losing a game of
Russian roulette, or deterministic, like being being drowned
in a rising tide.
So far, so good. But here comes the free will paradox, with knobs
on!
If the world is deterministic I am denied free will because I can
not determine the outcome of a decision. On the other hand, if
the world is random, I am denied free will because I can not
determine the outcome of a decision. Either element, determinancy
or randomness, denies me free will, so no mixture of a
deterministic world or a non-deterministic world will allow me
free will.
I am going to think about that for a little while before I post
again.
James
(JANET) James.Anderson@reading.ac.uk
------------------------------
Date: 5 May 88 15:36:34 GMT
From: sunybcs!sher@boulder.colorado.edu (David Sher)
Subject: Re: Free Will & Self Awareness
It seems that people are discussing free will and determinism by
trying to distinguish true free will from random behavior. There is a
fundamental problem with this topic. Randomness itself is not well
understood. If you could get a good definition of random behavior you
may have a better handle on free will.
Consider this definition of random behavior:
X is random iff its value is unknown.
This is I believe a valid definition of randomness. But in this case
free will may be a subset of random behaviors. Other more
sophisticated definitions may be proposed for randomness.
On a similar note to decide if we can allow machines to take
responsibility (which seems to be bothering our english contingent), we
must decide just what responsibility is. We already entrust machines
with our lives and have for thousands of years (since we invented
boats).
-David Sher
ARPA: sher@cs.buffalo.edu BITNET: sher@sunybcs
UUCP: {rutgers,ames,boulder,decvax}!sunybcs!sher
------------------------------
Date: 4 May 88 12:55:36 GMT
From: vu0112@bingvaxu.cc.binghamton.edu (Cliff Joslyn)
Subject: Re: Free Will & Self-Awareness
In article <30800@linus.UUCP> bwk@mbunix (Barry Kort) writes:
>The essential idea that
>individual behavior is generated by a mix of causal elements (agency
>motivated by awareness of the state-of-affairs vis-a-vis one's value system)
>and chance (random selection among equal-valued alternatives).
This is a central tennet of Systems Science. Absolute determinism is
possible and relatively common; absolute freedom is impossible; relative
freedom is possible and relatively common. Most (all?) real systems
involve mixes of relatively free and determined elements operating at
multiple levels of interaction/complexity.
It should be emphasized that this is not just true of mental systems,
but also of biological and even physical systems. As one moves from the
physical to the biological and finally to the mental, the relative
importance of the free components grows. Intelligent organisms are more
free than unintelligent organisms; which are more free than
non-organisms.
None of the above are absolutely free. No-one even knows what it might
mean to be absolutely free.
>--Barry Kort
--
O---------------------------------------------------------------------->
| Cliff Joslyn, Cybernetician at Large
| Systems Science, SUNY Binghamton, vu0112@bingvaxu.cc.binghamton.edu
V All the world is biscuit shaped. . .
------------------------------
Date: 3 May 88 16:04:00 GMT
From: channic@m.cs.uiuc.edu
Subject: Re: AIList V6 #86 - Philosophy
In his article Brian Yamuchi (yamauchi@speech2.cs.cmu.edu) writes:
> /* ---------- "Re: AIList V6 #86 - Philosophy" ---------- */
> In article <368693.880430.MINSKY@AI.AI.MIT.EDU>, MINSKY@AI.AI.MIT.EDU
(Marvin Minsky) writes:
> > Yamauchi, Cockton, and others on AILIST have been discussing freedom
> > of will as though no AI researchers have discussed it seriously. May
> > I ask you to read pages 30.2, 30.6 and 30.7 of The Society of Mind. I
> > claim to have a good explanation of the free-will phenomenon.
>
> Actually, I have read The Society of Mind, where Minsky writes:
>
> | Everything that happens in our universe is either completely determined
> | by what's already happened in the past or else depends, in part, on
> | random chance. Everything, including that which happens in our brains,
> | depends on these and only on these :
> |
> | A set of fixed, deterministic laws. A purely random set of accidents.
> |
> | There is no room on either side for any third alternative.
>
I see plenty of room -- my own subjective experience. I make mental
decisions which are not random and are not completely determined (although
certainly influenced) by past determinism. Minsky wondered why his
explanation seems to have eluded philosophers of the past. I am not
surprised because evidently he is just being swept away out of control
in an entirely random or totally determined universe. The philosophers
of the past, like myself, were probably like me -- intelligent, free will
sentient beings. Are these philosphers and myself in the minority?
I think not, but I am surprised that such views constitute RESPECTED,
let alone, PUBLISHED material on the subject. Certainly this objectivist
viewpoint helps the discipline of AI: people (i.e. funding agences) will
be more likely to believe that a machine can be intelligent if intelligence
can be reduced to a set of purely deterministic laws. But this BEGS THE
QUESTION of intelligent machines in the worst way. Show me the deterministic
laws that create mind, Dr. Minsky, then I will believe there is no free will.
Otherwise, you are trying to refute an undeniable human experience.
Do you believe your career was merely the result of some bizarre genetic
combination or pure chance?
The attack is over. The following is a plea to all AI researchers. Please
do not try to persuade anyone, especially impressionable students, that s\he
does not have free will. Everyone has the ability to choose to bring peace
to his or her own life and to the rest of society, and has the ability to
MAKE A DIFFERENCE in the world. Free will should not be compromised for the
mere prospect of creating an intelligent machine.
Tom Channic
University of Illinois
channic@uiucdcs.cs.uiuc.edu
{decvax|ihnp4}!pur-ee!uiucdcs!channic
------------------------------
Date: 5 May 88 06:12:45 GMT
From: vu0112@bingvaxu.cc.binghamton.edu (vu0112)
Subject: Re: Free Will & Self Awareness
In article <770@onion.cs.reading.ac.uk> jadwa@henry.cs.reading.ac.uk
(James Anderson) writes:
>3a) Random events in the choice mechanism deny free will to the
> extent that they prevent the agent from determining the
> outcome of a decision.
Interesting. So you understand freedom to be the ability for me to
determine my own actions, as opposed to them being determined by
external sources. Thus you're defining freedom in terms of a different
kind of determinism, a stance which seems problematic.
Why not instead freedom as simply the *lack* of *complete* external
control? This way you allow *degrees of freedom*. That is, I am free
if I am not determined. This in no ways implies that I in turn must
determine something else.
This also helps us sort out the difference between subjective and
objective uncertainty (freedom). Objective freedom is understood as the
(e.g. quantum) inherent uncertainty in processes, whereas subjective
uncertainty is the premise that I lack information about a possibly
determinate process. On my definition, both conditions indicate freedom
to me, but to you in the latter case we are still determined.
>So far, so good. But here comes the free will paradox, with knobs
>on!
I think my definition knocks your knobs off.
>If the world is deterministic I am denied free will because I can
>not determine the outcome of a decision.
For me, if the world is deterministic I am granted freedom when I cannot
determine the outcome of a decision.
>On the other hand, if
>the world is random, I am denied free will because I can not
>determine the outcome of a decision.
Yes, no matter if the world is determined or not, I can never determine
the outcome of a decision. This is an inherent epistemic limitation,
independent of the state of the world.
I think that's obviously correct. Further, by showing us that
determinism is impossible, I think you've just demonstrated that freedom
is necessary.
>I am going to think about that for a little while before I post
>again.
What I find fascinating is the implicit assumption that determinism is
the "normal" general case, and that freedom is come kind of strange
property rarely seen. I think that just the opposite is true, that
deterministic processes of any kind are very rare. Demonstrating the
freedom of *any* organism, let alone humans, is trivial. Demonstrating
our determinism is a silly philosophical waste of time.
--
O---------------------------------------------------------------------->
| Cliff Joslyn, Cybernetician at Large
| Systems Science, SUNY Binghamton, vu0112@bingvaxu.cc.binghamton.edu
V All the world is biscuit shaped. . .
------------------------------
Date: 5 May 88 07:00:47 GMT
From: quintus!ok@unix.sri.com (Richard A. O'Keefe)
Subject: Re: Sorry, no philosphy allowed here.
In article <1069@crete.cs.glasgow.ac.uk>, gilbert@cs.glasgow.ac.uk
(Gilbert Cockton) writes:
> > What about compatibilism? There are a lot of arguments that free will is
> > compatible with strong determinism. (The ones I've seen are riddled with
> > logical errors, but most philosophical arguments I've seen are.) (O'Keefe)
> I would not deny the plausibility of this approach. However, detection of
> logical errors in an argument is not enough to sensibly dismiss it, otherwise
> we would have to become resigned to widespread ignorance. My concern over AI
> is, like some psychology, it has no integration with social theories,
especially
> those which see 'reality' as a negotiated outcome of social processes, and
not
> logically consistent rules. If the latter approach to 'reality', 'truth'
etc.
> were feasible, why have we needed judges to deliver equity? For some AI
> enthusiasts, the answer of course, is that we don't. In the brave new world,
> machines will interpret the law unequivocably, making the world a much fairer
> place:-) Anyway, everyone know's that mathematicians are much smarter
> than lawyers and can catch up with them in a few months. Hey presto,
> rule-base!
The rider was intended to indicate that I neither endorse nor reject
compatibilism. In philosphy, I fear that I _have_ become resigned to
ignorance (though some recent moral philosophers have raised my hopes).
Concerning 'reality' as a negotiated outcome of social processes, do you
remember the old days when we were assured that colour terms were a social
construct, and that biological species were merely a theoretical construct
imposed on a world with no boundaries? An archaeologist once commented in
a lecture that I attended that a number of ancient peoples had the practice
of "killing" the deceased's possessions (breaking bowls, bending knives,
tearing cloth &c), and that by and large the artifacts that had been
damaged least (e.g. just a "token" chip out of the rim of a bowl) were the
ones the archaeologists found most beautiful (comparing reconstruction with
reconstruction). Negotiated social processes between people 5,000 years
apart?
Why have we needed judges to deliver equity?
(a) Because anyone who does that we _call_ a judge.
(b) They don't.
(c) The use of judges has nothing to do with the nature of reality.
The facts may be known to both parties, yet either or both may
simply be pushing to see what it/they can get away with.
Ever heard of "the Kerry alibi?"
(d) This is something of a loaded example, insofar as law _is_ a socially
negotiated field (particularly in the USA). But none of the admittedly
few books on jurisprudence I've read makes any claim that the law is an
instrument for reaching 'truth' or 'reality', only a mechanism for
reducing the level of dispute in a society to a workable degree. (It
doesn't even matter too much if a law is and is known to be unjust,
so long as you know what it _is_ and can rely on it well enough to avoid
its consequences.)
For an example of someone who _has_ looked at AI with an eye to sociology,
Plans and Situated Actions : The problem of human/machine communication
Lucy A. Suchman
Cambridge University Press 1987
ISBN 0-521-33739-9
------------------------------
Date: 3 May 88 15:46:10 GMT
From: otter!cwp@hplabs.hp.com (Chris Preist)
Subject: Re: Free Will & Self-Awareness
O.S. writes...
>> I would like to learn how to imbue silicon with consciousness,
>> awareness, free will, and a value system.
>
> First, by requesting that, you are underastimating yourself as a
free-willing
> creature, and second, your request is self-contradicting ans shows little
> understanding of matters, like free will and value systems - such things
cannot
> be 'given', they simply exist. (Something to beare in mind for other
purposes,
> besides to AI...). You can write 'moral' programs, even in BASIC, if you
want,
> because they will have YOUR value system....
Sorry, but not correct. While it is quite possible that the goal of 'imbuing
silicon with a value system' may never bge fulfilled, it is NOT correct to
say that values simply exist.
Did my value system exist before my conception? I doubt it. I learnt it,
through interaction with the environment and other people. Similarly, a
(possibly deterministic) program MAY be able to learn a value system, as
well as what an arch looks like. Simply because we have values, does not
mean we are free.
On the question of Free Will - simply because someone denies that we are
truly free, does not mean they have little understanding of the matter.
As Sartre pointed out, we have an overwhelming SUBJECTIVE sensation of
freedom. Questioning this sensation is a major step, but a step which
has to be made. Up to now, the problem has been purely metaphysical. An
answer was impossible. But now, AI provides an investigation into
deterministic intelligence. I believe it IS important for AI researchers
to make an effort to understand the philosophical arguments on both sides.
Maybe your heart will lie on one of those sides, but the mind must remain
as open as possible.
Chris Preist
P.S. Note that AI only provides a semi decision procedure to the problem
of free will. Determinism would be proven (though even this is debatable)
if an 'intelligent' deterministic system were created. However, if objective
free will exists, then we could hack away with the infinite monkeys, all
to no avail.
------------------------------
End of AIList Digest
********************
