AI and cognitive psychology rant (getting more and more OT - tell me if I should shut up)

[Michele Simionato]

Stephen Horne <steve at ninereeds.fsnet.co.uk> wrote in message news:

> OK - so why is it not possible to detect the superposition of that
> cat? Why is the experiment still considered a thought experiment only?

As I understand it, you *could* perform the experiment, modulo problems 
with the society for the protection of animals. Actually, those experiments
*have been* performed a number of times and are still performed. Of course, 
they don't use cats, but they call "cats" the wave packets used in the
experiments.

Quantum mechanics (any interpretation of it) will predict that if you repeat
the experiment many times, you will kill the cat 50% of times. Now, if you
really do the experiment, you will really get this rate of success or
failure (depending if you like cats or not), since quantum mechanics
works. 

The point and the crux of the discussion is: "what happens to
the cat *before* opening the box?" This is a matter of principle,
so the experiment is a thought experiment; thought experiment
does not mean it cannot be realized (it's the same for Einstein's
elevator experiment, the basis for the equivalence principle). 

Now, according to Bohr's interpretation
the question simply does not make sense: I don't know anything about
the cat if I do not observe it. It does not make sense to ask if it is 
alive or dead: the only thing that matters is that the cat will be dead
50% of times when I open the box, and dead the other 50%. 
Common sense says that the cat is alive or dead independently from the 
fact that I observe it or not. This is called realism. BTW, I am sure 
you are already familiar with the concept, but let me be verbose.

The orthodox quantum mechanics interpretation is NOT realistic: it
does not assume anything about the state of the cat prior to observation.
In this line of thinking, few centuries ago the Bishop Berkeley asked 
to himself something like this: what happens to a tree when I don't look 
at it? Am I *really* sure that the tree is still there? What if a evil
devil makes the tree disappear when I don't look at it, and reappear
when I look at it? At the end Berkeley solved the paradox by saying
that the reality is real because of God: He is watching all the trees
all the time, so we may be sure that trees are still there even if 
we are not watching them. 

In the last couple of centuries we have
lost our faith in God (fortunately/unfortunately) so now there is
nobody watching the cat inside the box. For this reason we don't
know if it is alive or dead, and we say that it is in a superposition
state. What happens when we open the box (i.e. when we perform the
measurament) we don't know; for sure it is something bizarre which
we model with the wave function collapse argument, but there is
no doubt that the argument is weak and should be replaced by
something better. We use it for lack of better alternatives (the
MWI is not considered a good alternative by the majority of 
physicists).

My point is that this is an INTERPRETATION: depending on your religious
belief you may find the realistic interpretation more or less appealing
(Einstein was against it).

This is not a point of Physics: both interpretation say that when we
measure we will get 50% of dead cats, and actually we get that. This
is Physics; the rest is speculation. You may adhere to the realist
interpretation, but in this case you must loose the property of
locality, and most people are so unhappy with this, that they
prefer to renonce to realism (there is a theorem that you cannot
get both realism and locality for free, and also experiments on
Bell's unequalities, so this is "for sure").

What it is really interesting it to understand how measurement works,
how to pass from microscopic to macroscopic, how to give a better
description of what happens behind the wave function collapse. 
This is an *hard* job, but there is an active line of research on 
*these* issues and I don't think we have lost hope. At the end, a theory
providing *new*  verifiable predictions will emerge. A theory
providing the *same* predictions but a different interpretation
will not have any success, since results are more important than
interpretations (at least, to most physicists). We are more pragmatic and
less academic than other academic people, I think ;)

I should have got the argument right, but please remember that 
these topics are typically not taught in courses, since they are not 
considered very relevant, and I have last studied Berkeley's argument
more than 15 years ago, so I may remember uncorrectly. Here I am giving 
my opinion and I don't have an autoritative reference on the subject now. 
So, my word should be taken  "cum grano salis". On the other hand,
the word of everybody should be taken "cum grano salis" ;)

HTH,

                       Michele Simionato