Turing tests, RQM and Pico



This post was rejected as off-topic in spr. Still I believe it raises
an interesting point, so I am posting it here.

In his blog entry of Sunday, March 12, 2006 Scott Aronson writes that

"... the Turing Test, introduced by Alan Turing in one of the most
famous philosophy papers ever written, is a game where you type back
and forth with an unknown entity in another room, and then have to
decide whether you're talking to a human or a machine. The details are
less important than most people make them out to be. Turing says that
the question "Can machines think?" is too meaningless to deserve
discussion, and proposes that we instead ask whether a machine can be
built that can't be distinguished from human via a test such as his."

Mhhh ...

... a machine that can't be disguinshed from WHICH human?

An anencephalous child?
Someone affected by Down syndrome?
A sneering boy pretending to be a computer?
Britney Spears?
Me?

... and can't be distinguished by WHOM?

An anencephalous child?
A computer?
My mom?
Scott Aronson?


I mean, in order to answer the question we must agree on what we mean
by human. Right?
So I essentially agree with Halpern, whom Aronson quotes as asserting
that " if human intelligence is open to question, then the Turing Test
is meaningless". I don't see how you one apply Turing's test without
such a "humanity" criterion, or without qualifying/sharpening the
test's value and significance.

However, once we sieve "human" out and replace it with "observer",
Turing's test model is highly relevant, sice it provides a conceptual
blueprint for operatively defining "observers" in relational quantum
mechanics. Observers are entities (people or devices, the distinction
is operatively irrelevant) with whom one can exchange information about
measurement outcomes so as to agree on a locus of intersubjective
agreement, a "reality", which will however depend on the observer's
community and on the adopted information-exchange protocol.

Still, the problem "humans" pose to AI and to physics in the light of
Turing text does not go completely away. One can equate "free will"
with unpredictability (cf. [1]) and see Geiger counters and humans as
both possessing free will. Some "human" devices however have the
curious property that they may not just give unpredictable output, but
they choose or even invent its format, its reference frame (e.g. the
unknown entity in the next room may signal "Hey, loser, I have no time
to waste answering your idiotic questions. Go get a life.".). Some
humans choose what they want to be and, for some reason, not everyone
chooses to become a mainstream theoretical physicist. I wonder whether
any computer with those properties exists within current AI's
conceptual horizon.

By the way, the "human" property i am referring to was actually
proposed as a defining feature of humanity in [2].

Cheers,

IV

-----------


[1]
http://groups.google.com/group/sci.physics.research/msg/56cc19cee663b0c0
[2]
"Definita caeteris natura intra praescriptas ... leges cohercetur. Tu,
nullis angustiis cohercitus, pro tuo arbitrio ... tibi illam
prefinies."
"Once defined, the nature of all other beings, is constrained within
the prescribed laws ... . Constrained by no limits, you may determine
it for yourself, according to your own free will ... .
Giovanni Pico "Oratio de hominis dignitate", 1486 , available at
http://www.brown.edu/Departments/Italian_Studies/pico

.