> I wasn't very clear in my last post. What I meant was this:
> (a) A conscious program written in C is compiled on a computer. The C
> instructions are converted into binary code, and when this code is run, the
> program is self-aware.
> (b) The same conscious program is written in some idiosyncratic programming
> language, created by a programmer who has since died. He has requested in
> his will that the program be compiled, then all copies of the compiler and
> all the programmer's notes be destroyed before the program is run. Once
> these instructions are carried out, the binary code is run, and the program
> is self-aware as before - although it is difficult or impossible for an
> outsider to work out what is going on.
> (c) A random string of binary code is run on a computer. There exists a
> programming language which, when a program is written in this language so
> that it is the same program as in (a) and (b), then compiled, the binary
> code so produced is the same as this random string.
> Is this nonsense?
No, not to me. All that seems to be making perfect sense.
> Is (c) fundamentally different from (b)? If not, doesn't
> it mean that any random string implements any program?
Well, this is obviously a very *special* random string! :-)
But no, (c) is fundamentally the same as (b).
But perhaps I can help make your point (though I'm not sure
what your point is). It may be that there is some sequence
of 1's and 0's that when run on a machine performs some
operation (say it implements a conscious friend), but which
is not the compiled output of *any* human readable code.
Vernor Vinge in "A Fire Upon the Deep" mentioned some Beyond
technology that didn't seem to be modular: it was just
spaghetti from our viewpoint. This is of a kind---in my
opinion---with the weird 322 King+Rook+Bishop vs. K+N+N
chess ending which evidently isn't made up of any concepts
or building blocks: all one has is a series of bewildering
chess positions that *just by sheer chance* it seems lead
to checkmate. (Normally in chess, any process that leads to
checkmate does so for clear reasons. Not this. It's just a
set of random looking arrangements of pieces that get
closer and closer to checkmate.)
Of course, taken literally, your last sentence is false: it
is *not* (of course) true that any random string implements
> We might not know what it says, but if the program is self-aware,
> then by definition *it* knows.
Yes, I think so. Whether it's the result of a compilation,
or it's just a raw machine code that is self-aware (and, say
passes the Turing Test), then he seems just as good as you
Sorry, but where are you going with this? :-)