Can a computer work anything out?
 

Can computers work anything out? For example, can a computer solve a
Sudoku or a chess puzzle?

No. Computers and their working programs don't work anything out.
Computers/programs have no target pattern, no goals, no work, no
termination, no halting and no last operation. For example, the end of a
computer game or program is not configured by a "stop" or "pattern". Why
is this? I will tell you.

Each step in a program, whether it is a pattern or not, is a stop. All
stops are a last stop. So, there are no computer/program parameters that
can define a "last stop" as a "halt" or "termination".

Let's be tough about this. Let's be materialistically realistic and not
flounder in the quaint anthropomorphic quasi-reality of the computer
scientist. Let's not make computer homunculi by glossing technical
computer language with the language of goal-seeking human activity.

Computers have no goals or targets. All they do is make sets of stops.
WE imagine these as tasks and terminations, IF it is useful to imagine
them so!

Re: Can a computer work anything out?
 

John Stafford wrote:
> On 7/15/09 1:23 PM, in article h3l75k$mpp$1@news.eternal-september.org,
> "John Jones" <jonescardiff@btinternet.com> wrote:
>
>> Can computers work anything out? For example, can a computer solve a
>> Sudoku or a chess puzzle?
>>
>> No. Computers and their working programs don't work anything out.
>> Computers/programs have no target pattern, no goals, no work, no
>> termination, no halting and no last operation. For example, the end of a
>> computer game or program is not configured by a "stop" or "pattern". Why
>> is this? I will tell you.
>>
>> Each step in a program, whether it is a pattern or not, is a stop. All
>> stops are a last stop. So, there are no computer/program parameters that
>> can define a "last stop" as a "halt" or "termination".
>
> But there are such programs that go on forever, or stop. Cellular Autonoma
> is an example, and there are other.
>
>
>> Let's be tough about this. Let's be materialistically realistic and not
>> flounder in the quaint anthropomorphic quasi-reality of the computer
>> scientist. Let's not make computer homunculi by glossing technical
>> computer language with the language of goal-seeking human activity.
>
> But that is exactly what _you_ are asserting when you infer that to stop or
> have a target is strictly a _human_ thing. By such reasoning we should
> invent all new language for computers - WHICH WE DID. A stop and a target
> (pattern, goal) is a computer term. IF/THEN/ELSE DO WHILE UNTIL and all the
> rest are merely mnemonics that are translated into machine code. It would
> make no sense to call an end/stop by machine code.
>
>> Computers have no goals or targets. All they do is make sets of stops.
>> WE imagine these as tasks and terminations, IF it is useful to imagine
>> them so!
>
> Computers sometimes do have goals and targets,

That's silly, and you know it.

but that's hardware
> implementation so we usually use firmware or software.
>
> You will never change the world by asking the rest of us to be willfully
> ignorant, intellectually impotent, stoned out of our minds.
>

Re: Can a computer work anything out?
 

> Can computers work anything out?

No. In fact, this is a very famous problem in computer science, and the
solution is that there are some problems which a computer (at least using
the model of a Turing machine) cannot solve, the most well known being the
Halting Problem (the problem of determining if a given program will halt; it
cannot be solved in general).

> For example, can a computer solve a
> Sudoku or a chess puzzle?

Yes, a computer can, both have been demonstrated as proofs-of-concept. In
fact, Sudoku solving is a well known contest problem, especially in contests
of who can write the most efficient computer program.

> No. Computers and their working programs don't work anything out.

Yes, they do.

> Computers/programs have no target pattern, no goals, no work, no
> termination, no halting and no last operation.

Sure they do: the operation of being shut down. If you are using a PC with
ACPI support, there is a specific sequence of instructions that cause the
hardware to cease executing further instructions and to power off. Prior to
ACPI, we used a sequence of instructions that caused the hardware to stop
executing further instructions and enter the "wait" state, where it was safe
for a human operator to power off the system.

> For example, the end of a
> computer game or program is not configured by a "stop" or "pattern".

Yes, it is -- PC operating systems general have an "exit" system call which
programs use to indicate that they have terminated and that it is time for
the OS to deallocate whatever resources the program was using.

> Why
> is this? I will tell you.
>
> Each step in a program, whether it is a pattern or not, is a stop. All
> stops are a last stop. So, there are no computer/program parameters that
> can define a "last stop" as a "halt" or "termination".

What are you talking about?

> Let's be tough about this. Let's be materialistically realistic and not
> flounder in the quaint anthropomorphic quasi-reality of the computer
> scientist. Let's not make computer homunculi by glossing technical
> computer language with the language of goal-seeking human activity.

....have you ever looked at a computer science textbook? Or the Church-
Turing thesis? Or even the Wikipedia entries on computer science?

> Computers have no goals or targets. All they do is make sets of stops.
> WE imagine these as tasks and terminations, IF it is useful to imagine
> them so!

In terms of physics, neither do humans: our actions are just undirected
chemical reactions. Let's get real here, and note that: computers are
machines that transition between states, and we generally configure them to
undergo a series of transitions that end in a state (or several states in a
particular pattern) that represents some computed value. We develop
programming languages to make it easier to express those transitions of
states, because it is too complex and expensive to try to define each state
one at a time (except in certain, highly restricted cases).

I recommend reading "Introduction to the Theory of Computation" if you want
a decent book on computing theory and computer science.

One final question: why did you post this in alt.atheism?

-- B

Re: Can a computer work anything out?
 

"John Jones" <jonescardiff@btinternet.com> wrote in message
news:h3l75k$mpp$1@news.eternal-september.org...
> Can computers work anything out? For example, can a computer solve a
> Sudoku or a chess puzzle?

Yes, they can.

> No. Computers and their working programs don't work anything out.

I guess you never heard about that chess match, where Garry Kasparov lost
two out of three games to Deep Blue?

> Computers/programs have no target pattern, no goals, no work, no
> termination, no halting and no last operation. For example, the end of a
> computer game or program is not configured by a "stop" or "pattern". Why
> is this? I will tell you.
>
> Each step in a program, whether it is a pattern or not, is a stop. All
> stops are a last stop. So, there are no computer/program parameters that
> can define a "last stop" as a "halt" or "termination".
>
> Let's be tough about this. Let's be materialistically realistic and not
> flounder in the quaint anthropomorphic quasi-reality of the computer
> scientist. Let's not make computer homunculi by glossing technical
> computer language with the language of goal-seeking human activity.
>
> Computers have no goals or targets. All they do is make sets of stops. WE
> imagine these as tasks and terminations, IF it is useful to imagine them
> so!

Hi, shitforbrains! You shouldn't make it so painfully obvious that you know
nothing at all about computers.

Re: Can a computer work anything out?
 

On Jul 15, 3:30*pm, "Lord Vetinari" <vetin...@ameritech.net> wrote:

> I guess you never heard about that chess match, where Garry Kasparov lost
> two out of three games to Deep Blue?

There is actually a little more to that story.

First off, the conditions of the match.

The computer selected several moves, and a group of people
decided which one to utilize.

Secondly, gary beat the computer in the first game and then
told them how to improve the computer.

Re: Can a computer work anything out?
 

"Tapestry" <estry.tap@gmail.com> wrote in message
news:ddf89d8f-aac6-4811-92a6-d2a2964f64ee@v20g2000yqm.googlegroups.com...
On Jul 15, 3:30 pm, "Lord Vetinari" <vetin...@ameritech.net> wrote:

> I guess you never heard about that chess match, where Garry Kasparov lost
> two out of three games to Deep Blue?

There is actually a little more to that story.

First off, the conditions of the match.

The computer selected several moves, and a group of people
decided which one to utilize.

Secondly, gary beat the computer in the first game and then
told them how to improve the computer.
************************************************** **

Heheheh....I never paid all that much attention to it. I can play chess
reasonably well (I was in the chess club in high school), but I don't enjoy
the game enough...it's just not my bag.

Re: Can a computer work anything out?
 

"John Jones" <jonescardiff@btinternet.com> wrote in message
news:h3l75k$mpp$1@news.eternal-september.org...
> Can computers work anything out? For example, can a computer solve a
> Sudoku or a chess puzzle?
>

Unless it has a program in it that can solve the puzzles you throw at it,
no.

Without programming, a computer is nothing more than a paper weight.

Re: Can a computer work anything out?
 

On Jul 15, 3:49 pm, "Jeff Strickland" <crwlrj...@yahoo.com> wrote:
> "John Jones" <jonescard...@btinternet.com> wrote in message
>
> news:h3l75k$mpp$1@news.eternal-september.org...
>
> > Can computers work anything out? For example, can a computer solve a
> > Sudoku or a chess puzzle?
>
> Unless it has a program in it that can solve the puzzles you throw at it,
> no.
>
> Without programming, a computer is nothing more than a paper weight.

Except there is an entire field, called "genetic programming" or
"genetic algorithms" -- modeled after biological evolution --
in which there is no human involvement, other than to
say "this program (which nobody wrote) works better
than that program (which nobody wrote) to accomplish
task X"

You start with completely random batches of computer
commands, and -- for those that run at all; most
will not -- and see which might go a tiny step
towards your goal. You keep only the best
of the best. Then you "breed" them, exchanging
sections of code between the survivors, and
you may throw in some random "mutations" as
well.

Rinse. Repeat, approximately one zillion
times.

Some very sophisticated programs -- e.g: designing
turbine blades for jet engines -- have been
produced by this approach.

More examples of applications at:

http://en.wikipedia.org/wiki/Genetic...s#Applications

Haiku Jones

Re: Can a computer work anything out?
 

On 15 July, 19:23, John Jones <jonescard...@btinternet.com> wrote:
> Can computers work anything out? For example, can a computer solve a
> Sudoku or a chess puzzle?
>
> No. Computers and their working programs don't work anything out.
> Computers/programs have no target pattern, no goals, no work, no
> termination, no halting and no last operation. For example, the end of a
> computer game or program is not configured by a "stop" or "pattern". Why
> is this? I will tell you.
>
> Each step in a program, whether it is a pattern or not, is a stop. All
> stops are a last stop. So, there are no computer/program parameters that
> can define a "last stop" as a "halt" or "termination".

On the contrary, the fact there there are such parameters allowes the
whole thing to "work".

> Let's be tough about this. Let's be materialistically realistic and not
> flounder in the quaint anthropomorphic quasi-reality of the computer
> scientist. Let's not make computer homunculi by glossing technical
> computer language with the language of goal-seeking human activity.
>
> Computers have no goals or targets. All they do is make sets of stops.
> WE imagine these as tasks and terminations, IF it is useful to imagine
> them so!

YEAH, but use an UNTIL there, OTHERWISE it's trash culture.

-LV

Re: Can a computer work anything out?
 

On 16 July, 00:50, LudovicoVan <ju...@diegidio.name> wrote:
> On 15 July, 19:23, John Jones <jonescard...@btinternet.com> wrote:
>
> > Each step in a program, whether it is a pattern or not, is a stop. All
> > stops are a last stop. So, there are no computer/program parameters that
> > can define a "last stop" as a "halt" or "termination".
>
> On the contrary, the fact there there are such parameters allowes the
> whole thing to "work".

In fact, I find interesting that in the formal description of a Turing
machine, the STOP is a (sort of?) meta-command. Maybe someone could
elaborate on this: the status of the STOP instruction w.r.t. the
specification of a Turing machine?

-LV