CD's To Take Over Software Distribution?

Computer trade publications are abuzz over plans by Microsoft and
by IBM to distribute software on CD-ROMs. Retailers are afraid that
they will be cut out of the software sales market.

Customers can get CDs packed with software which can be tried out.
If you find a program you want to use, you can call the publisher,
give your credit card number, and get a code which will let you
decrypt the full program from the CD-ROM.

Interestingly, shareware got its start as an alternative marketing
method by programmers who could not get their products into stores
because of major software companies with lots of money to spend
taking up most of the shelf space.

If retail software moves into a shareware type of distribution on
CD-ROM, the Big Boys could lose their marketing edge. On a CD-ROM,
there is no prime store position nor fancy store displays. All
software has the same chance and may the best program win.



Can Computers "Reason"?

We have read or been involved in several discussions recently about
whether or not a computer can "reason".

The least knowledgable of those claiming "No" is Dr. Joyce Brothers,
who in her syndicated newpaper column, told readers: "Computers can
give us instant dates and as many facts as we want about most things.
What they can't do is think, compare one fact with another, measure,
evaluate, reason, and come up with ethical judgments."

A newspaper bridge columnist, Phillip Alder, told his readers that
a computer will never be able to play expert level bridge because
they are incapable of playing hunches and making plays which are
intended to fake out the opponent.

Both of these statements are by people who are not computer experts
and most certainly have no scientific basis for their claims. We have
card games (and other games) in PsL which the computer examines the
play of the opponent, evaluates the expected results of possible
responses, and makes judgments about what the best play would be,
including plays designed just to "fake out the opponent".

These are not programmed responses. That is, the computer has not been
programmed with a set reponse to every play the opponent might make.
Instead, the computer goes through precisely the same thought processes
which the human does in deciding what to play.

In a debate on the MENSA forum on Compuserve, the claim was made that
computers are incapable of human type thought - that they can only
respond as they were programmed to do.

Said one member:
     Try typing "How are you today?". It will respond with one of
     the programmed responses that you taught it. You just taught
     it things that a human might say.

The counter-question is "Is a human's thought processes any
different?"

If you ask a human "How are you today?", it is going to respond
with one of the preprogrammed responses (vocabulary) that someone
taught it. If it doesn't, then its response would be unintelligible.

The human has also been taught "things that a human might say." If you
doubt that, go ask a 1-year old "How are you today?" or ask someone
who only understands Swahili, or anyone else who has not been
"programmed" to understand English.

When you ask a human a question, you must ask it in a way/language
which the human will understand, just like with a computer.

Then the human's brain will check with the body to see how it
feels and check memory for any information stored there about the
body's present condition.

Then it will translate this information into language (which SOMEBODY
taught it) to give you an answer.

This is virtually identical to what the computer goes through when you
type CHKDSK. It translates my request for information using the
vocabulary it has been taught; it examines itself to come up with an
answer; and it translates its answer back to the vocabulary which it
has been taught.

The only essential difference between how a human handles such a
question and how a computer handles it is the difference in
complexity, education ("programming"), and mental capacity between the
two.

Some day in the future, assuming that computer advances continue their
upward trend indefinitely, that difference should disappear altogether.

Twenty years ago, nobody believed that a computer could beat a master
at chess. That challenge was met.

Twenty years from now, computers probably still will not be able to
match the complexity and mental capacity of a human, but it will
doubtlessly be able to understand speech and respond in kind. When
the time comes when you are able to hold a spontaneous conversation
with a computer, you may have trouble NOT believing that the computer
is intelligent and capable of reasoning.