Path: utzoo!utgpu!cunews!micor!isishq!testsys!doug
From: isishq!testsys!doug (Doug Thompson)
Newsgroups: comp.society.futures
Subject: Re: Thinking Machines
Message-ID: <587861793DN5.42B@testsys.uucp>
Date: Thu, 06 Dec 90 15:33:35 EST
References: <9^}^-!+@rpi.edu> <1990Nov30.145228.21484@abcfd20.larc.nasa.gov> <574426895DN5.41B@testsys.uucp> <1990Dec05.072022.15170@kithrup.COM>
Distribution: na
Organization: SKAN Communications - Ottawa
Lines: 155

In article <1990Dec05.072022.15170@kithrup.COM> (Sean Eric Fagan) writes: 

> In article <574426895DN5.41B@testsys.uucp>  writes:
> >In article <1990Nov30.145228.21484@abcfd20.larc.nasa.gov> (John Burton) writes:
> >The highly organized and mechanized food industry extends
> >our ability to obtain nourishment. Later, people deprived of a
> >super-market can no longer feed themselves from the land - they have
> >lost the ability to recognize edible plants and trap edible animals.
> 
> Bull.  This is like saying that an Australian Aborigine has lost the ability
> to recognize edible plants in Siberia.  

No - the Australian cannot have 'lost' an ability his culture never
possessed - Siberian plant recognition.  However, the urban Aborigine
who no longer can recognize wild foods his grandmother ate can be said
to have 'lost' something that would likely have been hers had the new
food technology not been introduced into the environment. 

> The supermarket is part of our
> environment, and humans are not born with "the ability to recognize edible
> plants and trap edible animals":  it's learned.  Said Aborigine would not be
> able to survive in certain parts of *our* environment.

Nobody said anything about being 'born' with any of the abilities or
cultural patterns that technologies tend to alter, aside from being
born into a culture which makes certain kinds of knowledge accessible,
and makes certain other kinds of knowledge less accessible.

> Your other points are somewhat well taken, but you need to be careful with
> how far you extend your analogies.

Well, we *have* lost our ability to recognize edible plants in our
environment, outside of the supermarket, an ability our ancestors
possessed.  Of course it is a learned, cultural ability, not an
innate, genetic sort of thing.  But thinking is pretty much cultural
too, and not genetic. I doubt that our genetics have changed much in
500 years, but our thinking sure has!

(Of course homo-sapiens' thinking ability has something to do with
genetics - but it is not the genetic variables that I'm concerned
about here, just the environmental/cultural ones, assuming a constant
in the genetics)

If the supermarket weren't there, we'd have other food-aquiring
abilities, or we'd starve.  The point is, that when you add a
technology to the environment, like a supermarket, cultures change.
New abilities are learned and old ones forgotten.  Whether this is
good or bad is another matter.  That it *is* is the only point I was
trying to make. 

But now that you've got me started :-) . . .
another one is suggested though: cultures also choose which new
technologies they are going to embrace, and which ones to neglect. We
embrace nuclear energy and tend to neglect wind and solar power, for
example. That is a social and cultural decision into which a certain
amount of thought is invested.

The 'logic' of the decision is not absolute. There are pros and cons
to each alternative. How do you weight the 'pro' of cheap electricity
against the 'con' of possible catastrophic environmental degradation?
How do you weigh the 'pro' of decentralized ownership and management
against the 'con' of difficulty exercising central authority?

Well you can't come up with a genuinely 'logical' decision. Instead
you make a 'judgement' based on guesses about what will happen, but
mostly based on judgements about what is important, or what is good.
Is it good to have a broadly decentralized power system that is hard
to control and possibly impossible to closely regulate, with the
profits spread widely within the society? Or is it good to have a very
tightly controlled, centralized system with the profits accruing to a
much smaller number of people?

It depends on who you ask! As with most collective decisions, one
choice is good for some people, and the other choice is good for
others, and the choice made depends on which group has the power to
decide. 

There is such a thing as collective thought as well as individual
thought. Presumably all readers of this conference engage in something
that can be called 'thinking' when reading and posting. Presumably
most participants' 'thought' is influenced by what they read - the
thoughts of others. That is a collective thinking process.

Now let us think about a thinking machine that could participate in a
collective thought process like this one along side humans - and let us
think about a collective of thinking machines that could participate
in their own collective thinking process without humans.

When a society of people makes a choice, such as choosing nuclear
power over solar power, there is lots of discussion, there are
reasons, and there is power-broking. The opinion of some humans has
more weight in the decision than the opinion of other humans. One
person's 'logical reasons' are not the same as another person's. If we
add thinking machines into the process, how much power is our society
going to choose to give them? How will thinking machines rate the
dangers of, say, radioactive contamination if they know it will only
affect people, and not affect machines?  From the machine's point of
view, plants, animals, people, even oxygen in the air may be
considered 'contaminants' in an otherwise sterile cosmos - and the
machine might find the sterility more 'healthy', and thus decide to
dispose of the biosphere altogether.

I suspect people will only advocate the participation of machines so
long as they individually think that the machines will in fact give
them more power personally, and that as soon as the machine threatens
to assume a dominant role in a social/cultural/political
decision-making community, the people will gang up and drive it out,
if they can.

When machines are seen only as servants with no autonomous goals of
their own, they are not seen as a threat, and are given great
influence. As soon as they should articulate goals at odds with those
of powerful people, there will be conflict. Who will win?

Some AI people have been promising machines as smart as a man for a
long time, without a great deal of progress, in my view. But to take a
theme from Brad Templeton's recent posting, suppose we really succeed
in making a machine that is just as good as a man. Will that
machine-man be entitled to vote? What would we do with such a machine-man.
We already have millions of men and women around the world who are
underfed and under-employed. What benefit would accrue from adding a
few more mechanical ones to increase the unemployment queues? And most
important of all, to *whom* would such a benefit accrue!

One poster (name escapes me right now) postulated thinking machines
built into un-manned tanks. Now *there* is power! No one can forget
the sight in Tienanmen Square last year of a tank column stopped
briefly by a single defiant protester standing in their way. Would a
'thinking tank' have such qualms? Would such a thinking machine change
the way that protesters in Tienanmen Square 'think'?

Whose orders would such a tank obey? Yours? Mine? And what if they
just decided they'd only obey their own?

So if we add a thinking machine technology, the way we think (not our
genetic thinking aparatus) is likely to change, just like when we add
a food-delivery technology, the way we think about food, and how to
get it, changes.

Any technology which changes the way a culture accomplishes certain
important tasks, such as food-distribution or communication or problem
solving, will change the culture, and therefore the people within the
culture, over time.

A thinking machine would, I presume, change the way we solve at least
some kinds of problems, and thus the way we think about problems, and
ourselves. The machine would not, I think, be very much like a man or
woman at all, even it could mimic human behaviour in certain regards
and surpass it in others.

If such machines become very powerful and very useful, the control of
such machines could become a major political issue. If nothing else,
that will certainly change the way we think about a lot of things.

=Doug
---

isishq!testsys!doug