When I was doing some research on AI some years ago it struck me that the
researchers in the field were trying to develop some sort of AI that could
be somewhat friendly.  As I got to know them I saw that many were rather shy
and were more at home with machines than people.  Somewhere along the way I
came to the conclusion that many were trying to develop friendly predictable
software that could take the place of those oh so difficult and hard to
control  and understand human relationships.  

 

The other thing I noticed about AI development was that as each program was
introduced, people would say that is OK but it really isn't AI.  It seemed
that no matter how sophisticated the program, people would say "but that
really isn't AI".  AI was still something "out there".  If it could be
developed and marketed it was by definition not AI, yet.

 

Re: robots.  We wanted to develop flying machines but we didn't develop
giant machines with flapping wings a la birds.  We wanted something for a
specific function: to fly.  And we did.

 

Ditto robotics.  Lots and lots of robotic devices that do one or a few
things but do them well.  Will they ever mimic human beings?  Why should
they?  As long as they do the thing they were designed to do and do it well,
over and over again without error.

 

Arthur

 

 

From: [email protected]
[mailto:[email protected]] On Behalf Of Keith Hudson
Sent: Wednesday, July 14, 2010 5:33 AM
To: RE-DESIGNING WORK, INCOME DISTRIBUTION, EDUCATION
Subject: [Futurework] Automation

 

Here's something that falls well within FW's brief. I have for a long time
been extremely sceptical about the dreams of humanoid robots being the next
big breakthrough consumer good. The following article confirms this. If
these ever happen (which I doubt) they're still decades away. 

But, having worked in industry for many years and seen the introduction of
much automation I've been expecting a great deal more development of
automation than has seemed to have happened so far.

However, having said that, it's also a fact that the automation of very
simple repetitive tasks continues apace and some quite trivial operations
can have immense consequences, both in manpower and cost terms but also
accidentally. For example, during my lifetime the trivial task of loading
and unloading goods on and off ships and other vehicles has been transformed
by simple standardized freight containers, reducing dockside manpower at
least 100-fold, long distance costs by at least 20-fold but also meaning
that, for example, the whole world's need for toothbrushes is now
efficiently made and delivered from just one town in China. 

Fifty years ago when I was building an extension to my house, sand was
delivered by two men in a truck who dumped it on the pavement, and then
spaded it into my territory -- a two-man, two-hour job. Recently, after
moving house and wanting a ton of lawn sand, a truck with one driver came
along, and then used his crane to lift a jolly-bag right over my garden wall
to a place 30ft away where I wanted it -- a five-minute job, if that.

Even if we'll never have androids looking after all of our intimate needs
(most of them anyway!), we haven't seen the last of automation yet. The
following is in today's Independent.

Keith

-----  

ELECTRIC DREAMS: IS IT THE END FOR ROBOT DEVELOPMENT? 

Michael Fitzpatrick

We were promised a life of leisure thanks to hard-working robots and
fiendishly clever cyborgs. But the android fantasy has largely been
terminated, argues Michael Fitzpatrick





_______________________________________________
Futurework mailing list
[email protected]
https://lists.uwaterloo.ca/mailman/listinfo/futurework

Reply via email to