VLSI@DEC-MARLBORO@sri-unix.UUCP (08/09/83)
From: John Redford <VLSI at DEC-MARLBORO>
I don't see increasing technological complexity as all that much of
a problem. Or at least, it's no more of a problem now than it has
always been. Face it, no one really knows the detailed workings of
everything around them. You may understand the innards of a '56
Chevy transmission, but do you understand FM radios? How about air
conditioners? The chemistry of polyethylene? The four-color printing
on the cover of your SF paperback? And how do they get all those
little metal bits on a zipper to line up so well?
No one understands everything, and no one needs to. All you need
to know is the user interface. Radios turn on when you twist this knob.
Zippers close if you pull on this tag. Some interfaces, like those on
personal computers, are in such an embryonic state that they do require
some knowledge of the machine's innards to use it properly. But soon
only specialists (like us) will care what goes on inside. Look at the
computer interface described in Pohl's "Beyond the Blue Event Horizon". The
protagonist never touches a keyboard. Instead, he converses with holograms who
act as counsellors. His science advisor, for instance, is a hologram of
Einstein.
The biological world has already mastered this. You don't need to
know about DNA to get a tomato to grow. All that's necessary is a few
basics and perhaps a couple of simple trouble-shooting procedures, like
if the leaves start to curl up and turn brown, give the plant more water.
If some obscure problem comes up, like your tomatoes are turning blue
and trying to strangle passers-by, then call in a specialist. A
system meant for common use should rarely need uncommon knowledge.
John Redford
DEC-Hudson
--------sts@ssc-vax.UUCP (Stanley T Shebs) (08/11/83)
User interfaces are fine, but I think the original question was about the overall reliability of a highly technological society. Our society is less inherently reliable than it was a century or two centuries ago, and this has become more obvious in the past several years. Consider: although one does not need to know the internal workings of an FM radio to use it, how many people can fix it using only raw materials? If you have replacement capacitors and ICs, it requires only a moderate amount of specialized knowledge, but suppose the radio is old and the parts are no longer available. This has already happened to me once with my personal computer. I had used a fairly obscure memory chip on one board, and, well, one of them failed. Guess what? The part was no longer available. Rather than try to find a supplier, I ended up redesigning part of the board. Now I wonder about those people with 8008's in their products. The situation now is that it is often cheaper to replace black boxes than to repair them, and now many of those black boxes are *designed* to be replaced. Most are not repairable. A highly technological society depends on a massive infrastructure, which in turn depends on the high availability of resources. This infrastructure has become more and more necessary to supply replacement parts, and it is rather fragile. Any kind of extended disaster could cause a massive breakdown (imagine us running completely out of oil *now*. Could everything be kept running long enough to develop alternate energy sources? Can't research solar cells if your lab equipment breaks down and can't be repaired). I'm reminded (sorry to carry on so long) of my grandmother's home town in the middle of Utah. It is basically self-contained, in the sense that anything breaking down can be fixed by someone in town. Scrap metal is saved, and I have seen some real artists fashioning replacement parts from it. Needless to say, they are suspicious of things like computer controls in cars. stan the leprechaun hacker ssc-vax!sts (soon utah-cs)