SEUL: Summary of Scientific American Article

[John Munoz <jmunoz@nospam.mho.net>]

It is hard to do something like this justice, but here is a stab at a
short summary.

Scientific American: Stupid Computer Tricks: How Virtual Reality, Speech
Recognition and Other Good Ideas can Hurt Business.  Summary of Article
in July 97 issue by W. Wayt Gibbs.
        Everyone is predicting and some are working on producing the
next
generation of user interface.  At a Silicon Valley expo in march,
visionaries unveiled their working prototypes, many of which were way
"Cool".  
        Realists have no product to demo, but ask:  If processor power
doubles
every 18 months, why do corporations spend 43 percent of their capital
budgets (213 Billion) on computer hardware?  Why are corporations
spending 500 billion total (adding networks, software, and support) in
the US and 1 Trillion worldwide?(1)
        The aim of a capitalist corporation is to decrease cost by
lowering
labor and overhead needed to produce the product or service they sell,
and to raise the number and price of the products they sell.  The
effectiveness of corporations is measured by a productivity growth
index, which hass fallen over the past 30 years (since after WWII). 
Productivity growth indexes in the US showed an improvement of 4.5%
annually in 1960, which has dropped to 1.5% in recent years.  Most of
the economic groth of the '90s can be attributed to high employment
rates, trade, and production capacity(I think they mean high use of
existing plant).  There are exceptions.  Telecommunications companies
gained 7% annually from '73-'83.  "Highly routine tasks ..were easy to
automate" (2).  The productivity gains that can be attributed to IT
investment come in the transaction-proccessing business functions:
moving trades, processing checks, p-rocessing orders.
        There are many explanations for this apparent lack of
productivity
gains in the countries investing heavily in IT.  One is that Economists
metrics fail to quantify the benefits of many service improvements.  A
counterpoint is that laptops, cellular phones etc allow people to work
longer, and result in more unreported work time.  Considering the
cumulative or exponential nature of productivity growth, even a tiny
error would rapidly accumulate resulting in visible mismatch with the
GNP's of the affected Nations.  Another defense of IT spending argues
that US industry only has 2-5% of its capitol stock invested in
hardware, or about 12% capitol stock in hardware, LANs, software, and
support.(3).  Another defense of IT investment is : Electric motors did
not improve productivity growth till 40 years after Edison installed the
first dynamo in 1881.  In 1919, half of American plants were wired for
electrical power.  Perhaps the same will be true of computer technology. 
        Hospitals are one example of an industry that does not seem to
have
benefitted from computerization.  In his book "The Squandered Computer",
Strausman says that in 1968, hospitals employed 435,000 administrative
staff to serve 1.4 million patients daily.  In 1992, it took 1.2 million
admin staff to serve  853,000 daily patients.  He attributes part of
this decrease in efficiency to computers.
        A typical corporate PC desktop costs about 3000 US$, or 1,000
US$ a
year over it's practical lifespan.  Gartner Group (Stamford, Conn)
research indicates annual cost is 13,000 US$ annually.  This seems like
alot, so here it is broken down:  A LAN card, wiring and Standard
software account for 1,730 US$, Technical Support costs 3,510 US$, and
the LAN itself and technicians to support it add 1,170 to this total. 
The biggest expense is the time an employee spends "futzing"
(non-productivity enhancing tweaking like color scheme, etc) with the
computer rather than working with it, which is estimated at 5,590 US$. 
Games, waiting for tech support and programs to start ranked high,
accounting for 5 hours a week per computer, and totalling an estimated
100 Billion a year of lost productivity.  Boeing removed games from MS
Windoze PC's, and Sun banned managers from using presentation software
to produce meeting slides for presentations.
        Steven Roach, cheif economist at Morgan Stanley estimates 80 to
90
percent of the 213 Billion US IT budget goes to replace obsolete IT each
year.  If Computers lasted more than three or four years, the savings
could be enormous.
        Creeping Featurism is a second area that is taxing computer
productivity.  The biggest single user problem is mapping their goal to
the functions in the program.  As if that were not bad by itself, these
features bloat the software and accelerate the obsolescence of computers
only a few years old.  For example, Microsoft's Word family of word
processors went from 311 commands (2.0c) in 1992 to 1,033 commands (Word
97) in 1997.  The article blames this on trade periodicals and
advertising that emphasize checklist comparisons between applications.
        Nolan Norton Research (Melbourne Australia) over four years
shows that
employees take 4 to 10 percent of their time to help co-workers solve
computer problems.This adds 10,500 US$ to the cost of a pc annually. 
This cost varies according to the extent that the interface accomodated
the users work patterns. 
        The Win95 interface team studied win3.1 users and found typical
users
had trouble managing overlapping windows, and nearly half avoided using
more than one program at a time (probably to avoid GPF's)  A large
fraction were bewildered by the file manager tool, and many had trouble
double-clicking the mouse (Duh?).  the average beginner took 10 minutes
to open a program if it was not visible.
        The few real success stories have one thing in common:  They
thoroughly
studied the work flow of their intended audience, and made a specialized
OS to match it as closely as possible.  The New York Stock Exchange and
American Express are two companies that benefitted from this approach.
These real success stories succeded without any fancy new gizmos: none
of them used voice recognition, 3-d graphics, etc.  What they did do was
spend lots of time modeling work flow and testing and re-working the
specification till the pproduct was right.  The NYSE volume of trading
has doubled, and error rates have fallen ten-fold.  Am-Ex's bank
authorization procedure training times fell from 12 hours to two, and
task times fell from 17 minutes to four.
        Donald Norman (4) suggests that specialization can help achieve
a
"better match between peoples software and their jobs".  Silicon
Graphics chief scientist William Buxton agrees, saying PC telephone
convergence is "totally wrong.""  We need to diverge".
        Wayt Gibbs(5) says "the Web has lowered software distribution
costs to
virtually nil.  Many companies hope the Net will similarly slash their
distribution, transaction and marketing expenses."  The expansion of the
internet onto every desktop and into every business presents many more
opportunities for futzing.  One way companies are fighting this is by
designing their own internal interfaces to their intranets.  Companies
are trying to design a Simple interface to their internal work.  This is
along the same lines as Netscape Communicator's channels.  Business
wants a customized interface designed around their internal workflow
with minimum opportunity for futzing.
 
------------------------------------------------------------------------

1) expenditures according to Paul A Straussmann, chairperson of Method
Software, former cheif Info. Officer of Xerox and the Pentagon. 
2)Tom Landauer, former Bell core cognitive scientist now at U of
Colorado.
3)Daniel E. Sichel of the Federal Reserve Board.
4)Former vice president of Research at Apple
5)Staff Writer, Scientific American
-- 
John Munoz:  jmunoz at mho.net. Remove nospam from Reply address, or
change at to @  No unsolicited ads please send spam, get added here:
alfoster@llv.com glen.quantc.com completefree@hotmail.com Support the
Anti-Spam bill.  Join at http:/www/cauce.org/
----------------------------------------------------------------------------
Simple End User Linux Mailing list
To be removed from this mailing list send a message to majordomo@txcc.net
with the line
unsubscribe seul-project
in the body of the letter.
----------------------------------------------------------------------------