COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	STANFORD UNIVERSITY PRESS RELEASE - by Sandra Blakeslee - Nov. 11, 1977
C00018 ENDMK
C⊗;
STANFORD UNIVERSITY PRESS RELEASE - by Sandra Blakeslee - Nov. 11, 1977


A company which claims it is ready to manufacture and sell truly intelligent
androids, or robots, within the next two years - to the tune of $4,000 apiece -
has the nation's artificial intelligence experts up in arms.

The experts unanimously state that the claims are fraudulent.

The company, Quasar Industries, located in Rutherford, N.J., is currently
touring shopping centers around the country with a model of an android
which they claim comprehends 4,000 words of vocabulary and can vacuum, wash
dishes, and teach the kids French.

Crowds in the sundry department stores are dazzled, according to numerous
press reports.

The robot reputedly answers any questions in English and jests with onlookers,
saying such things as, "My cousin is R2-D2" (a lovable little robot from the
popular film "Star Wars") and (said to a heckler in a store in
Portland, Oregon) "How would you like tire marks on your belt buckle?"

This robot is a fake, according to scientists from Carnegie-Mellon University
in Pittsburgh, PA.  The scientists recently undertook a first-hand investigation
of the matter and found the robot to be a "radio-controlled puppet".

Quasar Industries sees it differently.

"We programmed the robot to approximate human behavior," said Anthony Reichelt,
president of the company. "The only limitations to what it can do is the boundary
of imagination and the cost factor."

Quasar Industries has no relationship with a brand of television set formerly
manufactured by Motorola and now made by a Japanese company.

Numerous publications, including Newsweek, Parade, and Newsday, have in recent
months printed articles about the robot - called Sam Strugglegear - in which
the claims made by Quasar Industries are reported as fact.

"It is a fraud that has gone on far too long," says John McCarthy, director of
Stanford University's Artificial Intelligence Laboratory.  McCarthy is one of the
earliest pioneers in the field.

The technology is nowhere near being able to manufacture a robot which can
do all, let alone safely, what Strugglegear's owners say it can do, McCarthy
said in an interview.

The company has some sort of expensive, remote-controlled gimmick, he said, which
in no way possesses the sort of intelligence claimed for it.

"the press has been extraordinarily irresponsible in not bothering to ask the
opinion of artificial intelligence experts on the matter.  They don't have
to believe us, but they should at least ask.

"If the press can't distinguish between promoters and scientists, then the public
will be cheated again."

Artificial intelligence experts from Carnegie-Mellon, who have taken part in
a $5-million nationwide effort over the last eight years to develop a
machine to analyze speech, share McCarthy's concern.

On Oct. 24, according to CMU graduate students Brian Reid and Mark Fox, a well-known
department store in the heart of Pittsburgh advertised the appearance of a 
"Domestic robot."

"Knowing of CMU's pioneering work, particularly in the field of speech recognition,
various friends called CMU to ask how this robot might be so much better at
speech recognition than our talented and dedicated research team," they wrote
in a memorandum sent to colleagues at Stanford, MIT, and elsewhere.

"Rising to the challenge, four courageous members of our department went downtown
to investigate.  They found a frightening sight: in the men's department,
among the three-piece suits, was a 5'2" image of an aerosol can on wheels,
talking animatedly to the crowd."

Upon closer investigation, the CMU experts saw that the robot was surprisingly
crude.  "It didn't seem able to tell that an object was blocking its path.
Covering the faceplate did not change its behavior at all.  Since the
robot seemed able to navigate around the room without hitting anything, we 
found it quite curious that it had no detectable sensory reactions.

"Feeling more dubious, we began to look around the room for evidence of
remote control.  Lo and behold, about 10 feet from the robot, standing
in the crowd, we found a man in a blue suit with his hand held
contemplatively to his mouth.

"After watching for awhile we noticed that whenever the robot was talking,
the man in the blue suit could be seen muttering into his hand."

Soon after this discovery was made, the robot was taken away to another
suburban department store for an evening demonstration.  A second team
of CMU scientists, not satisfied they had yet learned the truth,
followed.

At the second store, "they found a furtive-looking and rather disagreeable
person loitering in the back of the room.  He was carrying an airline
flight bag with his hand stuck down in the bag."

The investigators challenged the man, suspecting that he was controlling
the movements of Sam Strugglegear.  The man became "very excited and called
for store officials to come get us away from him" the CMU scientists said.

The robot is a reasonably good parlor trick, they concluded, more in the
domain of magicians than computer scientists.

Quasar Industries still maintains they will be ready to market a domestic
robot within 18 to 22 months.  It will, they said, be able to
perform 12 household tasks such as "server your dinner, vacuum your
rugs, babysit for your kids, and insult your enemies."

Such claims amuse, annoy, and sometimes downright anger artificial
intelligence experts.

It is not clear what they expect to gain from their venture, McCarthy said.
Whatever sort of machine they can produce will be so inferior compared with
the expectations they are raising as to make a successful business based
on sales unlikely.

"Perhaps they make money from promotional tours and investors" he said. "Maybe
it's a case of self-delusion"

Stanford graduate student Paul Martin, who is writing a thesis on the problems
of machines understanding natural dialogue, believes much of the gullibility
on behalf of the public, the press, and department store managers stems from 
recent technological innovations, such as hand-held calculators and space
missions.

"Star Wars," viewed as a national phenomenon, also has helped set the stage for
belief in benign machine intelligence.

But there is a huge difference, Martin said, between having a machine which can
recognize separate, simple words (this is possible now) and one that can 
understand the flow of language as spoken naturally.

How far away are we from the real thing?  "Different people believe different
things,"  McCarthy said.  "We are certainly not near the manufacturing
stage.

"Everyone in the field would agree there are important scientific problems
waiting to be solved before we can get to that.

"On how long these problems will take to solve, you find a diversity of
opinion.  I tend to hedge by saying it will take between five and 500 years.
If you ask me to center it, I would say 40 years.

"The basic problems are how to make computers recognize patterns in a noisy
environment ad how to go about representing common sense knowledge about the
world, how to put that in a computer.  In fact, what common sense knowledge
about the world is there?  How can a machine be taught to solve
problems?

"One way these problems express themselves is in safety issues.

"A household robot has to be very reliable, mechanically and intellectually.
It must distinguish between children and animals as objects.  It cannot
think the baby is part of a stack of pillows.  It has to have a wide
knowledge of different kinds of objects to be intelligent."

In response to a reporter's question whether the robot could tell a cup from a
saucer, Reichelt responded that it didn't have to because it didn't take them
off the tray.  McCarthy then argued that, in that case, the Quasar robot
is too limited to be really useful.

"Say it could carry a dinner tray from the kitchen to the dining room,"
McCarthy said.  "Someone in the kitchen would have to prepare the tray.
The robot would be on a tread, according to the company.

"But it still couldn't be done safely.  It could trip on a rug.  It could
run over something.  It could fall down the stairs."

Sam Strugglegear, also called KLATU in some news reports, is not yet out of the
news.

A few weeks ago, while performing at a shopping center in Philadelphia, the
imitation robot reportedly frightened a 10-month-old child.

The child's mother has filed suit for $100,000 in damages, saying that the
robot "accosted, touched, assaulted, an terrorized" her child.

The CMU scientists were reminded by these recent events of another famous
"intelligent" robot built by Baron Wolfgang von Kempeler in 1769.  This
robot, known as "the Turk," would trounce people at chess, all the
while rolling its eyes and nodding its head.  Thousands of people,
including Edgar Allan Poe and Napoleon, examined it thoroughly and could
not figure out how it worked.

It wasn't until more than 70 years later that the truth was discovered.  The
robot cleverly concealed a dwarf who was a talented chess player.

Kempeler's description of his own robot, circa 1771, the CMU investigators said,
is probably the best summary of Sam Strugglegear:

"A mere bagatelle, not without merit in point of mechanism, but whose effects
appear marvelous only from the boldness of conception and the clever choice of
methods adopted for promoting the illusion."