Page 110 - Modern Robotics Building Versatile Macines

90 Modern Robotics

Nevertheless, the layers worked together in interesting ways. The
result would be that the robot could explore a room, avoiding both
fixed and moving obstacles, and appear to search “purposefully”
for things.
In October 1985, Brooks presented his robot to an interna-
tional robotics research symposium that was attended by many
of the world’s foremost robot designers. Brooks’s robot, Allen,
startled observers by its seemingly intelligent navigation and
exploration. When they realized that the robot had no “cognition
box”—no AI brain in the traditional sense—many researchers
in the audience were dismayed. (In Flesh and Machines, Brooks
recalled learning that two had whispered to each other “Why

PARALLELS: ARTIFICIAL LIFE AND ARTIFICIAL INTELLIGENCE

The field of artificial intelligence (AI) owes most of its beginnings
perhaps to the work of Alan Turing (1912–54), a mathematician
and pioneering computer scientist who speculated about the ulti-
mate capabilities of computers while the machines were still in their
infancy. In 1950, Turing proposed the famous “Turing Test,” which
basically suggested that a computer could be considered intelligent
if its conversational output could not be distinguished from that of a
human under controlled conditions.
In 1956, a seminal conference at Dartmouth College laid out the
key problems and objectives of artificial intelligence, raising issues that
are still at the heart of the field today. Early approaches such as those
by Marvin Minsky (1927– ) and John McCarthy (1927– ) focused on
developing artificial reasoning and problem-solving capabilities as well
as finding ways to encode knowledge so it could be accessed and used
automatically.
An alternative was the “bottom-up” approach that tries to generate
sophisticated behavior from simple interactions. The earliest example was
the neural network (which was also refined by Minsky), where process-
ing elements are arranged in a network resembling that found in the
nervous system of an organism. The system is then given a problem (such
as recognizing an image), and those elements that respond correctly are