Consciousness

(Tuis.) #1

Chapter


Twelve


The evolution of machines


HUmAnoID RoBots AnD
sImULAtIons
Robots that look and move like people have
been built for fun, companionship, and
domestic tasks as well as for research. sony’s
Qrio and WowWee’s Robosapien were meant
for entertainment, while Honda’s Asimo and
the ‘social robot’ nadine were intended to
help people and ultimately to act as carers.
Able to walk and carry things, they have lim-
ited cognitive abilities, including recognising
people and remembering and responding
to speech, but make no claims to subjective
experience. Projects that copy aspects of
human functioning to investigate conscious-
ness take three main approaches.

1 Humanoid robots
the ambitious Cog project, conceived in 1993
by Brooks and colleagues at mIt (Brooks
et al., 1998), aimed to learn about human
cognition by trying to implement it. Consist-
ing of a humanoid body with arms and a head, it had
dozens of motors, a core of hundreds of connected PCs,
moveable eyes, and integrated auditory, vestibular, and
tactile sensory systems.
Built on the principles of embodied cognition, Cog was
given no detailed internal representations of the world
but learned through the coupling between its own actions
and perceptions. Its basic social skills included finding
faces and eyes, sharing attention with a human, following
pointing gestures, and imitating head nods. An original
aim, that Cog would acquire the cognitive abilities of a
young child, was never achieved, but there were many
surprises, including the way its carers started treating
Cog as if it mattered what they did to it. this was even
more obvious with the ‘sociable robot’ Kismet, which
was just a moveable head with large eyes and moving
red lips, with simple routines designed to be cheap, fast,
and just adequate. It could move towards or away from
things to suit its cameras, giving the impression of being
interested in things and people, and had three ‘mood’
variables expressed by making sounds and changing its

C


on


C


e


P


t


12.2


other, but he denies being a property dualist or
indeed any kind of dualist (2002). He stresses that
although consciousness is causally reducible to its
neurological base, it is ontologically distinct from
the brain in the sense that it must be experienced
(Chapter 17). He explains that ‘biological brains have
a remarkable biological capacity to produce experi-
ences, and these experiences only exist when they
are felt by some human or animal agent’ (1997, p.
212). Even so, Searle does not claim that brain tissue
is necessary for consciousness. He argues that other
systems could be conscious too, but only if they had
equivalent causal powers to those of the brain. How-
ever, he does not say what those causal powers are.


MACHINES WILL NEVER DO X


There are some things that no machine can
possibly do because those things require the
power of consciousness.

Turing (1950, p. 447) offers a selection of things said
to be impossible for a machine:


Be kind, resourceful, beautiful, friendly [. . .] have
initiative, have a sense of humor, tell right from
wrong, make mistakes [. . .] fall in love, enjoy
strawberries and cream [. . .] make someone
fall in love with it, learn from experience [. . .]
use words properly, be the subject of its own
thought [. . .] do something really new.

It is a good list; nearly seventy years later machines
still cannot do most of them. Yet, as Turing points
out, the claim is based on people’s extrapolation
from machines they have actually seen, rather than
any principled reason why machines could not do
such things. People too easily jump to several con-
clusions: first, that machines cannot do X; second,
that because we can do X, we must have something
machines cannot have; and third, that this extra
thing is consciousness.


The last is particularly interesting and relates to
what is often called ‘Lady Lovelace’s objection’.
Ada Lovelace, Lord Byron’s daughter, studied
mathematics, became fascinated by Babbage’s
ideas, and wrote the only full account we have of
his Analytical Engine. She famously said that ‘The
Analytical Engine has no pretensions to originate
anything. It can do whatever we know how to order

Free download pdf