facial expression. the interesting thing was how people
responded to Kismet, talking, coaxing, and mirroring its
facial expressions. they behaved as though Kismet was
alive and had feelings.
But could things really matter to Cog or Kismet? this
depends what you think about real mattering and real
suffering: are they special biological or human attributes
forever denied to machines, or are they just more of the
kind of thing these primitive robots already have (Dennett
1998b)?
2 Anthropomimetic robots
Rather than imitating the appearance of a human, owen
Holland’s CRonos robot imitates aspects of the function-
ing of a human body in ‘a strongly embodied approach to
machine consciousness’ (Holland, 2007). Holland wanted
to embody in a robot the kinds of inner representations
that would allow it to interact with the real world in
human-like ways. CRonos is an upper-body robot with
plastic bone-like structures, elastic tendons, and functional
joints, making it wobbly and hard to control, but it allowed
its creators to test various internal models of world and
self that, they argue, provide insight into human con-
sciousness and self-consciousness. CRonos was subse-
quently developed into the more ambitious eCCe robot,
which continues the search for human-like mentality in an
anthropomimetic machine.
3 Disembodied simulation
existing entirely as programs inside computer systems,
Rodney Cotterill’s CyberChild is a simulation of a baby
based on the known circuitry of mammalian nervous
systems. With just two senses, hearing and touch, Cyber-
Child can feel pain when its simulated stomach is empty
or its nappy unchanged, and can feed, cry, and control
its simulated bladder. In other words, it’s a crying baby
simulated in silicon. Cotterill hopes to be able to infer the
presence of consciousness from its behaviour, and even to
discover the nCCs this way. He claims ‘there appears to
be no fundamental reason why consciousness could not
ultimately develop and be observed’ (2003, p. 31). not
everyone would agree.
it to perform’ (in Taylor, 1843, p. 772f ) This suggests
that the machine could not be creative, and the
same argument has often been applied to modern
computers: ‘Working in a fully automated mode,
[computers] cannot exhibit creativity, emotions,
or free will. A computer, like a washing machine, is a
slave operated by its components’ (Buttazzo, 2001,
p. 26). But this argument seems less and less appli-
cable as time goes by. Computers can already write
poems, essays, and screenplays, make pictures and
compose music. Some do this by simple algorithms
combining ready-made segments, some use neural
networks and parallel processing, and some use
genetic or evolutionary algorithms.
These are computer versions of the familiar evolu-
tionary algorithm (Chapter 10): they 1) take a seg-
ment of computer code or program, 2) copy it with
variations, 3) select from the variants according to
some specified outcome, and 4) take the selected
variant and repeat the process. Is this real creativ-
ity or only as-if creativity? That rather depends on
what you think real creativity is. If you think that real
creativity requires some special power of conscious-
ness, then perhaps machines are not really creative.
But what if human creativity depends on the evolu-
tionary algorithm, using exactly the same processes
as those described above? This would mean the
copying, selection, and recombination of old memes
to make new ones. In that case, biological creativity,
human creativity, and machine creativity would all
be examples of the same evolutionary process in
operation and none would be more real than the
others (Blackmore, 2007b).A variant on this argument is what Turing calls the
‘mathematical objection’. There are some things that
machines cannot do, so if we can do any of them that
proves we have something extra – consciousness.
As we have seen (Concept 12.1), there are some
functions that are noncomputable, meaning there
are questions a machine could never answer cor-
rectly, however much time it is given (Turing, 1950).
Penrose (1989, 1994a) claims that mathematicians
can intuitively see noncomputable truths, and this
real understanding requires conscious awareness. So
consciousness itself must be beyond computation.
This is why he thinks we need an entirely new kind
of physics to understand consciousness, and so proposes the theory of objective
reduction in the microtubules (Chapter 5).