June 25, 2012 12:45 PSP Book - 9in x 6in 09-Junichi-Takeno-c09
198 New Architecture of Robot Consciousness and the Robot Mind
is a manifestation of a thought, and as such, there can be another
thought to deny it. Discussions in this regard are not over. The author
currently holds the same views as Alexander.
9.9 Problem of Free Will
The problem of free will was mentioned in Chapter 4. The topic is
discussed in this chapter in detail while referring to the model of the
consciousness system.
Research on human free will by Benjamin Libet, professor
emeritus at the University of California, San Francisco (UCSF), and
other researchers highlighted the theme of free will (Libet, 2004).
These research studies doubt the common thinking that humans act
based on their own free will. New scientific evidence found by the
researchers is that the brain is activated about 0.55 sec before a
human arbitrarily decides on and performs a behavior. Humans are
unaware of the fact that activities regarding a behavior have already
begun in the brain, and they become aware just when the behavior
is about to be performed.
When humans think, they become aware of the content of the
thought in some cases. For this reason, there are two types of
cerebral activities by humans: those noticed and those not noticed.
Cerebral activities occurring before one becomes aware may be
an integrated process of what has been learned empirically all
of one’s life since one was born and the functions that one had
already possessed when one was born. The latter refers to the
innate functions acquired in the evolution process of life since life
appeared on the Earth. Because these functions are built into and
belong to individuals, the cerebral activities are performed with self-
responsibility, irrespective of whether one is aware of them or not.
This is the basic idea underlying the thought of free will.
Free will means not to receive instructions from others and that
all decisions are made by oneself. On the other hand, when one
becomes aware of the behavior selected by the cerebral activities,
one is given the opportunity to scrutinize the proposed behavior
and then perform the behavior. It can be said that one is acting with
self-responsibility. We cannot be aware of the scrutiny activity of the
brain, but we have the option not to do the behavior once it surfaces
June 25, 2012 12:45 PSP Book - 9in x 6in 09-Junichi-Takeno-c09
Problem of Free Will 199
to our consciousness. This may be called free will. Wrist-flexing is
not a very common social activity; so it may be difficult to use it to
discuss the importance of suppressing behavior as it relates to free
will. If wrist-flexing were replaced by the gesture of aiming a gun at
a person, for example, the role of free will to suppress a behavior
would be more keenly appreciated.
In later research, Libet found that an awareness of conscious
will occurs about 0.15 to 0.2 sec prior to the physical action, and
just at that moment, one can select between stopping the behavior
or going ahead with it. Libet formulated a hypothesis that the
possibility of this selection is the entity of human will. The brain is
always ready to perform behaviors, but consciousness controls the
execution of behaviors. In short, humans do have free will and are
held responsible for their actions.
The author’s model does not specifically use control signals
for suppressing the execution of a behavior. However, the author’s
consciousness system can execute suppressed behaviors without
jeopardizing the consistency of cognition and behavior (Suzuki et al.,
2005; Takeno et al., 2005) Therefore, it is possible to suppress
behaviors using signals to execute or not execute. A suppressed
behavior means the act of imagining the behavior. A process
seems to exist in the human brain in which the brain imagines a
behavior just before the behavior is executed and determines the
possibility of releasing the suppression by evaluating the condition
of representations firing at that time.
It is possible for the consciousness system proposed by the
author to have a similar function to the human mechanism of free
will.
The problem here is the relationship between will and respon-
sibility, i.e., who is responsible for the behaviors performed by
humans? Generally, if a behavior is performed with awareness, the
performer is responsible for the behavior. If, on the other hand,
a behavior is performed without awareness or unconsciously, the
performer may not be held responsible. And even if the unconscious
performer is held responsible, the responsibility seems to be lighter
than that assumed for a conscious performer. If a robot has the
function of will, can the above concept be applied also to the robot?
The problem is, who is responsible for the behavior of a robot? If we
June 25, 2012 12:45 PSP Book - 9in x 6in 09-Junichi-Takeno-c09
200 New Architecture of Robot Consciousness and the Robot Mind
make the robot responsible, is there any practical meaning to stating
that the robot accepts the blame? If a robot has a self, should it be
responsible for its own behavior like a human? The author believes
that although robots may have nearly the same functions as humans,
the maker of the robot should be basically held responsible for the
conduct of the robot, at least for the time being. This would be much
like the case of an automobile, in which the owner or driver is held
responsible, and various legal regulations should be established to
ensure the safe use and operation of robots. Robot owners should
also be subject to various legal controls to promote the safe use of
robots. We should do our best to ensure the safety of robots so that
conscious robots can work in society for the benefit of humans.
9.10 Summary and Observations
This chapter introduced the word’s first self-aware robot capable
of becoming conscious of itself reflected in a mirror. The scheme
of consciousness is made up of recurrent neural networks. The
basic idea is as follows: (1) The process for the robot to cognize
information coming from sensors and the process to execute
behaviors share a common neuron group K, and (2) the common
neuron group K is represented by neural circuitry. Another feature
of the author’s robot is that behavior information is returned to
the common neuron group K via somatosensory nerves. The robot
thus cognizes information about the external world supplied via the
sensors and information on somatic sensation, i.e., both external and
internal information, as representations, simultaneously and at all
times. This means that the robot distinguishes between the self and
others. This scheme of consciousness has been demonstrated to be
capable of learning and performing imitation behavior.
Using this scheme of consciousness, the author has constructed
emotion, feelings, and association areas in the consciousness
system of the robot. The structure, though simple at this stage of
development, shows the possibility of the robot’s consciousness in-
teracting with emotion and feelings. Subconsciousness and explicit
consciousness have been described. Lastly, the sense of a robot,
June 25, 2012 12:45 PSP Book - 9in x 6in 09-Junichi-Takeno-c09
Summary and Observations 201
i.e., the problem of qualia, free will, and problems of responsibility
related to free will, have also been discussed.
Human consciousness has not yet been completely defined, but
the author’s scheme satisfies 90% of Husserl’s 10 properties of
consciousness. As presented in the online version of the Discovery
Channel (www.discovery.com), the scheme of consciousness devel-
oped by the author is a solution to the problem of mirror image
cognition, i.e., becoming aware of the self reflected in a mirror.
This page intentionally left blankThis page intentionally left blank
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.