'Machine Minds: Frontiers in Legal Consciousness' by Evan Joseph Zimmerman
argues that
Research is at the point where we might have to confront the possibility of what computer scientists call “strong AI” in the coming years. A strong AI could be intelligent by most reasonable definitions of the word, and possibly have a subjective experience. It stands to reason that we must seriously consider whether such a machine should have its will recognized, protected, and enabled. That is to say, such a sufficiently advanced machine might be said to carry responsibilities, and rights, on its own, as a legal person.
This question is wrapped up in significant philosophical and technical questions of acceptability and feasibility. The novelty of this Essay is that it provides a positive reason to grant such rights with a basis in technical law, along with a concrete definition and justification for personhood. The organizing principle this Essay arrives at is: Personhood exists to protect conscious individuals from suffering and allow them exercise their wills, subject to their intelligence.
This Essay examines several documents across a variety of fields, as well as historical records. Examples such as corporate law, personhood law, slavery law, and standing law demonstrate that the story of person-hood is a history of grappling with what is fully conscious, and how to allow these consciousness’s to exercise their will and avoid pain. However, because law was made by and for humans, by examining the law surrounding animal welfare and humans in a vegetative state, one finds that the law privileges humanity. Hence, our laws imply that a computer, if it is intelligent enough, should be considered conscious; however, our laws as they are would arbitrarily not provide it personhood simply because it is not made of flesh and blood.
Zimmerman states -
In one episode of that ever-prescient television show, Star Trek: The
Next Generation, the android Data’s right to refuse to be dismantled for
research purposes is questioned. Data is an android that is one-of-a-kind in
its (the crew calls it “his”) intelligence, and is treated by his crew as a living
being with feelings. Despite this, a scientist wants to dissect Data despite
the fact that this could destroy data’s personality, which causes Data
to refuse the procedure. In response, the scientist claims that as a robot,
Data is Star Fleet’s property, and thus has no rights to refuse any procedure.
A court case ensues, in which one side appeals to, essentially, a
form of species solipsism; organic beings are known to be life, and as Data
is an android, he is thus clearly not alive. The opposing side argues that
Data has feelings, and that his consciousness is as difficult to prove as any
other person’s. At stake is the potential destruction of Data’s soul, which
is not likely to survive the procedure. The argument, of course, is over
whether such a thing even exists.
The complexity of computers, particularly with their potential to become
intelligent, raises profound legal questions. Although computers may
not become brilliant overnight, several observers of the field consider such
a thing a serious possibility, and technology has shown itself to be unpredictable. Despite the quip that artificial intelligence is always ten years
away, a breakthrough may be just around the corner, and the law should
be prepared. It is extremely important that this matter is treated wisely,
carefully, and with an eye on the emerging future moving at the speed of
light and imagination. Proper care depends on our ability to abstract beyond
our own experiences, as a being need not think like us for it to be
said that it is thinking. It also requires knowledge—though not necessarily
expertise!—of difficult technical issues.
Those in software circles are fond of saying that “Technology moves
fast, and the law moves slow.” The likes Europe’s Luddites litter the 19th
century. Yet, in the 21st, it seems impossible to avoid the fact that machines
outfitted with advanced computational power form an increasingly
large part of our lives and are changing more quickly than ever: in homes,
in vehicles, at work, in the air. Many in the technology industry believe
that machines have changed in front of our eyes as the law seems to have
hardly noticed. Often the technologies are so complicated that the regulations,
may seem to our tech observer as built on an incomplete or faulty
understanding, inadequate, cumbersome, outdated upon pronouncement,
and obstructionist in the reality. Whether or not this is true, outside of a
few articles in the past two decades, the literature has mostly not addressed
the issue of artificial intelligence and personhood.
Any such judgment(s) will have to hinge on consciousness. The point
of the law is not to protect tools. It is to ensure the protection and allow
the full life of those who can feel, even if, like for animals, in degrees,
and even if those feelings are alien yet almost inconceivably compared to
our own. A different type of being need not think the same way as a human
being for it to be said that it is thinking, and consequently to be deserving
of privileges and protections. Surely it carries obligations and liabilities;
so, why not rights? Consequently, consciousness is the key factor
in determining the legal status of intelligent computers. A recognition of
this requires a willingness to admit that there are serious philosophical
questions that the law may not be able to address, but be forced to consider
anyways. It is important to address these issues long before we have to
scramble to piece together a last-minute solution, while we have years to
consider the question. These are very pressing concerns that address the
core of our legal system, and by highlighting and thinking about them
now, perhaps for the first time in a long time, the law can be prepared for
when technology arrives.
Unlike the previous literature, instead of negating reasons for denying
personhood, it discusses an affirmative reason to grant it, and instead of
approaching it philosophically, this Essay considers the technical legal
standpoint. I examine case law and influential historical conceptions of
personhood, including corporate personhood, and liability to determine
that personhood is intended to privilege consciousness, as it has a will and
can feel pain. By examining case law and statute for such issues as slavery,
women’s suffrage and education, corporate personhood, vegetative
state humans, children, and animals, to conclude that personhood exists in
a tiered form, even if sometimes we dare not speak its name. Furthermore,
popular historical justification is used to suggest that the basis for
such stratification is the level of intelligence, i.e., the complexity and
depth of the subjective experience of, the persons in question.
In Section I, this Essay frames the issue for future scholars considering
such cases by explaining how these sophisticated machines work with a
technology primer. In Section II, this Essay provides working legal definitions
for terms like “intelligence” and “consciousness” that a court could
use without forcing it to take a position in perhaps the most significant and
ancient debate in human society, that is, how to assess consciousness. In
Section III, this Essay examines the historical record, statutes, and case
law to assess what personhood really is, and then whether machines
should be granted personhood, and contrasts my conclusion with the existing
literature. The main thrust of this Essay is ultimately to answer this
question: Could a machine be a legal person, and is the law currently able
to withstand the intellectual challenge of an intelligent machine?
