Application of human rights to artificial intelligence

¶ … Technological Changes & Advancements Affect the Law

by altering the cost of violating and enforcing existing legal rules; (2) by altering the underlying facts that justify legal rules; and (3) by changing the underlying facts implicitly assumed by the law, making existing legal concepts and categories obsolete, even meaningless. The legal system can choose to ignore such changes. Alternatively, it may selectively alter its rules legislatively or via judicial interpretation.

David Friedman in Does Technology Require New Law?

Artificial Intelligence is one of the most controversial and interesting topics in the modern human repituar and within the issue are countless social, emotional, and legal discussions. As artificial intelligence becomes more and more like human intelligence, capable of creating lasting relationships, expressing human emotion and developing influence on social phenomena the controversies about the AIs "existence" as "living" beings becomes far greater. Many are likely to believe that these issues are not currently pressing as a machine is a machine, no matter how much it looks like or acts like a living being and technology may mirror this limited concern as it has yet to create AI that is systematically reflective of the human condition. Yet, technological forays into AI are and will continue to become much more life-like, which will likely create lasting questions and debated answers regarding how laws should be altered to include or exclude AI in the collection of rights most humans consider innate aspects of being. Yet, this becomes the enduring question of extending rights to artificial intelligence, what exactly is "being?" Does "being" or "living" require an exchange of oxygen, fluids, electricity or other elements to characterize it as living or does it allow an exchange of different elements to represent life? Is consciousness the designation of life? It would not seem so given recent debates regarding medical decisions made about people in persistent vegetative states, or other syndromes like it where the medical community is fairly certain the traditional consciousness is not present. www.questiaschool.com/PM.qst?a=o&d=5002137129" (Cranford 27) the questions regarding AI and what constitutes a demarcation of "life" are actually rather complicated and for all intents and purposes theoretical, as AI is a development of human technology, and therefore subject to human debate and theory, not that other life forms are not subject to human superior debate but AI is tailor made to be discussed and debated as a wholly human creation.

This work will discuss the current debate on AI, offer a brief illumination of the state of AI technology and discuss questions regarding what the future may hold for Ai and the need or lack of need for concrete legal designations for AI technology. It is clear that as a theoretical concept AI debates ask more questions than they answer and this will be felt in this paper but the main question, worth discussion is; "What rights should AI being have and based on what characteristics of being?

Some argue that the ultimate test is if the other "being" is controlled by a person and yet there are countless example sin human history where individual humans have been physically and mentally controlled by another and in fact some would argue that the to an alarming degree humans are currently controlled by artificial intelligence. We live by the clock, the computer the electronic bank balance and likely eventually we will live by the desires and standards of much more complicated systems. Multimodal interfaces are the thing of the future and they are specifically designed to allow mimic human interaction and human thought patterns to produce results that are akin to the individual interacting with another individual, not entirely unlike them. The passage below describes the present and future "life" of multimodal interface interaction between computers and humans and a passage within it; "The... system will track and incorporate information from multiple sensors on the user's interface and surrounding physical environment to support intelligent adaptation to the user, task, and usage environment."

Oviatt 287)in other words the multimodal interface will be to some degree in control of the situation, molding the user to its purpose and altering the environment to meet the needs of the user and the AI even if the user, we presume a person does not know there is a need for alteration.

The advent of multimodal interfaces based on recognition of human speech, gaze, gesture, and other natural behavior represents only the beginning of a progression toward computational interfaces capable of relatively human-like sensory perception. Such interfaces eventually will interpret continuous input from a large number of different visual, auditory, and tactile input modes, -which will be recognized as users engage in everyday activities. The same system will track and incorporate information from multiple sensors on the user's interface and surrounding physical environment to support intelligent adaptation to the user, task, and usage environment. Future adaptive multimodalmultisensor interfaces have the potential to support new functionality, to achieve unparalleled robustness, and to perform flexibly as a multifunctional and personalized mobile system.

Oviatt 287)

Builders of AI contend that such usability will greatly add to the human experience and allow the person freedoms to contend to different issues than those controlled by the AI system. They also argue that the more "life-like" these systems become the more capable they will be of mimicking human intelligence and teaching people how to act and be in accordance with their own needs. Many would like to see AI take the form of a system that replaces the need for humans to themselves engage in mundane tasks such as cleaning, nursing and even companionship when none is available to the individual by natural means. These AI of the future will likely challenge what it means to be a "living human being" and this will be more tricky the more they look and act like people. The ultimate goal of the creator being to design and implement a tool that is almost indiscernible from a living breathing being, such as something we might see in the movies. Some argue that this is far from a forgone conclusion and that if and when such creation is possible it must be patterned off something else, in look at least it must be expressive of an entirely new visual representation, yet AI that is capable of thinking acting and learning like a human will likely be personified to such a degree that these questions must not go unanswered.

Legal rules that assume that a brief examination is sufficient to determine whether someone is alive or dead and that the latter status is irreversible might produce unfortunate results in the context of cryonic suspension. Rules that consider as legal persons only human beings or organizations of human beings and take it for granted that persons have the characteristics associated with human beings and human organizations will be wholly unsuited to a world of advanced artificial intelligence, when and if that world arrives. In these cases and many others, new technology requires new law.

Friedman 71)

Laws that restrict the designation of "being," "living," and "existence" as something that must be an entirely carbon-based organic system have yet to be developed, as there has been no fundamental need to do so. Yet, with technology adapting to create systems that are "clones" of living, breathing, organic beings does exist and will ultimately require legal distinction. It is not then difficult to take the leap to an understanding of the fact that non-organic "life-forms" will also require this distinction in the future. This is especially true if such "life-forms" take on the roles and challenges that are inherent to animal or human existence. Who is to say that an AI that learns in the same manner as a child, requires constant feeding from some source and develops cognition of emotion and knowledge that is similar to that of an animal or human does not deserve some protection for the unique manner in which it did all these things and, "survived" doing them?

As a culture we have based a good deal of our ideals and ideas about the human condition on a set of inalienable rights that protect us from encroachment by others, be it by individuals or institutions. Does it not seem logical then that at some time in the near future such debates as those debated during the civil rights movement regarding disenfranchised people, some of whom spent a good deal of their existence in a state of chattel as if they were nothing but livestock, will occur? It is also crucial to understand that the manner in which artificial intelligence is being "discovered" and invented is through a clear and decisive dissection of human intelligence. The scientific community is seeking to fully understand how the human mind (likened to a supercomputer with astronomical functionality) really works to create characteristics such as language and social learning.

Donald, (http://www.cs.dartmouth.edu/~donaldclass/GradAI/noteasy1.gif)

The scientific community currently stresses that the human mind and all of its functions are far from easy to recreate or even mimic in a non-discernible fashion but this may change in the future as so many other things do and have. In a descriptive syllabus for a graduate seminar in AI professor Donald gives insight into the form that AI research is taking and the bottom line is that internal brain functioning, switching, neurotransmissions, and patterns are being dissected to give the student a greater sense of the workings of the human mind so these same students may go forwards and attempt to recreate, decidedly small scale (likely single or minimally multiple) functions of human thought to create action. It is therefore not difficult to imagine that these same students will be a part of the future with regard to the development of life-like interactive toys or tools. (Donald NP) the point being made is that research developed to recreate a person in an artificial form is being done on real human science and characteristics. Science is attempting to map and dissect the whole of human physical and mental potential to both understand it better and create imitations of it for science and potentially to better the human condition. Many of them com at it as if they are designing better toasters, even though it is far greater than that. They are designing elaborate toasters that can learn from humans and can learn to anticipate what type of toast and how dark the individual would like it to be, based not on cues but on an artificial sense of empathy. This is of course a simplistic analogy and in reality mind reading toasters is a still yet more simplified concept than an AI system, as AI being developed today has the hope of science behind it that would have it be conscious, think, learn, emote and even hold viable and lifelike conversations with its human companion.

There may also be a time when, like the movies, research into such systems does not require the support of an institution and can be created and manipulated within "rogue" environments or at the very least environments with little or no legal supervision or control. This is not to say that a conspiracy of AI development currently exists as most if not all of the current research regarding AI is open and debated, with the exception of the fact that within science and even the law there are many protections for entities that wish to keep scientific knowledge a secret for the protection of their own rights, the security of nations and most often for the protection of any future right to profit from their discoveries.

A leading expert of corporate social responsibility reasons that companies are not populated by enough people with universal ideals of social responsibility. This is a statement that many experts agree on, that culpability also lies in the individual, working for any corporation, to speak out when problems are noted and force action. (Lindorff 880) Sims proposes that there needs to be a "universal ethical principle orientation" that he defines thusly:

Right is defined in terms of self-chosen universal principles of good and bad. The individual follows self-selected ethical principles. If there is a conflict between a law and a self-selected ethical principle, the individual reasons and uses conscience and moral rules to guide actions. A person at this stage may be more concerned with social ethical issues and not only rely on the business organization for ethical direction. www.questia.com/PM.qst?a=o&d=106810913" (Sims 103)

The problem then arises when the corporate and scientific climate in which these individuals work does not stand on the ideas of openness and communication about concerns and problems. The system does not reward scientists for speaking out about known problems and whistleblowers are rarely tolerated within the system, as they represent a threat to secure information, that builds profit as well as company public image. There are many cases of this sense of secrecy in the science world, that is clearly contrary to open communication and sharing about science and technology, one of sciences many founding principles.

There is, to begin with, the kind of secrecy that everyone deplores but that is fostered by institutional cultures of self-interest, both public and private -- when scientific facts that the public has a right to know are intentionally hidden and knowingly withheld to preserve the economic or political standing of powerful organizations. Examples include drug companies that fail to disclose reports of adverse reactions to their products, (57) car manufacturers that hide technical defects in their vehicles, (58) employers and polluters who conceal data about illness caused by their activities, (59) and governmental agencies that paper over malfunctions in technologies that are deemed key to the success of their missions. (60) Impediments to criticism and communication have also arisen within universities, the traditional strongholds of scientific openness, as a result of private sponsorship that contractually demands secrecy. (61) www.questia.com/PM.qst?a=o&d=5018256079" (Jasanoff 21)

Jasanoff then goes on to discuss the fact that the legal system support secrecy in science and business as a development of controlled rights to future profitability and limited liability. The scholar stresses that corporations often fund science and therefore control the outcome of information regarding it as a result of commercial confidentiality and in the case of the public sector as what Jasanoff calls an overuse of the function of "national security." (Jasanoff 21) Within the moral fiber of the consumer driven society there must be more exploration of the impact that companies have on the communities they serve and safety must be a paramount concern. Yet, it is clear that in the current system the rule of secrecy to protect image and profit drive the scientific community, as soon as such researchers and developers go on someone else's payroll. Therefore individual culpability can only be a limited answer and solution to the problem. Corporate social responsibility should be the umbrella to all decisions made with regard to client safety and with regard to AI overall principles of right and wrong that stress open communication and legal precedence setting as essential to discovery and implementation. As one can see from the above supported statements the scientific community, built around the idea of open discovery and dialogue is in no need of developing "rogue" conspiracy to develop AI technology as it is primed for secrecy already. The issue of disclosure only really comes to pass if safety is questioned (and people are willing and able to express concerns) and if the systems intention is for the development or release of a consumer product.

The current state of AI is arguably developed, though most will note that the products of such "builds" are arguably inhuman in character the language of science legitimately places the goal of the science of AI in a position to seek life-like capabilities in a robot.

In 1993, after years of research on behavior-based insect robots, Brooks and his team at MIT started to construct a robot shaped like a human. They named it COG, an abbreviation for "cognition," and also the tooth of a gear. COG was designed and built to emulate human thought processes and experience the world as a human. Brooks and his team further assumed that people would find it easier to interact with a robot and aid the robot in its learning process when the robot could respond in a somewhat human way. Consequently, the machine should have limbs, sensory organs, and a physical resemblance to humans. Unlike other artificial intelligence systems, like medical expert systems, COG was meant to test theories of human cognition and developmental psychology. (Ahmad NP) picture of COG does not elicit the idea that COG is in any way demonstrative of a human being, as he looks as he should like a robot, yet already this researcher is capable of taking the leap of pronoun from it to he. The robot is gendered and characteristically personified, as the goal of the MIT experiment was achieved through the very means discussed above, a greater understanding of the workings of the human mind, not the secrecy. Though COG may not necessarily look human those who have interacted with "him" are often struck by the fact that "he" acts a great deal like a human, learning in the same manner as a child.

Ahmad (http://www.acm.org/crossroads/xrds10-2/robotcog.html)

Ahmad goes on to say that technology will eventually be paired with biotechnology which will likely help create much more human looking robots and he then gives examples of other AI humanoid robots, "Honda's Asimo, Sony's SDR-4X, and Kitano Symbiotic Systems' baby robot PINO are some of the other humanoid robots being developed around the world. To date, these are not commercially available," (NP) [PINO is currently available to purchase as a toy] and some products that are already available, that come in the form of electronic pets, (Sony's Aibo) capable of learning from their human owners and carrying out simple tasks for them as well as recharging itself when needed, a functionality which is clearly demonstrative of a traditional human controlled AI function. "Aibo is an autonomous pet dog robot which inhabits a home, sings, dances, reads emails, re-charges, and learns from its owner." (NP) According to Ahmad the; natural progression in technology ensures that humanoid robots will soon follow commercially available pet robots. We may just be on the verge of a robotic revolution in which robots and intelligent autonomous machines become a common part of our daily lives. The day when we start to communicate with them in human language, teach them our daily chores, and share our responsibilities will be when we have achieved the ultimate science fiction goal of making human-like robots. (Ahmad NP)

It must also be stressed that when one is looking at the language and stresses of those who seek to create this collaborative human robot reality there is a clear sense of wonder and a very limited sense of systematic and/or legal control. The progression of human science is logical and formed through the very innocuous intentions of seeking to create toys and/or computer interfaces that help humans, but without complete and full thought of the consequences of the change. Though this may seem like a trite example the issue of dynamite comes to mind, as the development of dynamite was intended as a great discovery to aide mankind in its ability to clear land and therefore extract resources and if used correctly would actually save lives and labor rather than take lives, but to date and even in the lifetime of the inventor dynamite likely took more lives than it saved and often as an instrument of war. www.questiaschool.com/PM.qst?a=o&d=100876772" (Stevenson, and Byerly 154)

In a full length book about the extension of animal rights one scholar discusses COG, and the character surrounding him. Wise discusses one of the biggest detractors of AI as he has been transformed, by interacting with and becoming apart of the COG experiment;

Daniel Dennett, who slashes away at the idea that any nonhuman could be conscious, has been advising a group building a robot named COG at MIT's artificial intelligence laboratory. Dennett claims he gets "an almost overwhelming sense of being in the presence of another conscious observer" when COG's camera eyes track him as he moves about a room. He predicts that COG "will be conscious if we get done all the things we've got written down." 209 www.questiaschool.com/PM.qst?a=o&d=9932383" (Wise 156)

Wise was so fascinated with the turn of Dennet from a detractor to a believer, and with Dennet's belief that COG might someday become conscious, a characteristic currently designating life he made an appointment to meet COG and tour the lab at MIT.

The friendly Ph.D. student who led my group of three about the lab told us that the long-term goal was to make COG "as smart as a six-month-old." Then he showed us YUPPEE, an artificial intelligence "dog," and KISMET, a "robot baby," and videos of all three in action....After listening to him use a raft of words that denote mental states to describe how the three robots were "feeling," such as "likes," "dislikes," "cries," "gets upset," "wants," "lonely," "interested," "bored," and "intimidated," I asked him if he expected COG to achieve consciousness within, say, the next fifty years. He plainly hoped it would. But he shook his head and said that thirty years ago, people thought that artificial intelligence was "just around the corner and we're still not even close."

COG is very clearly a cumbersome and inhuman frame with many needs, as a work in progress Ahmad notes that some of COG's functions have not been utilized to their fullest potential and that COG still needs many more sensory receptors and other technology to be more lifelike, as well as a logical progression to biotechnological "skin." (Ahmad NP) it is also clear that COG was developed in 1993 and will be further developed along side other models in both commercial and academic production, with an emphatic note that commercial production clearly has greater resource allocation (and a higher degree of secrecy) than academic production. There is also a concurrent humanoid robot project in the MIT lab that is dealing specifically with the face the Lazlo face project the description of the project includes the overarching goal of not only putting a face on Lazlo a limbless humanoid robot but one on COG to further generate a sense of human connectiveity, in a platform that is universal to humans, the face;

To date neither Cog nor Lazlo have faces. Our goal is to design and fabricate an iconic, humanoid face for each robot that fosters a suitable social contract between the robot and humans. Another goal of this project is to shift the robot aesthetic to a design language that utilizes strong curvilinear and organic forms through state of the art design processes and materials. With a well designed face, Cog and Lazlo will be able to convey an appropriate social aspect. This will assist them in regulating interaction, receiving appropriate stimulus and, in the longer term, learning imitation tasks. As of this writing, we have fabricated and installed the first prototype of the face on the robot head platform, Lazlo. (Edsinger, O'Reilly, Scassellati, Scarpino and Breazeal NP)

As one can see from pictures of other humanoid Robots there is a clear expression of progress noted in the abilities and the appearances of humanoid robots many of whom have AI characteristics built in and some who are waiting for such technology to meld with their lifelike appearances.

Here two of the more advanced corporate AI robots are shown interacting with humans and responding to human touch.

Honda's ASIMO)

Sony's SDR-4X)

Left: (http://www.hiptechblog.com/2005/12/14/hondas-asimo-robot-that-guides-people-runs-and-serves-coffee/)(Right:(http://androidworld.com/prod02es.htm)

These robots are in production phase and will likely be available for purchase in some form in the near future, if not currently. Likely the most disturbing of all is the image of the techworks studios working real robot, used as a prop in films (or soon to be) but disturbingly lifelike. It is fact looks a bit like a human in a mask. Granted this prototype is utilized as a prop and is not likely to have the characteristics of a human, in learning or other capacities it is difficult not to give the robot a pronoun and a name, given its almost startling human appearance. One can see this robot or ones like it in sci fi films that depict some sequences in form and some in digital animation, yet it is clear that if appearance is the judge realistic moving humanoid robot forms are much closer than it would seem, at least for Hollywood.

As was noted in brief above the toy humanoid robot PINO developed by Kitano Symbiotic Systems is currently available to purchase in the form of a commercial toy. Here is an ad copy of PINO's "characteristics;"

The more you play with him the more he will learn, becoming a fully programmed robot and friend. Pino can be sad, happy, lonely, angry and sleepy. He will let you know by the colour of his visor. If he's sad or lonely cheer him up by playing with him. If he's happy watch him sing and dance. He learns up to 5 games as he grows up and will develop his own personality depending on how well you treat him. Will interact with others of his kind, singing dancing and talking Pino. Sensors help him learn and communicate with all his friends human and robotic. Requires 4 x AA batteries. For ages 8 years and over." ("PINO Your Robot Friend" NP)

It may not seem at all disturbing to provide toys like this on the market for children. As they are just that, toys. It may not seem at all disturbing to provide toys like this on the market for children. As they are just that, toys. Yet this particular toy is a development of AI technology for commercial use that compares to little else. The technology that makes it possible for PINO to learn and develop a personality based on the "treatment" the owner provides it is not the least bit unlike raising a real human child and though at its current state it is unlikely that anyone would think to develop a system of laws that govern its acceptance a "living" being with rights and privileges does not mean that this is not at all possible in the future, or that the children who "train/rear" or care for PINO may not be the next generation of scientists fascinated with making robot toys and tools even more lifelike and with even greater capabilities to learn and grow into sentient beings, with thoughts and emotions that not only mimic human characteristics but actually "live" them. It is also assumed that most of these now grown children will develop a sense of objectification of these AI objects, as a normal aspect of development, yet it this really something we should assume?