The term, Artificial Intelligence (AI), has changed since it was first coined by John MacCarthy in 1956.Footnote 1 AI, believed to have been created with Kurt Gödel's unprovable computational statements in 1931,Footnote 2 is now called deep learning or machine learning. AI is defined as a computer machine with the ability to make predictions about the future and solve complex tasks, using algorithms.Footnote 3 The AI algorithms are enhanced and become effective with big data capturing the present and the past while still necessarily reflecting human biases into models and equations.Footnote 4 AI is also capable of making choices like humans, mirroring human reasoning.Footnote 5 AI can help robots to efficiently repeat the same labor intensive procedures in factories and can analyze historic and present data efficiently through deep learning, natural language processing, and anomaly detection. Thus, AI covers a spectrum of augmented intelligence relating to prediction, autonomous intelligence relating to decision making, automated intelligence for labor robots, and assisted intelligence for data analysis.Footnote 6
This spectrum, however, will be further expanded with the development of the Artificial General Intelligence (AGI), also known as super-intelligence. The AGI, which is a set of algorithms, learning and developing multiple self-intelligence independently to resolve multiple problem domains, will accelerate the displacement of human labor. Just as Jeremy Rifkin's The End of Work foresees the end of farmers, blue color workers, and service workers due to the First, Second, and Third Industrial Revolutions,Footnote 7 the AGI, after the Fourth Industrial Revolution with AI, robots, biotechnology, nanotechnology, and autonomous vehicles, can be developed to displace general, current human jobs. This issue has been a big conundrum to answer in contemporary society.
This negative denotation of AI displacement of human labor with the Fourth Industrial Revolution will depend upon how we define the nature of the mind because there is more to the mind than the brain.Footnote 8 Among multiple intelligences, based on the Gardner's theory,Footnote 9 there are intelligences highly relying on the mind such as musical intelligence, interpersonal intelligence, intrapersonal intelligence, metacognitive skill, and naturalistic intelligence. To the extent that our world needs and cares about only visual-spatial, linguistic-verbal, and logical-mathematical intelligences in the brain, AGI, so-called superhuman intelligence,Footnote 10 may cause the human era to come to an end.Footnote 11 Furthermore, some scholars suggest even curiosity and creativity, which is relevant to musical and intrapersonal intelligence, can be defined and interpreted in a new way of “connecting previously disconnected patterns in an initially surprising way,” and can be reached and realized by AI.Footnote 12 At a certain point, the “Singularity,” defined as the acceleration of technological progress,Footnote 13 will cause exponential “runaway” reactions beyond any hope of control, surpassing human intelligence and leading to irreversible changes to human civilizations, where most human affairs may not continue.Footnote 14
When seen from a different angle, however, AI can be viewed more optimistically. Robots and machines equipped with AI can coexist harmoniously just as humans’ relationship with animals. They may replace humans in many traditional jobs, but humans also may be able to find new jobs to supervise, educate, and design AI. Especially, when AI machines are given similar moral status as humans, humans will be able to feel empathy toward robots, and the necessity to protect them as a human equivalent will be possible.Footnote 15
In accordance with Moore's Law,Footnote 16 AI has developed exponentially in the last ten years, and noticeable development allows us to predict with more accuracy both positive and negative results from AI. AGI seemed very far from reality when AI was used to solve only computational problems and played games against humans. For example, the AI computer Watson, created by IBM, won first prize in Jeopardy in 2011 and AlphaGo defeated Lee Sedol, human professional Go player in 2016. Nowadays, AI has become realized in more concrete forms, such as androids, autonomous vehicles, autonomous killing machines, and various video avatars, affecting a wide variety of humans’ life, and as a result, human rights and humanitarian law.Footnote 17 AI is not a story of the future anymore; it is the new normal. AI affects contemporary everyday life. AGI is also believed to be fully developed within twenty to fifty years. The current moment is a good time to thoroughly analyze the human rights implications of AI both philosophically and empirically before the AI development passes the Singularity.
So inductively investigating and analyzing what would be the human rights implications of AI or AGI is critical. I organized the AI discussion into four different realms. The four realms are devised according to the level of AI development toward AGI, as well as the human rights implications depending on the level of similarity of AI to humans. These realms may be overlapping between stages of development and timeline.
The four realms are: (1) humans developed AI to benefit themselves (The First Realm); (2) humans experience negative side effects due to AI development (The Second Realm); (3) humans project emotions onto anthropomorphic AI, especially social robots, and feel obligations to protect them (The Third Realm); and (4) finally, AI, especially AGI, start claiming their own rights (The Fourth Realm).
In the first and second realms, AI as passive beneficial objects of human life was heavily discussed. Paternalistic attitudes toward AI remain. Until general international law regulating AI is formed or drafted, democratic accountability for governments’ misuse of AI should be regulated in a uniform way by the current legally binding universal human rights system.
Many international human rights principles will apply to the AI as passive objects, and scholars have thought about what human rights will be affected by the AI development. Various working groups for global governance, including governments, international organizations, and private entities and institutions, produced statements and principles to regulate AI development. The primary goals sought by this regulation are accuracy, transparency, human-centered design, lawful, ethical, and robust safety, privacy and data governance, diversity, non-discrimination and fairness, societal and environmental well-being, and accountability.
The question whether AI is an active subject for human rights and protection has not been successfully resolved among scholars and practitioners in the third and fourth realms of AI discussion. There has been no clear answer to whether AI is a human or human-like and enjoys human rights accordingly. Just as animal rights is a different category from human rights and are provided based on sympathetic motives, AI may be able to enjoy rights as a separate category instead of human rights. However, this idea has the possibility to remove AI discussion from international realm to domestic legal realms. Furthermore, this temporary solution will not last long and needs more clear resolution on whether AI/AGI is entitled to human rights before the Singularity passes. AI has bigger, comprehensive impacts on all humankind, so global cooperation and governance among states, international organizations, and private entities to deal with this AI issue for the humankind is necessary.
