Artificial Intelligence: The Human, the Multiple, and the Cartel

As Artificial Intelligence (AI) evolves rapidly, it is critical to evaluate human and machine domains and thereby formulate a vision of unique ethical, psychological, and practical human space. In this article, I argue that there is a set of three core principles with twenty human properties that will ultimately guarantee human control of AI despite its rapid evolution as evidenced in particular by generative text engines (such as ChatGPT). In particular, these properties include creativity, innovation, and critical thinking. Moreover, it is unlikely that an AI singleton will emerge in terms of a global dominant entity (however conceived) due at least to nine counter-factors resting on compelling arguments. Rather, AI is likely to develop as a multiple of large differentiated systems and even a cartel of the same, although there is a risk of an autonomous AI system becoming uncontrollable. Furthermore, an ethical dimension is integral to every human stance and act and this is reflected in the creation and development of AI: it is not neutral. It is essential to identify the ethical side of AI on all levels from the institution to the state and to devise and implement regulatory systems of oversight. Where opposing state ideologies are found, conflicting sets of ethical values will be built into AI thus leading to incompatibility and friction across borders. Then, the AI-human relationship is subject to three principal areas of challenge: ethical monitoring and regulation, ethical human control, and explainability, transparency, and control. Dynamic regulation is desirable since AI is continuously and rapidly evolving, although regulation is typically slow. Ethical human control is essential but open to interpretation, and without explainability and transparency, control will be poor. The underlying conditions for the successful management and ethical control of AI are human critical thinking and creativity.