How can we make Artificial Intelligence more manageable in light of its History and Future?

Abstract

The contemporary world is fast-paced, and so is the need to adjust appropriately to the issues arising from technological advancements. Gone are the days when man used steam engines to transport products to multiple destinations, so the most powerful AI humans can create based on classical computers in the foreseeable future may also easily slip down to the servant's place with tech evolution for powerful quantum AI coming out. From ancient machines meant to make work easier, there have been tremendous advancements in the era of technology. The end goal may be to design machines with the ability to think like human beings and do tasks intuitively, but it also may unexpectedly come to a being that is beyond human cognition before humans figure out what they are actually doing.

This thesis holds both theoretical and practical significance: questions of how to prevent loss of control and how to address associated risks are theoretical in nature, whereas governing AI constitutes a practical challenge. The critical issue humanity confronts is how to control AI and prevent its unintended consequences. This thesis systematically explores this crucial problem by adopting a historical-philosophical approach, combining conceptual analysis, comparative evaluations of existing AI governance frameworks, and critical examinations of ethical implications through various philosophical lenses.

This thesis aims to show that if the machine (AI) is left untamed, its actions cannot be guaranteed and thus may have adverse effects on humanity shortly; also, the long-term threat index will be increased exponentially if the basement of governance for AI based on classical computers is vulnerable, the governance of quantum AI will be even worse since the classical AI are still in humanity's cognition, but the power of quantum is far beyond the limit of human's cognition not even mention their combinations, vice versa.