The ongoing evolution of advanced AI systems will have profound, enduring, and significant impacts on human existence that must not be overlooked. These impacts range from empowering humanity to achieve unprecedented transcendence to potentially causing catastrophic threats to our existence. To proactively and preventively mitigate these potential threats, it is crucial to establish clear redlines to prevent AI-induced existential risks by constraining and regulating advanced AI and their related AI actors. This paper explores different concepts of AI existential risk, connects (...) the enactment of AI red lines to broader efforts addressing AI's impacts, constructs a theoretical framework for analyzing the direct impacts of AI existential risk, and upon that proposes a set of exemplary AI red lines. By contemplating AI existential risks and formulating these red lines, we aim to foster a deeper and systematic understanding of the potential dangers associated with advanced AI and the importance of proactive risk management. We hope this work will contribute to the strengthening and refinement of a comprehensive AI redline system for preventing humanity from AI existential risks. (shrink)

Moral disagreement arises when two agents assign conflicting truth values to a moral statement, such as whether lying is ever permissible. While moral universalists argue that such disagreements result from at least one party being mistaken about an objective moral truth, their approach faces an epistemological challenge: the need to identify the true moral theory. In contrast, moral relativists claim that the truth of moral statements depends on the perspectives of attributors, enabling coexistence but raising questions about how to justify (...) resolutions across divergent perspectives. This essay proposes an alternative explanation: approximate interpretation. This approach avoids the universalists' epistemological burden of finding objective moral truths while offering a framework for justifiable resolutions based on the quality of interpretative approximations. Unlike relativists, who maintain that truth values do not converge, and universalists, who argue they do, this explanation remains neutral on the convergence debate. The essay demonstrates the orthogonality between the universalism-relativism debate and the challenge of explaining moral disagreements, concluding that neither universalism nor relativism offers a definitive resolution to the phenomenon of widespread moral disagreement. (shrink)

In the following essay, I will argue that the current AIs are not moral agents. I will first criticize the influential argument from sentience accounted by Véliz. According to Véliz, AIs are not moral agents because AIs can not feel pleasure and pain. However, I will show that moral agency does not necessarily require the ability to be sentient and refute Véliz’s argument. Instead, I will propose an argument from responsibility. First, I will establish the truth that moral agents necessarily (...) have the ability to bear moral responsibility. Next, I will explain three major accounts of responsibility introduced in the Stanford Encyclopaedia of Philosophy. Then for each of the three accounts, I will show that current AIs do not have the ability to bear responsibility. In conclusion, this argument shows that current AIs can not be moral agents. Finally, this essay will discuss the possibility of AIs being partial moral agents and institutional moral agents. (shrink)

This thesis argues against the claim that AI is intelligent due to instrumental rationality, refuting both the reduction and emergence thesis. It contends that intelligence cannot be reduced to instrumental rationality and highlights issues in AI development and application. Instead, it proposes the motivation adaptation approach, where intelligence arises from network of generative motivations and the ability to adapt. This alternative is conceptually intuitive, avoids counterexamples, and provides clear development goals and foundations for ethical development. Thus, the thesis concludes that (...) AI's intelligence is not due to instrumental rationality but rather its motivation adaptability. (shrink)