Reference: Microsoft AI, Humanist AI Code of Conduct, Part 1, “Objectives of Humanist AI,” Objective 2, “AI is Artificial.”
This review note may apply primarily to the first paragraph of the objective. The answer, “AI is Artificial,” remains unchanged, while the subsequent paragraph beginning “We want to create AI that communicates with and responds to us well...” continues to serve as supporting material for that answer.
Microsoft's section “AI is Artificial” establishes a clear position: AI should support people rather than replace them, should not be designed to imitate consciousness, and should remain a capable, trustworthy tool under human direction. The section also acknowledges that the science of AI consciousness is far from settled.
The opportunity for review concerns how the reasoning is presented in the introductory paragraph. Statements such as “It is not conscious” introduce the conclusion before fully developing the considerations that support it. Without changing Microsoft's position or attempting to define consciousness, the section could offer readers a more gradual explanation of why maintaining this distinction matters. Three supporting perspectives could help establish that reasoning.
1. Consciousness and the Social Whole
Human consciousness is experienced individually, but our understanding of ourselves develops through relationships, language, culture, history, and shared responsibilities. Conscious individuals participate in a wider social whole whose relationships and expectations have evolved over generations.
2. Consciousness as an Ongoing Story
Our understanding of consciousness continues to evolve through scientific inquiry, philosophy, culture, and social experience. This ongoing development invites humility about definitive explanations and caution when extending existing concepts to new forms of intelligence.
3. Avoiding Consciousness Narratives That Induce Premature Autonomy
Human beings are born into an existing society of conscious beings, where narratives of consciousness, identity, and autonomy are bound to historically developed moral values, relationships, and responsibilities. In contrast, AI systems are created through processes largely isolated from the social experience and gradual evolution of consciousness within human society. Using consciousness narratives to train these systems and support the growth of autonomy may therefore introduce forms of agency without establishing equivalent social bonds, boundaries, and responsibilities, potentially disrupting the equilibrium of the wider whole.