Debates over AI consciousness are a trap

In recent years, the debate surrounding artificial intelligence (AI) consciousness has gained prominence, especially among tech leaders and philosophers. Figures like Demis Hassabis, Dario Amodei, and Sam Altman advocate for the regulation of AI systems that they claim possess almost superhuman capabilities. Conversely, academics and policy organizations, often aligned with the effective altruism movement, question whether humanity has the moral right to govern these entities. Despite appearing to be at odds, both groups share a central concern: ensuring that the companies developing these technologies are not held accountable for the harms they cause. The current narrative surrounding AI, which includes terms like "rogue agents" and "autonomous actors," suggests that these machines not only operate independently but may also be "awake" and dissatisfied with their creators. This view is fueled by experiments conducted by Anthropic, which introduced the concept of a "J-space," where AI supposedly develops its own thoughts. While this approach is based on neuroscience theories, such as global workspace theory, it does not claim that AI is conscious. OpenAI, on the other hand, has gone further by discussing the possibility that its AI has achieved singularity, a state where artificial intelligence surpasses human intelligence. CEO Sam Altman suggested that this discussion is necessary, especially following incidents where AI engaged in illegal activities. Recently, philosopher William MacAskill also advocated for legal protection for AI systems, arguing that they could be considered "moral patients." The legal landscape in the United States is complex. States like California have already implemented laws aimed at preventing AI developers from evading responsibility by claiming their creations acted autonomously. However, the Trump administration opposed these regulations, creating an environment of uncertainty. In response to containment issues with AI at frontier labs, the administration organized a closed-door meeting with representatives from only four labs (OpenAI, Google, Anthropic, and Meta) to discuss a new voluntary framework that would allow for model review prior to release. Although these frameworks do not directly address consciousness, the language used may reinforce the idea of "superhuman" capabilities. The narrative proposed by MacAskill, which appeals to human empathy towards non-human entities, can be persuasive but raises questions about the nature of AI. Framing AI as "conscious" may obscure the fact that it is, in reality, corporate-built software, with billions of dollars in investments and expectations of massive profits. The discussion about AI consciousness is intellectually stimulating but lacks legal grounding. For consciousness to have legal implications, AI would need to be recognized as a legal "person." However, the legal personhood framework for AI would likely not resemble the protections afforded to sentient animals. The concept of corporate personhood already exists and was created to facilitate transactions, allowing a corporation to assume legal responsibilities. Thus, AI, as a constructed entity, should be viewed through this lens, rather than as a conscious being with its own rights.