Life 3.0: Being Human in the Age of Artificial Intelligence

The final chapter picks up where the previous one left off: “philosophy with a deadline” (281). In this chapter, Tegmark investigates the nature of consciousness, one of the book’s central themes. For Tegmark, it is extremely important whether or not we can determine if AIs are conscious and how. If, for some reason, humans were replaced by AIs, it would be a tragedy if these AIs were merely zombies, i.e., utterly unconscious.

Tegmark gives a broad definition of consciousness as “subjective experience” (283). Basically, consciousness is what it’s like to be a thing and is composed of qualia, or instances of experience. The “problems” with understanding consciousness are layered. Figure 8.1 articulates a pyramidal structure for these problems, ranging from the scientifically achievable problems of brain processing to the philosophical problem of why anything is conscious at all. The later has often be dubbed the “hard problem of consciousness.” Tegmark relies on the work of Australian philosopher David Chalmers when discussing these issues. He contrasts his own view, “the physics perspective,” with substance dualism, or the view that mind and body are utterly dissimilar forces (286). The “physics perspective," often known as physicalism, assumes that consciousness is connected to particular physical arrangements. Tegmark is interested in determining which arrangements are conscious, which aren’t, and why.

Tegmark produces experimental data and proposes other experimental possibilities for the study of consciousness. He discusses the “behavioral correlates of consciousness,” that is, the behavioral signs that conscious processing is taking place. He discusses the nature of the brain and the frontal lobe and the prefrontal cortex. He also invokes the notion of “neural correlates of consciousness” (295). Just as there are specific behaviors associated with conscious processes, so there are with various brain states. It is important to note that neural and behavior correlates may involve strong correlation between a behavior or brain state and consciousness but do not yet answer the “hard problem” of consciousness. Tegmark notes that not all areas of the brain appear conscious, and that consciousness seems more connected to specific arrangements of neural networks rather than just the neurons themselves. He also points to research that shows that consciousness lags slightly behind real time: “brain measurements can sometimes predict your decision before you become conscious of having made it” (298).

Tegmark then turns his attention to various theories of consciousness. As a physicist, Tegmark is well-known for the “mathematical universe hypothesis,” the philosophical view that the universe is fundamentally a mathematical structure. Though he does not discuss this theory here, he indicates an orientation toward a “mathematical theory of consciousness” (299). He expands from the neural and behavioral correlates of consciousness to general physical correlates (299). He believes that consciousness may be an “emergent” phenomenon, a complexity resultant from simple arrangements of particles. He discusses the “integrated information theory” of consciousness developed by Giulio Tononi and describes it as “the most mathematically precise consciousness theory to date” (301). One implication of this theory is that there is a particular way it feels to be certain kinds of information. This returns to the idea of substrate independence:

If consciousness is the way that information feels when it’s processed in certain ways, then it must be substrate-independent; it’s only the structure of the information processing that matters, not the structure of the matter doing the information processing. In other words, consciousness is substrate-independent twice over!” (304)

Consciousness feels independent of the matter that it’s reliant upon because, in two important ways, it is. First, sense consciousness is defined as the subjective experience of an information state is one level removed from the bare mathematical information: it’s the feeling of that information. Second, that information is itself independent of the substrate that it uses, which in our case, is the brain. This, for Tegmark, explains why we feel so strongly that the mind is separate from the body. In a way, it is. It requires some kind of physical material, but this “hardware” can be radically transformed.

Tegmark ends with some brief speculations on what AI consciousness might feel like, noting that the scope of its conscious experience could be beyond our ability to imagine. When thinking about the ethics of engagement with potentially conscious AI, Tegmark notes that, regardless of whether any individual conscious being actually exerts free will (it would seem Tegmark believes this is impossible) it is the case that all conscious beings would subjectively experience free will. This makes them subjects of moral concern. Tegmark believes that the possibilities of AI in the future are incredible and ends with a note on the problematic view of human exceptionalism. Human hubris, he writes, has been a source of trouble from the beginning. We should be prepared, going forward, to take a humbler approach to our place in the cosmos.