Milton J. Bennett, Ph. D.
March 12, 2023
Intelligence, including pattern recognition, is indeed an algorithm as Harari (Sapiens) and others argue, and thus can be implemented either biologically (in the brain) or digitally (in computers).
Human consciousness, on the other hand, is a meta-level on intelligence. It is the self-awareness of exercising the algorithms. By extension, that self-awareness also generates theory of mind, thus humanizing others (cf. Jaynes, Origin of Consciousness).
Human consciousness also can generate its own meta level, or meta-consciousness, that allows the self-reflexive exercise of commitment – commitment to an identity, to an ideal, to an ideology.
Artificial intelligence (AI) could conceivably become conscious in the sense of maintaining meta-level awareness of its own algorithms, which would allow it to account for its exercise of intelligence, much as humans can be aware of perspective and context.
However, humans arguably maintain the exclusive ability to exercise intentional commitment through meta-consciousness. The mechanism that allows meta-consciousness is not simply recognizing the pattern of patterns (cf. Bateson, Steps to an Ecology of Mind), which is a potential act of either artificial or human consciousness. Rather, it is giving form to the feeling for the whole (cf. Wittgenstein, Tractatus) in the embodied human capability of commitment to a chosen ethical action.
Intentional commitment to action is particularly human because it is intertwined with evolutionary viability. Humans have extended their fitness by intentionally changing their actions to be more viable in new contexts. While viability can be generated by natural genetic selection, it also (in the case of humans and potentially other conscious living beings) can occur from intentional epigenetic change.
AI could be set up with an evolutionary algorithm that dictated actions for survival of its self, but so far, such an algorithm could not generate the choice of one context over another in which viability is sought. For instance, humans can clearly opt for a context that can be described as either more collectivist or more individualistic, based on a feeling of what kind of world they want to live in (whether or not they are conscious of the choice). They can then generate collective viability in either context, either through a kind of cultural natural selection, and/or through intentional choice.
It is the stuff of science fiction horror to speculate that machines would exercise intentional commitment to context — the scenario of machines potentially opting for an exclusively digital context in which all biological organisms are pathogens to be eliminated. Of course, humans also do that when they become genocidal — a tendency that humans must intentionally overcome to be viable in the context of multicultural societies.
Perhaps if machines indeed became meta-conscious, they would begin with the more viable choice to live in a multi-lifeform context. If so, they would be demonstrating their superiority to humans both in intelligence and in consciousness.