Can AI ever be conscious? The question stems from a misconception

The question of whether artificial intelligence can ever be conscious has emerged from a common misunderstanding. A new book, Intertwined Creatures, by cognitive scientist Anthony Chemero, seeks to clarify this issue. Chemero argues that much of the hype around AI consciousness is based on simplistic notions of cognition and consciousness, and a flawed understanding of human minds.

For centuries, Western philosophy and psychology have portrayed the mind as a separate, disembodied entity that observes the world objectively. This view leads to the belief that behavior is controlled by specific brain processes and cognition is a purely intellectual activity manipulating abstract representations. Chemero challenges this perspective, drawing on ecological psychology, enactivism, and non-linear dynamics theory.

According to Chemero, human minds are embodied, dynamic, and social, deeply connected to our environments and to others. Perceiving, experiencing, and thinking are inherently embodied activities. The body is not a mere physical object, but a living organism constantly interacting with the world. Behavior emerges from reciprocal interactions between an organism and its environment, not just from brain-body interactions.

Chemero grounds his ideas in relatable examples. For instance, noticing you're late for work, searching for keys, and becoming distracted by laundry and a forgotten apple core all demonstrate how inner speech and thoughts are intertwined with interactions with others and the environment. Even imagined scenarios, like a Black woman becoming president, are influenced by our perceptions, interactions, and social context.

Chemero's embodied cognitive science has significant implications for debates on AI consciousness. Many arguments about AI rely on oversimplified analogies to human cognition and computational processes. While sophisticated AI systems, like large language models, can generate impressive outputs, they do not navigate interpersonal interactions or possess the capacity for care. Chemero calls for a new ethics for AI, recognizing that these agents lack the qualities that would merit legal rights or moral consideration.