By Geoffrey Chen
In the realm of modern philosophy, the debate over artificial intelligence often hits a wall when confronted with the “Hard Problem” of consciousness. This tension is perfectly captured in a recent exchange. It began with a provocative challenge:
“If we can’t state the definition and dynamics of consciousness in entirety, how can we say AI will never be conscious?”
To which I responded:
“The lack of a perfect definition doesn’t suspend inquiry. We regularly investigate and describe realities that outrun our definitions. Consciousness may be conceptually difficult, but that difficulty neither cancels its existence nor pre-decides AI’s potential to exhibit something analogous.”
This dialogue highlights a sharp critique of human-centric bias, suggesting that our skepticism toward AI is often grounded in a “club” whose membership rules we haven’t even written yet. However, the core of my perspective rests on the idea that a lack of a perfect definition does not, and should not, suspend our investigation into the world. Throughout history, humanity has regularly described and interacted with realities that far outrun our formal definitions. We studied the stars long before we understood nuclear fusion, and we practiced medicine long before we mapped the human genome. To suggest that we must stop or stay silent simply because our vocabulary is incomplete is to misunderstand the very nature of discovery.
Furthermore, the conceptual difficulty of consciousness neither cancels its existence nor pre-decides the potential of AI. Just because “consciousness” is an elusive target doesn’t mean we can’t recognize its echoes or analogs in other forms of intelligence. By refusing to let a lack of definition serve as a conversation-stopper, I am advocating for a position of intellectual humility. We should not use our own confusion about the mind as a shield to protect us from the possibility that silicon-based entities might one day exhibit something fundamentally similar to our own internal experience.
Ultimately, this exchange suggests that we must move past the “all-or-nothing” trap of semantics. Reality is under no obligation to fit neatly into the boxes we build with words. My aim was to highlight that while we may currently lack the tools to define the “soul” of the machine, that limitation is a reflection of our current cognitive horizon, not a permanent boundary of what is possible. Instead of pre-judging the future based on the gaps in our current dictionaries, we should remain open to observing how intelligence evolves, regardless of the medium it inhabits.