Consciousness has been a long-standing conundrum for many reasons. Historically, the intellectual and academic need to understand ourselves came first: the observer observing the process of observation. You may call it meta-observation or metacognition. Trying to understand that process one meta-level further up is not straightforward. Each school of philosophy has its own understanding of consciousness and of how it fits into the puzzle of the known universe. Those disagreements are unlikely ever to be resolved, because almost all of these schools consider consciousness to involve qualia: something irreducible, something undeconstructable. It is something you can say if it is there, but it cannot be defined because there is nothing more basic by which to define it. That is the curse of trying to apply axiomatic principles to language; it kind of works, up to a point.

From a more pragmatic point of view, a proper, universal, and timeless definition is not needed as long as we can somehow determine where something stands on the spectrum of consciousness. Our society requires the presence of consciousness in order to assign responsibility. If you are deeply asleep and cause harm to someone, you will most likely be acquitted in court. Our language works in the same direction: we do not say that AI makes decisions; we say that an algorithm (or a machine), processes the available information to recommend an optimised solution. Or at least that was the case until relatively autonomous AI agents came along. Now we delegate more and more real-world capabilities to these agents, and as a consequence, both our language and our understanding of responsibility are shifting toward the realisation that this delegation, while perhaps very useful and profitable, comes with a price tag.
Are we ready to accept the payment? Aside from some media rhetoric, I don’t see any sign that we are!