Computers are designed to automate intellectual labor.  In the process they do what straightforwardly invites description as "representing" or harboring "information," making "inferences," and taking "decisions" perhaps based on "goals," "constraints," or "criteria."  Computers equipped with "sensors" are naturally said to "sense"; and if it's visual sensing even to "see" things.  Note that the doings of these devices could scarely be described at all, much less straightforwardly, without resorting to these or similar locutions.

Such facts on the ground -- or, rather, in the circuitry -- are the most potent arguments for artificial intelligence.  Evidently computers think.  The more compellingly evident it becomes, the heavier the burden rests on AI deniers  to  produce theoretic reasons against able to withstand or undermine this evidence.