An oddly vague and empty article that gives up at the first step of analysis by begging the question of a "myth," and justifying it with the oft-repeated gross oversimplification of AI fiction that LLMs "merely respond with what has the highest statistical probability, given the user’s prompt". This is like an explanation of digital photography that says an image is captured merely by a set of pixels corresponding to the highest statistical probability that photons struck points on a plane; a science-sounding explanation that exposes nothing about the theory, techniques and engineering of devices for image capture.
While there's no reason whatsoever to assume AIs are conscious, the reliability of this assumption stems from lack of criteria for the phenomena of consciousness combined with the obvious lack of even rudimentary similarities between AI computer architecture and life forms.