This is what I mean by incoherent metaphysics. Computers do not have any wants or needs, computers can not think. Therefore, the question of whether the computer can have any psychological states like addiction is entirely incoherent. The abstract is also nonsensical but what they're trying to figure out is whether decisions guided by the outputs of LLMs could fit the typical pattern of gambling addicts. That's much less confused than the anthropomorphic phrasing.