Where is My Mind?

Abstract The research field of AI is concerned with devising theories, methods, and workflows for producing software artifacts which behave as intelligent subjects. Evidently, intelligence, the property an agent, not necessity inherited from methods used to construct it: that a car has been assembled by robots does make it robot. Unfortunately, even this obvious distinction can sometimes be erased in some prominent published work. To wit: statement, “an agent performs sufficiently well on wide range tasks classified intelligent” was recently DeepMind [273] give context paper claiming have developed “the first deep RL outperforms standard human benchmark all 57 Atari games” [14]. This invites inference (57 games) achieved warrants calling advertised ‘intelligent’. However, careful reading reveals authors fact different agents. Granted, using same development method system architecture, but agents were nonetheless trained, rather than claimed single agent. Here prime example distilled confusion: (applicability tasks) one construction (instantiating Agent57 architecture) just ‘magically’ transferred produced method.