What differences do we see when we compare humans and AI? I put forward five comparisons which are the root causes of the machine penalty: appearance, identity, behavior, mind, and essence. First, machines have different appearances, including bodies, voices, and faces, leading to humanlikeness or an uncanny valley. Second, AI have specific identities assigned to them based on their design, compared to humans who are freer to select their own identities and often have many identities. Third, AI behavior has traditionally been more formulaic and predictable, making it difficult for AI to pass as humans or pass the Turing Test. Fourth, AI minds are seen as having somewhat less capacity than humans for agentic thinking and much less capacity for experiential feeling. Fifth, the essential nature of machines includes their being mechanical, not biological, and beliefs they are not sentient and, therefore, do not have true emotions. The five human–AI comparisons follow a progression with appearance being highly based on observation and applying to an individual AI, whereas machines’ essential nature is a belief that is applied across all types of machines.