The anthropologist talks about a study where, after assuming vervet monkeys each had knowledge of their troupe's complete dominance hierarchy, the anthropologist and her colleagues were able to predict complex forms of those monkeys' group behavior. The anthropologist concludes that, since the prediction was so accurate, the assumption that led to that prediction must also be true: the vervet monkeys each knew about the complete dominance hierarchy.

The primatologist responds by claiming this conclusion doesn't follow. She uses an analogous argument: assuming that certain bank machines want to satisfy customer requests can lead to correct predictions of their output. By the anthropologist's logic, this means the machines actually have desires, which doesn't make sense.

The primatologist doesn't reject any of the contextual facts presented by the anthropologist. The primatologist doesn't even necessarily take a stand on whether the vervet monkeys did have knowledge of the dominance hierarchy or not. Whether that conclusion is itself correct or incorrect, the primatologist's concern is that it doesn't follow: she is concerned with the anthropologist's reasoning that if an assumption helped lead to highly accurate predictions, that assumption must be true. She shows the problems with this reasoning by making an analogous argument about bank cash machines, where the same reasoning leads to an absurd conclusion.