Exactly, the assumption (known as the inductive hypothesis) is completely fine by itself and doesn’t represent circular reasoning. The issue in the “proof” actually arises from the logic coming after this, in which they assume that they can form two different overlapping sets by removing a different horse from the total set of horses, which fails if n=1 (as then they each have a single, distinct horse).
Exactly, the assumption (known as the inductive hypothesis) is completely fine by itself and doesn’t represent circular reasoning. The issue in the “proof” actually arises from the logic coming after this, in which they assume that they can form two different overlapping sets by removing a different horse from the total set of horses, which fails if n=1 (as then they each have a single, distinct horse).