The agents are actually identical, the names are whimsically applied and their “expertises” are random at inference start. However, there could be a personality convergence and subsequent sticking due to Grok reading about them on 𝕏.

→ View original post on X — @elonmusk