This is the most common result, and not necessarily a bad thing at all. Motiva Message Testing AI will continue experimenting across versions, gradually investing more sends in the versions that are getting positive responses and fewer sends in poorer-performing versions.
Don’t worry if you never reach a single winner - there can be many good reasons for this: maybe you have different sub-audiences responding to different variations. Maybe your variations are so similar to one another that responses would naturally be pretty similar. Keeping testing, learning, and improving!