Ants get around by a combination of pheromones (chemical signals) and a special internal sensory system that functions almost like a natural pedometer for them (in that—if I’m remembering this correctly—it essentially “unwinds” when an ant takes steps in one direction, and them “rewinds” when they take steps in the opposite, so in this way worker ants can essentially sense whether they are moving the correct direction with regards to their original path or not, and then in combination with pheromone “breadcrumb trails” can almost always retrace their exact steps back to the colony like clockwork). Of course, there are many ways these navigation systems can be thrown off, like a human moving an ant some distance from its original path, or cutting off one joint of each of their legs so that their internal “pedometer” system will register too few steps (this is how the system was originally experimentally observed; some ants had their legs shortened, and others artificially lengthened, and it would cause them to either stop short of the colony or walk too far very consistently. Not the nicest of experiments if you’re someone who feels bad for ants, but it did provide some valuable insights).

So to answer your question, no, if you were to move the ant far enough off course, it would not be able to find its colony again, or at least would have great difficulty in doing so most likely.

And ants definitely do not join other colonies. Ants are an example of a eusocial organism, meaning essentially that they live as a group of closely-related individuals that function as a cohesive collective, with only select members of the collective having reproductive privileges (the queen and the male drones), and the rest being strictly non-reproductive. It’s a social structure seen in ants, bees, wasps (although not all species of any of these), termites, some beetles, some marine shrimp, and (possibly most bizarrely) naked mole rats and another species of mole rat (the only known eusocial vertebrates), and probably others. Eusocial organism colonies are all the children of (usually) a single reproductive female “queen” and a small number of reproductive males (which in eusocial insects tend to be short-lived and literally exist solely to fertilize eggs and produce workers, as well as also being children of the queen themselves) and therefore are all genetically very close and almost act like the cells of a single “superorganism” more than individuals, as the workers can’t reproduce and have literally no other reason for existing except serving their colony’s needs. A worker ant can’t join another colony any more than one of your white blood cells could be put into another person’s body without being rejected and destroyed by the other person’s immune system; ant colonies are a packaged deal, and all other colonies are seen as competitors for the same resources that need to be destroyed on contact.