OK, I'll bite.
Imagine two cylindrical electromagnets arranged coaxially one above the other, each has an N and an S. If the currents flow in the same direction, each has N in the same direction (up, say), and so the S of the upper is near the N of the lower, and they attract.
Now reduce these electromagnets to single loops. They still attract. Now imagine these loops are very very large in radius compared with their separation. They still attract. Imagine their radii are so large that when you see the two close up, you cannot even tell the wires are curved. Attraction still occurs. It does not seem logical that the whole circle of wire is needed to result in an attractive force. In fact, attraction occurs even if the wires are short lengths.
Without the math, of course, I haven't proven anything, But I hope I've made it plausible.