ResonantW

Griffiths brings up a point that it might make sense that current in a fat wire would want to concentrate itself mainly along the axis of the wire, given that parallel wires attract each other.

He then asks us to figure out why this might not happen, but to be honest I would think it should! The only thing I can think of is some sort of electric repulsion in the conduction particles' rest frame, but I feel like there are more subtle points in transforming the frames like this. Where am I going wrong?