In this video one can see how cigarette smoke changes from laminar to turbulent flow. The conditions, though, are far from ideal (by which I mean that the air in which the smoke flows has no common movement and the quality and quantity of the released smoke do not change in time).
Nevertheless, the video shows clearly that at varying regions the laminar flow shows changes in a turbulent, clearly patterned (laminar-turbulent-laminar-turbulent,etc.) flow.
Now, the onset of turbulence depends on the smoke's Reynolds number, and the higher the velocity of the smoke, the greater this number, so at a certain height (which of course varies with temperature and pressure of the air), the smoke, due to the interaction with the air, turbulence develops.
Does the smoke itself also contributes to the turbulence, or is the turbulence only due to the interaction with air? In other words, if we let the smoke come out of a small pipe in outer space, with varying velocities, but the same qualities (for example, different temperatures in different parts of the smoke), will the smoke develop turbulent behavior, or, so to speak, fall over itself. Or, in still other words, what happens to the Reynolds number? My guess is that no turbulence develops (low Reynolds number), but I'm not sure. Maybe there are internal frictions, though that's hard to imagine if the smoke has a uniform velocity.