I imagine if deep learning is going to lead to molecular assemblers, it will more likely be in the automated design of enzymes and similar structures that could

I imagine if deep learning is going to lead to molecular assemblers, it will more likely be in the automated design of enzymes and similar structures that could be easily mass produced by say the genetic engineering of some microbe.