Well, we didn't need AI to start down the path. The core 'can this be done' science is done. Yes, we can control chemical reactions by manipulating reactant position.

Next the question is how do we scale it? We need LOTS of manipulators. We're... not actually great at building lots of manipulators. We might use this atomically precise tech to build manipulators and have them make copies of themselves.

Or we might leverage other technologies like protein engineering, DNA origami, and nanoelectronics to make less effective manipulators that we could build today and use to bootstrap toward the nanotech that CAN build everything atom-by-atom.

Either way, things are accelerating. Finally.

My guess is that this is highly useful for either:

* Building chips (but it's on a silicon surface, not with silicon)

* Carbon nanotubes (looking forward to a space elevator, folks, one day ;))

* It's a more notable movement forward in nano construction in general and I just know too little to understand its impact.

Given the zero comments to date: would someone who knows this area like to accept an invitation to share something really interesting about this paper, please? :)

https://wiki.reprap.org/wiki/Drexler_Arm

and put a post-it between them with COMPUTONIUM???? written on it

Can we please keep it together for a modest number of decades longer please.