Intel reinvents the transistor as a 3D Tri-Gate (opens in new tab)

If you were to make a gear at 22nm then blow it up to say 5 inches wide, it would be the most terribly machined gear you ever saw. Heck it might look like a roughly notched rock. Precision has to be refined at this level to make complicated nano technology

You have to understand just how these chips are made. You start with a wafer, send it to thin films cover it in a uv resistant coating affectionately named photo resist, send it to photo lithography, expose it to uv light through a precision laser etched crystal reticle to overlay a schematic design for the bottom layer of the chip, then you ash the wafer to remove all the parts of the photo resist that the UV light did or didn't soften/harden (there are positive and negative photo resist types) then send it to implant and implant materials such as boron, phosphorus and other materials that will change the electrical properties of the wafer where implanted, then UV expose again, ash it, now yo have the bottom the wafer is annealed to bring the implanted particles to the top. Sometimes there is etching before any of this happens. Then you build the wires above the silicon connecting all the parts, and sometimes growing silicon in upper layers just to add more parts to the circuitry. All through the use of etching(plasma, acid and others), photographically, ashing among other things and there are many many quality checks along the way just to make sure everything is going well. (no reason to waste a month of process time on a chip that won't work because of something that happened at the beginning of the processing. Basically everything is done in layers. It can take months for chips to go from silicon to shippable chips. I suppose you might be able to build mechanical object with photo-lithography and I'm sure it has been done, but until processes are refined to where the objects don't look as rough under a microscope as they do, then I don't think it is very feasible.

Edit: On another note, now that we're able to make 22nm chips, by using the same process we should be able to produce larger nano objects with greater precision, like say 100-200nm. But as for 22nm precision, it's going to take a while to get machine shop quality parts when scaled to large size at the 22nm size.

There will be some areas of research where Intel's photolithography improvements are relevant and others where they're completely orthogonal, like repairing human tissue. That's just how these things go. :)

Even as an undergraduate back in 2005, I was fairly easily making ~50nm structures with interference lithography and the equipment wasn't even that expensive. One day that technology should be useful for filtration, quantum dots, and such from the list you linked.

Edit: Fixed typo

Actually I thought he was fine