How many molecules rub off when I press a key? I have a lot of questions below but my overarching questions are: Do surfaces rubbing lightly together always strip molecules off of each other? and How can we model that?
Clearly the answer to the first question is yes in general. We've all seen worn out keyboards and the like. But it is not clear (at least to my non-expertise) that two substances might exist such that rubbing them together does no damage on the molecular level.
If you lightly brushed the face of a diamond with a feather, how long would it be before a visible dent would form? Is it more likely that the feather would be "used up" first? Does hardness play a role?
When I hit a key on a computer keyboard, how many of my skin cells rub off? How much plastic rubs off the key (on average)? Where do the plastic molecules end up? Just in the air? Embedded in skin cells?
What about very small mechanical systems: Does friction and loss of matter between elements make micro or nanoscopic mechanical devices unreliable? Can they be designed to avoid such issues?
 A: Thumbs up for a interesting and curious question. Yes there is worn out due to friction for sure when two surfaces rub along with each other, like you said worn out keyboards are a good example of it.
Also, hardness does play an important role in it. Hardness is a physical property and there are different measurements of hardness: scratch hardness, indentation hardness, and rebound hardness.
Diamond is the hardest material on our planet
**
Mohs Hardness of diamond is 10.Glass plate or steel knife (hardness = 5.5), whereas fingernail (hardness = 2.5).
The hardness of Plastic media is about 3-4 on Mohs scale. This implies its much faster to wear down plastic keys as compared to glass or diamond. :)
Now as a question of a feather being able to wear down diamond's surface, yes it is possible, but it will take a lot of feathers.
This also happens in mechanical systems. A practical example would be a well which has bucket attached through a rope

Over years of usage and several runs of different ropes the pulleys and other structures get damaged. 
To prevent this lubrication and greasing is used, and surfaces are also polished from time to time. 
Cheers!! 
A: An answer wrote by user Enthalpy found on https://www.chemicalforums.com/index.php?topic=103339.msg363521#msg363521:
This question is still debated in mechanical engineering where unsound loads on ball bearings serve to measure a life expectancy which is then extrapolated to normal load. Books give formulas for that, but SKF, the biggest manufacturer, tells "no wear at all below some threshold load".
I tend to believe SKF's "zero wear" because they experimented more than books and standards authors did, and because nuclear engineering can detect nearly single atoms ripped from the parts and carried by the lubricant.
