I'm sorry I was not clearer.
My reply was intended to encourage getting some first hand practical experience RE: the "brittleness" concern.
I think it is probably NOT a good idea to make the crankshaft bearing shell from ceramic for the IMPORTANT reasons of "embeddability" and "conformability" as mentioned by others.
Here are some specs comparing bearing steel and the (silicon nitride?) ceramic material used in ball bearings. Note it is about half as dense, 50% stiffer, much harder, and only expands about 25% as much when warm.
For what its worth, other specs suggest the thermal conductivity is beween 0% and 50% less than steel.
The specs do not indicate how just-plain-RUGGED the material is when used in ball (and recently roller) bearings.
It is not uncommon to take apart a high speed HC spindle bearing that has failed CATastrophically (siezed, 15,000 rpm to zero in about 2 seconds, plenty of stinky grease and phenolic retainer smoke) to find the steel rings smeared, discolored, and cooked, and the ceramic balls looking perfect and feeling smooth.
I am not aware of specs that describe the ductility, although I kind of recall that installing a (rare) ceramic ball bearing inner ring with inerference fit (commonly required) can be a problem, and might be PART of the explanation of limited use of full ceramic bearings.
Especially because the many performance increases just using cermaic balls.