The Cartesian begins with a hidden assumption - let's call it the internalist assumption - that in order for me to know that "here is one hand," there needs to

The Cartesian begins with a hidden assumption - let's call it the internalist assumption - that in order for me to know that "here is one hand," there needs to be no possibility that we are in a simulated world (for if it's possible I'm in a simulated world, then I don't really know that here is one hand). Moore's point is: why is the internalist assumption any better than the assumption that here is one hand?