"no no I aggregrate preferences over all computable agents weighted by the universal semimeasure it's not my fault that they all want to do what I was gonna do anyway"

To spell it out more, math describes relational structure, because it lets symbols stand in for the thing. Because mass is defined by what it does, in QM the operator M plays a role in the Hamiltonian exactly parallel to how actual mass constrains the dynamics of a particle’s position and momentum. That is: the way the symbol for mass relates to the symbol for position is isomorphic to how actual mass relates to actual position in the world. But because qualia are intrinsic rather than relational, there is simply no structure for math to be isomorphic to in the case of conscious experiences The way to push back is to say that blue merely is relational, defined by its position in a color space, but the pushback to *that* is: no, the constraints of color space aren't enough to fix blueness. Even if Mary knows exactly how all the color brainstates relate to eachother (indeed, if she knows the exact math isomorphic to a whole brain) she still learns something new when she sees color.