Reword and move dispatch around (Rust-GPU#27)

Author: LegNeato

blog/2024-11-21-optimizing-matrix-mul/index.md

@@ -177,7 +177,7 @@ examples.
 
 #### Dispatching workgroups
 
-Each workgroup, since it’s only one thread (`#[spirv(compute(threads(1)))]`), processes
+Each workgroup, since it's only one thread (`#[spirv(compute(threads(1)))]`), processes
 one `result[i, j]`.
 
 To calculate the full matrix, we need to launch as many entries as there are in the