A number of recent experiments have raised interesting questions about the quantum back-action of mesoscopic quantum conductors. Example experiments include quantum point contacts being used to detect quantum dot qubits, and single-electron transistors being used to measure the position of nano-mechanical oscillators. On one hand, back-action can be "bad", as it leads to fundamental quantum limits on how well one can make a measurement. I will discuss theoretical work which attempts to demystify the quantum limit in these systems by making connections to the physics of mesoscopic noise and to information-theoretic ideas. I will also discuss how back-action can be "good": it can be used to tailor the dissipative environment seen by a quantum system (i.e. a nano-mechanical oscillator). I will discuss theory and recent experiments showing that the back-action of incoherently tunneling Cooper-pairs can be used to cool a cantilever down to millikelvin temperatures.