Computing “In the Noise” Room at the Bottom

(I've clearly spent too much time with Feynman and quantum but...) Quantum computing is statistical in nature. Experiments need to be performed multiple times to get reliable results. Which got me to thinking... Quantum effects occur everywhere at room temperature. One example is the thermal noise in a resistor. It ought to be possible to impose forces, such a magnetic fields, on the resistor to "shape" the thermal noise in various ways. Using this shaping effects it ought to be possible to impose some logic "on the noise". From there, it seems reasonable that statistical measurements could yield "computed results". The useful point is that it ends up using quantum effects at room temperature, albeit somewhat slower than in a supercooled quantum computer. It seems reasonable that there must be a "speed vs temperature" curve for this "in the noise" computing. One interesting experiment would be to create a Toffoli gate in a resistor using N vs S magnetic fields as the control elements. Since this would be symmetric it would also be reversible. The end result is a "room temperature" quantum computer.