Variable resistor - mechanised oscillator?

Imagine you have a motor with a gear on top, then you used a gear train to significantly increase the speed output from a gear train, you linked the final gear to a reciprocating mechanism (with enough torque to drive it), and then linked this to a potentiometer you would create a varying voltage and current.
If you were to use a capacitor to remove the DC bias, could the output signal be used as a carrier wave?

Why would you want to do that?

The last of the electromechanical generators was the Alexanderson alternator, installed for CW telegraphy in the early XX century.

Anyways, such an arrangement would only produce frequencies in the low Khz......and perhaps not even that.

If you need to measure rotational frequency of your motor there are all sorts of encoders or mount magnet on motor shaft and use Hall effect pick-up. ... Just guessing

I read somewhere that at nanoscale some nano-motors can be used for this purpose. It was a pretty old article.

Nano-motors for frequency generation - very interesting! What are the advantages? Frequency stability, phase noise, level of harmonics?
... seems we drifted away from posted question : ).

The question was 'Why would you want to do that?'