The signal cannot overcome the breakover voltage of the transistor alone, and R1 is there to bias the transistor on to smooth an easy passage for the AC signal. In effect we get a varying DC which is interpreted by C3 as AC and is in opposite phase to the input signal. When constructed so, an amplifier is said to be operating under class A conditions. This is nowhere near as efficient as the class C amplifiers we looked at which amplified the RF carrier, but it does give the best gain and fidelity which is what we need for audio. If we linked several together like this, gradually increasing voltage through each stage, we could end up with a high voltage signal which is a highly magnified copy of the original. For high power we would eschew transistors and use valves, the response is not necessarily as good but their power-handling capabilities are far superior.
Lastly the audio amplifier will contain, in the early stages, a filter to eliminate higher frequencies. In AM, the frequency of the modulating signal determines the bandwidth, and in commercial broadcasting 4.5kHz is the maximum permissible frequency that can be transmitted.