23W with extremely high gain though. It was difficult to find a number, but this source [0] gives the gain at 47dB (dBi or dBd isn't specified). With that high of gain, the effective radiated power is over 1 million Watts (because it is concentrated into a very narrow beam).
Also using phrases like "0.1 billion-billionth of a Watt" is misleading, and not how radio signals are actually described. For example 0.1 million-billionth of a Watt sounds pretty small too, but describes the power level of a usable, and not particularly uncommon LTE signal level.
Here's a related thought: if we ever manage to send a probe to the closest star, what a tough problem communication with home will be. The distance is about 2000 times higher than where Voyager is now. So the signal will be about 4 million times weaker due to the quadratic attenuation alone, and maybe 100 million if we consider the absorption by the interstellar medium. On top of that, when the probe will phone home, it's going to be a few seconds of arc away from a phenomenally bright source of electromagnetic radiation: the star itself.
You can fix this by using a bigger transmitting antenna (but Voyager's is already 3.7m, not small at all), or a bigger receiving antenna (the current ones are 20m). Or a stronger signal. But it appears we'd need to go from 23W to GigaWatts .
One thing is sure: we won't be able to send a pound-size probe.
The gain of a 3.5m parabolic antenna on the tx end, and they're using from 45 to 70 meter size dishes on the earth end to communicate with them. The NASA goldstone 70m dish has some serious gain.
Also using phrases like "0.1 billion-billionth of a Watt" is misleading, and not how radio signals are actually described. For example 0.1 million-billionth of a Watt sounds pretty small too, but describes the power level of a usable, and not particularly uncommon LTE signal level.
[0] https://www.quora.com/How-can-Voyager-send-a-signal-strong-e...