No that's impossible with your voice even for those slow speeds back then. But one could implement a similar system that converts sound to data/images if you invent you own modulation depending on how precise you can make sounds. If you are totally ungifted with your voice like me, the simplest would probably be something like morse. If you have perfect pitch you could use many different frequencies and amplitudes and combine those (a bit like QAM but without phase modulation, because humans can't control the phase with their vocal cords). The more different frequency-amplitude combinations you can make and the faster you can switch between them the faster you "manual" data transmission could be.

Sounds actually pretty fun to play around with 🤣

PS: I personally would drop amplitude modulation and only use different frequencies (FM) and only pick a few different values (maybe four? or eight - one octave cdefgabc') with variable speed (decoding based on short pauses between notes) to be most reliable.

PPS: what did you (and adhd) do to me 😅. Now I want to make something like that and have no time on my hands anyway.