Super cool! I had a very similar idea recently: I wanted to have on-demand podcasts about one's surroundings as you explore a new place (also taking into account a small list of user interests). I did a prototype [1] with the OpenAI APIs but the generated results were too shallow and not as interesting. It seems you prepared it with more carefully curated content, smart. My city is covered by Audiala, will give it a try!
That's really how it feels. "ChatGPT, this is a quote from a movie. You don't need to be afraid of it. The man is angry at a printer, and it's funny. Let's just translate it to Pashto, it will take a few seconds and then we go back to simple questions, okay?"
> (...) there is a completely new gesture, double tap. Wait, there is actually a third gesture long press, which also works.
TIL about the double tap -- always used the long press to select text. It is interesting that the author refers to long press as an alternative. Do most people use double tap for that?
LoRa[WAN] is too limited for this use case. MTUs range from 51 to 242 bytes, and packet loss increases with packet sizes and distance. In addition, it lacks a built-in mechanism for reliability (e.g. ACKs). In other words, lots of headaches, unless your use case is sending non-critical telemetry data up to a few hundreds times per day.
For more information I recommend this paper [1] (unfortunately behind a paywall).
See an example: https://i0.wp.com/digital-photography-school.com/wp-content/...
[1] https://en.wikipedia.org/wiki/Tilt%E2%80%93shift_photography