That’s too much for any human to read. For ultra high-definition displays, 50000 non-overlapping texts would mean 165 pixels per text (if you have nothing else besides the texts). My expectation is that you need at most a few hundreds of recognizable texts (the closest to the camera), and all the rest could be just merged into a soup of pixels.
Anyway, you may have a look at @prisoner849’s work: