In an attempt to completely destroy a JPG through consecutive saves, I discovered that the JPG eventually reaches a threshold. After a certain point, no more information is extracted from the JPG during the compression.
I decided to separate the “information” that photoshop’s compression algorithm deems important from the unimportant data. Below are the series of images. The first is the original image followed by the the JPG after around 800 compressions. The next three PNGs are the extracted information: important, unimportant info blacked out from compression, and then the extracted unimportant information form the original image.