Above the generous threshold of about 128 bits, any given number is vanishingly unlikely to have been derived through any other process. If your 134-byte file turns out to be valid SVG for the Cartoon Network logo, then functionally speaking it is the Cartoon Network logo because it's incredibly unlikely you just pulled those particular 1072 bits out of your ass.
Similarly, if you are in possession of a 30-million digit binary number that happens to be valid h264 for an ISIS recruitment video, "oh what a stunning coincidence" isn't going to wash.
So "files are just binary numbers" doesn't get you anywhere, at all.
To be more specific, it is a number used as an argument to a function. The bit stream input into a video decoding algorithm is what makes the video. The video binary itself is mostly nonsense.
But the other way to look at is that it would be possible to have a set of data that is essentially illegal to use one function on and legal for another function to be used. Could someone unknowingly generate or acquire a set of data that is illegal to access and own? There has already been cases of terrorists hiding information within legal media.
Above the generous threshold of about 128 bits, any given number is vanishingly unlikely to have been derived through any other process. If your 134-byte file turns out to be valid SVG for the Cartoon Network logo, then functionally speaking it is the Cartoon Network logo because it's incredibly unlikely you just pulled those particular 1072 bits out of your ass.
Similarly, if you are in possession of a 30-million digit binary number that happens to be valid h264 for an ISIS recruitment video, "oh what a stunning coincidence" isn't going to wash.
So "files are just binary numbers" doesn't get you anywhere, at all.
https://qntm.org/number