All files are made up of “text”, or rather, numbers. How each program interprets those numbers differs depending on the kind of work they do. Any program can open any file, but the way it translates those numbers won’t make any sense if the file wasn’t intended to be opened by that kind of program. So, if you opened an MP4, you might see a little bit of metadata that was encoded in a way the text editor can understand, and then you’d get a ton of random symbols, some that are numbers and letters you recognize, but a lot of them would be specialized characters from farther on in the list of characters whatever font is being used might have.

Think of it this way: take two human languages that use the same writing system, like German and French. Suppose you ask a Frenchman who also speaks English to translate and write down a few specific sentences. You then take those sentences to a German who also speaks English (but not French) and ask her to translate it into English. Obviously she can’t. She might be able to sound out the words, but neither of you will know what it means, and it probably wouldn’t sound right to a French speaker. Or better yet, you can ask her to try and guess what each word means. She’d likely come up with mostly nonsense (minus a few cognates and loanwords). This isn’t an exact analogy, but that’s basically what’s going on.