Each note is uniquely identified by its path relative to the notebook's root. However, in some cases it is more convenient to refer to a "note ID", which is the unique part of its filename. For example, the note ID of the file 200911172034 An interesting concept.md is 200911172034. You could have several notes named "An interesting concept", but only one with the ID 200911172034.
The purpose of using a unique identifier in your note filenames is to create stable links between your notes, which will not break even if you change the title of the linked note. See this reference for more information.
There are several flavors of note IDs and zk supports most of them. You can set it up in the note configuration.
A random ID enables short and memorable unique identifiers. By default, zk is configured to generate random IDs of four alphanumeric characters. I found this to be the sweet spot between an easily memorable and usable ID and enough candidates. This default setting can generate 1 679 616 unique IDs.
Another common ID is a timestamp in the YYYYMMDDHHMM shape. This is less readable than a short random ID, but has the added advantage of being sortable by creation date. However, I find this not so useful in practice.
Sequential (incremented) IDs are currently not supported by zk. They get ugly very quickly when deleting outdated notes and have an irregular shape.