Why do servers cache Web pages even though they know the Web page will be changed in the future?

For performance, some Web pages get cached. This is because the server can then speed up page load time:

Without a cached copy  +--SERVER--+                                                   +--CLIENT--+ | [file] --> load file from disk --> read file --> sends over --> loaded  | +----------+                                                   +----------+  With a cached copy +--SERVER--+                                             +--CLIENT--+ | [file] --> read file from cache --> sends over network --> loaded | +----------+                                             +----------+ 

This speeds up the operation.

However, when a file changes, the cache does not change. Even though the cache can speed up load time by a lot, the file is still old. The Web site has to regenerate the cache again.

So why do Web servers cache pages even though the file will have to change in the future?