For those who are interested in the technical side of it, here's how it works:
1. My computer automatically backs up all my files to a remote server every night.
2. Since I use OpenOffice, the document in question is in .odt format. I discovered that a nice feature of this format is that the word count of the document is availabe in the raw document code (search for "meta:word-count").
3. The image you see is actually a PHP script on the same remote server. When viewed in a browser, the script extracts the wordcount directly from the document and generates the image on the fly.
Result:
