Saw this today on metafilter.  A CS grad student in Dublin has found a way to figure out what is under those black ink blots in classified documents.  I was just wondering last night if there was some way to do this.  We had been discussing censoring soldier's letters at lunch yesterday.  I wonder how this will effect declassification of documents.  Perhaps they will start converting declassified documents to random fonts and sizing, like the cliched ransom notes.  Maybe they will start censoring random words or entire paragraphs, a la Yossarian.