Everyone knows that the government keeps secrets. Described as "matters of national security," many a detail is blacked out of the reports that are released to the public. And whether you think those redacted lines of text are covering up an alien landing at Roswell or a plan for world domination, it sure would be nice to see past the secrecy.
Thanks to a little bit of ingenuity and crowd-sourced data, now we can. The project is called the Declassification Engine and is the brainchild of a trio of Columbia University professors. At present the Declassification Engine is primarily a set of data-analysis tools with the entire Internet as their playground — a powerful and successful tool, even at this stage.
Oh, but there's more. Deeper down the anti-redacting rabbit hole, we find the envisioned future of the Engine: a synchronous platform that not only finds older less-redacted versions of documents, but also has a vast knowledge of terms used in often-redacted documents and can use these tools together to theorize upon what the text beneath never-declassified documents might be. The more the Engine knows, the more it is likely to find out in the future.
While the text revealed in this way is merely a "best guess", it can nonetheless be very accurate, owning to the formulaic way in which governments document their knowledge.
Is the Declassification Engine legal? That's something of a grey area, but its creators do emphasize that they are trying to keep their project on the right side of the law, while simultaneously telling the public those things the government would rather we remained oblivious to. If you'd like to help make the Declassification Engine a potent reality, the project is currently looking for funding, and a lot of it.