Apparently, I write like a robot. Slashdot reports that researchers at Indiana University have written a program designed to detect whether a given piece of text was written by a human or the result of a bot, designed to write for you.
For those that don't know, there are several bots that will write papers for you on various subjects. There is the Computer Science Paper generator, and the Post Modern Generator which is hilarious (just hit refresh to get a new paper). There are probably others but you can google for them yourselves.
The Inauthentic Paper Detector, which assumes that anything rated over 50% is authentic thinks I am a computer program. I ran my previous two blog posts and one old post through the detector and here are the results:
- What Else Won't Work? - The detector gave me a 23% chance that it was written by a human.
- Apple And Free Speech - This rated only 22.5%. I'm starting to feel dehumanized.
- On Liberty - This is one of my favorite posts so I thought I'd try it. It only rated this post as having a 12.1% chance of being authentic. Can you believe that?
is intended for detecting whether a technical document is human written and authentic or not. Predictions may work for documents intended for entertainment (novels, news articles etc.). The main purpose of this software is to detect whether a technical document conforms to the statistical standards of an expository text. You can easily take a human written technical document and add some nonsense text somewhere in the middle, or paste a document generated by an automatic paper generator. We are trying to detect new, machine written texts that are simply generated not to have any meaning, yet appear to have meaning on the surface.
It's just funny that it thinks that my posts make sense on the surface but are otherwise devoid of real meaning. I've been told exactly the opposite.
Run your own blog post through the detector and post the results in the comments.