Daniel Hillis and Bran Ferren at Google filed a patent in March of 2010.  Normally I don’t spend a lot of time focusing on such matters, but when a patent filing actually mentions MY personal invention, I get a nice smile.  That smile gets a little bigger when the patent is actually incorrect!

Specifically, the filing says ‘the system determines a numerical score for each comment ranging from 1 to 5.’ which is actually wrong.  Slashdot comments are scored from -1 to 5, but just seeing the Slashdot mention in such a document is awesome.

The filing goes on to say  “because there is no restriction on the users that may participate, the reliability of the ratings is correspondingly diminished.”  In Slashdot’s case, the restrictions on the users are actually fairly complex…  You need to create a user account, use it regularly in accordance with certain typical usage patterns, and when granted moderation access, you must use that access reliably in order to continue to get access  To say that there is “No Restrictions” is a ludicrous mis-statement in the case of Slashdot at least…. in fact, in order to participate in Slashdot’s moderation system you need to be that most elite of human: A Slashdot User… or at least someone willing to accurately fake it for awhile!

As for the claims of ‘diminished reliability’, that seems pretty speculative to me.  We tried a number of variations on our moderation system in the early years, and I found that the most significant factor in diminished reliability was simply to let people have infinite moderation powers all the time.  They just get bored and stop, occasionally after getting drunk with undeserved power and abusing others

Ironically enough, the Slashdot TAGGING system might actually be better prior art on this particular patent.  It was designed to do a lot of what is described in this filing, although I’ll admit that my ability to read a patent-ese is sub-par.

The Slashdot tagging system was worked out years ago by myself and Jamie McCarthy.   It was designed to sort of reach consensus in a community to decide what tags should appear on Slashdot stories.  Users have weight within the tagging system… as user #1, I had the maximum weighting, and over time, my weight was distributed to other users who reliably either reinforced or predicted my actions.  Essentially a pyramid scheme were I was in effect “Delegating authority to evaluate content”, although in this case the content was 5-15 letter tags and not whole stories.  Eventually this information became part of our story selection process. If you’ve ever seen the color coding on stories, that’s a big part of how it all works.

The longer term intent behind the tagging system is that it would ultimately replace the moderation system: it was designed with a number of really awesome possibilities not present in the original system mentioned in this filing.  We had dreams of blending the tagging system with the moderation system to achieve subject specific moderation labels (like letting the politics stories be tagged left/right/independent or letting idle section stories have tags that let you further refine the “funny” label.

Anyhoo, thanks for the smile Googlers Hillis and Ferren!  If you get bonuses, I’ll take a beer!