Small Step For Spam, Giant Leap For AI?
from the spam-for-hawaiians-only dept
While anti-spam software companies try to claim successes, most people still point to their failures. The problem with spam filters still revolves around the tricky determination of spam versus real email. If the filter is a little too twitchy, you wind up with mysteriously missing emails that you actually wanted to read (false positives). However, to avoid those false positives, you'll probably let more real spam (false negatives) through, which defeats part of the reason for using a filter. Although it seems like the spam problem isn't going away, the battle between spammers and automated spam filters could produce advances in artificial intelligence as spam provides a sort of Turing test. So the solution to the spam problem may be just ten years away, because everyone knows AI is always ten years from being self-aware.
1 Comments | Leave a Comment..
- Why Apple Will Not Be Part Of The Real Tablet Revolution
- DailyDirt: We've Got The Droids You're Looking For...
- DailyDirt: No 9000 Computer Has Ever Made A Mistake Or Distorted Information...
- State Department Spent $1.2 Billion On An Asset Monitoring System... That Ignores All Non-Windows Equipment
- Feds Say They Can Search Bradley Manning's Friend's Laptop Because They Can





Reader Comments (rss)
(Flattened / Threaded)
uh, right...
[ reply to this | link to this | view in thread ]
Add Your Comment