There are a number of distributed computing projects that use idle computers to help solve computational problems. (Seti@home is the most famous one that I’ve contributed to.)
I wonder why there aren’t any distributed projects that attempt to create a distributed semantic processing system. (Here’s a book that I’m putting on my to-read list.)
Unfortunately, I’m not even certain what a “distributed semantic processing system” would look like, or how it would perform. I imagine such a system would be able to provide correct answers to the following questions:
- Does a fire cause pain?
- Does a fire truck cause pain?
- Does a camp fire cause pain?
But how the system would be designed and seeded with information, I presently have no idea.
I suspect that a true test of a semantic processing system would be its ability to respond to a silly question with a joke:
Question: Does a fire truck cause pain?
Answer: Only if it runs you over.
But that is mere conjecture.