GReader Shared Posts

Wednesday, January 30, 2008

Collected Human Intelligence

I listened to two podcasts this morning that gave two very interesting perspectives. I'm now trying to reconcile them.

First was On Point with Tom Ashbrook. It was an episode about the movie that premiered at Sundance -- The Linguists. They included the statement:

Most of what humans know isn't written down anywhere. It's simply in people's memories. (Timestamp - 6:30)
Second was Escape Pod with Stephen Ely. It was the most recent episode -- Artifice and Intelligence. A fairly standard story about an emergent, sentient AI with the statement:
He could create a convincing imitation of an artificial intelligence, with access to the sum of human knowledge online... Edgar believed a ghost driven AI could operate on the same level as a real machine intelligence. (Timestamp - 4:05)
Here are the questions that I'm wrestling with:
  1. How would an AI parse all of the conflicting information online to develop a comprehensive picture of 'human intelligence'?
  2. A machine wouldn't have access to all our daily offline activities that give our lives context. Even with all the information posted to the Internet, how well would an AI be able to converse with humans?
  3. Does the adoption of more 'conversations' mean we are providing more context for a future AI entity to learn how to be more human?