- From: Bullard, Claude L (Len) <len.bullard@intergraph.com>
- Date: Wed, 29 Jun 2005 08:44:23 -0500
- To: 'Jonathan Borden' <jonathan@openhealth.org>
- Cc: 'www-tag' <www-tag@w3.org>
The point being humans tolerate Turing Test failure better because they know it really doesn't matter to them. If however, you are setting up machines to do this, it might matter depending on what you expect them to do with an answer. So don't fire on first notification. The Turing Test isn't interesting to the SemWeb unless a machine tests another machine. It doesn't matter unless you intend to hook up fire control systems to the semweb. Don't do that. URIs can be ambiguous. URIs are information resources. len From: Jonathan Borden [mailto:jonathan@openhealth.org] Bullard, Claude L (Len) wrote: --- Humans fail Turing Tests consistently. The test is only interesting if the machine can tell that the resource is an information resource. That would be a test of the architecture. --- The point being that the definition of an information resource as one for whom one's HTTP URI returns a 200 in response to a GET is a perfectly reasonable one. One should not assume that this is *at all* inconsistent with Mark Baker's assertion that http://www.markbaker.ca identifies *him*.
Received on Wednesday, 29 June 2005 13:44:31 UTC