To begin with, I'm going to reiterate what others have said about your
argument being cloaked in technical terms without enough context to
give them utility.
On 14-Feb-11, at 9:17 PM, Anand Manikutty wrote:
[snip]
There is some detail in these subsequent emails (Suresh, Venky,
Udhay, Sirtaj, ss, Eugen, Vinayak, Deepa), but I am going to pass over
them unless one of you is claiming to be a Singularitarian. But even
if you are, please see below how you ought to proceed.
Yep, and I guess we should all stop discussing politics unless we each
confess to being an MP. Got it. Why are you still carrying on this
conversation?
I have a reference here to Clay Christensen's paper entitled "Theory
Building in Management Research" : http://www.innosight.com/documents/Theory%20Building.pdf
. It is generally applicable to research in all the social sciences
and that includes this theory of Singularity.
The process to be followed for a theory (any theory!) is the
following :
Step 1. Observe, describe and measure the phenomena (The
'constructs' step)
Step 2. Categorization based upon attributes of phenomena (The
'frameworks and typologies' step)
Step 3. Statements of associations (The 'models' step).
In Step 3, various hypotheses are tested. Regression or other
analysis may be used.
Now, anomalies may be observed for a theory. In such a case, the
theory must be revisited and revised as necessary.
Oh good grief.
Nobody so far has responded to the question of a canonical set of
hypotheses. Eugen offers the idea of a feedback loop, but that is
not a baseline hypothesis. What I want to see is an academic paper
that clearly spells out the hypotheses and how they have been
tested. Furthermore, Eugen has not answered the question of the
elasticity of substitution. So because he is unfamiliar with this,
he is unable to defend the theory. Note that this is not any old
objection. This objection comes from none other than William Nordhaus.
What "question" re: elasticity of substitution? It is as much a
statistical heuristic as anything else, and therefore a viable fitness
function for a GA. If you can clearly state the question (rather than
just throw out "but what about elasticity of substitution?" as if the
term itself provides all possible relevance), perhaps someone can help.
-Taj.
