From: Stathis Papaioannou (stathisp@gmail.com)
Date: Sun Nov 11 2007 - 01:49:51 MST
On 11/11/2007, Wei Dai <weidai@weidai.com> wrote:
> If contemplating these questions is an optimization process, it's an
> optimization process with an unknown objective function, and the objective
> function seems to change over time (i.e., what we consider satisfactory
> answers change as we learn new things and come up with new insights). But
> then any process can be described as an optimization process for an unknown
> objective function that changes over time, so the concept would have no
> explanatory or practical value. I think when we talk about optimization
> processes, we usually mean those with known and fixed objective functions.
An intelligence must have *some* idea of what the criteria for a
satisfactory answer to a question might be, even if it is something as
vague as "it just seems right"; otherwise the question reduces to a
request for random information. If these criteria could be formalised,
that could be used in the objective function of an optimization
process to answer the question, along with information on how the
criteria should change in the light of new learning. Is it wrong to
speak of an optimization process where the objective function is
subject to revision?
-- Stathis Papaioannou
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:00 MDT