>Paul,
>
>Decision options don't have to be deterministic at all.
Of course they don't.
I mentioned how we have developed the mathematics for a new class of
stochastic optimization methods which, for
the first time, overcome the classic curse of dimensionality which
has limited well-known
classical methods like decision trees.
However, I also mentioned a basic feature of these methods --
under situations of serious uncertainty (both good uncertainty and
bad, hopes and fears),
the optimal strategy which pops out is often NOT a rigid future plan
or rigid commitment, say, to
a single energy source for the entire world. Even simple decision
trees lead us to things
like wildcat drilling, buying information, and keeping options open
to the extent that
this does not lead to paralysis or unaffordable delay. More sophisticated
intelligent systems can account for the fact that there is OTHER INTELLIGENCE
out there as well, such as that of energy producers; if we make
decisions without
accounting for that reality, we are acting on a false premise.
One question in that vein: is it possible that there is a creative
chemist out there who
knows a noncarbon liquid fuel better for cars than hydrazine hydrate
(HH)? Perhaps even in
Exxon Research itself? I wouldn't assert that there is, but it's a
possibility worth
allowing for -- along with a lot of other possibilities.
But still I have to admit that I see more hope for better batteries
right now than for
anything else to store energy for cars and trucks. As an example,
someone really needs to follow up on the
kind of new directions proposed by Sadoway last week... click on Sadoway at
http://www.ieeeusa.org/policy/phev/program.asp..
Best of luck to us all,
Paul
No comments:
Post a Comment