From: Eugene Leitl (eugene.leitl@lrz.uni-muenchen.de)
Date: Thu Aug 31 2000 - 09:49:43 MDT
Robin Hanson writes:
> Under standard accounts, decisions are made by combining positions
> on values and beliefs about facts. One considers all the possible
> relevant states of the world, and beliefs about facts are about how
> likely each state is given each action, while positions on values
> are about how much one would like each state compared to the others.
Of course we have multiple problems here:
1) we must measure the current state of things, using a patently
unreliable instrument: ourselves
2) the world is a very large and highly complex place, clearly
we can't claim to have sampled its state exhaustively.
Sometimes (I suspect pretty often) things are so muddied,
that even supposedly rational strategies are worse than a
purely randomly chosen strategy. Being super-rational might
be bad for you, at least on the short run.
3) to choose the best outcome (assuming, the outcomes we seek are
identical, which is certifiably not true) from countless future
trajecotries we must extrapolate these highly sketchy
measurements using a most vague internal model of the world
(a world which is not only nonlinear and hence fundamentally
long-term unpredictable), but often even perverse from round
to round, as artefacted by co-evolution.
So even if we agree on the goals (which we must communicate,
introducing another source of errors), due to above reasons we will
not adapt a homogenous strategy (which, incidentally, would be
predictable, and hence potentially exploitable -- just as pests blight
the crop monoculture more easily than an alpine meadow).
Surprise, surprise: diversity is good for you.
> People on this list seem to disagree with those not here about many
> decisions, so it can make sense to ask: do those disagreements tend
> to be more about values, or more about facts?
Of course our values are different (albeit less heterogenous than from
a randomly picked Jane Doe), but we even tend to disagree about these
pesky facts. Speaking of which: let's talk about guns.
> There is less scope for being "right" in disagreements about values.
> Once we understand what we want, and opponents decide they don't want
> that, there isn't that much more to say to them. "Advocacy" then
> consists mostly in seeking out people who haven't considered the
> issue and getting them to pick sides. And if there end up being
> lots more of them than of us, we may just lose big time.
Sure, that's possible. Right now most people adhere to traditional
values (I just had an argument with a physicist at the gym today, as a
Christian and believer in an immortal soul he had a distinctly
different view on the matter of transhumanism).
> There is more to say when we and opponents disagree about facts,
> because in that case we fundamentally share the same interests, and
> are just trying to deal with asymmetric information. But in that
> case we should take very seriously the fact that others disagree
> with us. Even if they haven't articulated their reasoning to our
> satisfaction, we have also failed to do so for them, and assuming
> they aren't stupid we have to realize that they may know things
> that show why we are wrong.
Hey, one should always reexamine one's beliefs. While my core values
are probably pretty much permanently write-protected the rest of it is
most assuredly not. I think I'm even trying to be over-flexible,
trying to compensate for onsetting age value calcification.
> So which is it, do our disagreements with opponents tend to be more
> about values, or more about facts?
I'd say it can only be decided from case to case. Duh.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:30:41 MST