From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Wed Sep 10 1997 - 18:45:00 MDT
I don't think I liked the way Holly Pearson phrased the issues involved. His
impending economic collapse seemed too much like a social conflict, not a
blind consequence of increasing technology. There may be winners and losers,
but they are irrelevant and unintentional. Like most technological
consequences, it is not in the least inevitable; it only takes more technology
to counter.
Our economy will either collapse or transform. *Which* will depend on whether
we're smart enough, and flexible enough, to redesign it along the necessary lines.
I have had thoughts along those lines, and they shone through in "Staring Into
The Singularity". Hence the quote, I suppose.
From "The Economy Of Plenty", an unfinished work: (©1997 by Eliezer S. Yudkowsky.)
-- The Economy of Plenty --
The basic problem is economies of scale. At our level of technology, it only
takes a billion people working to support six billion. And then the other
five billion starve, and you start over with one billion and two hundred
million, until you've got a self-supporting community of ten million people
and the six billion are dead.
There are four obvious solutions. First, give everyone a Universal Standard
Wage. Second, break up humanity into units of ten million and forbid resource
transfers between them. Three, raise the standard of living until there's
enough work for everyone. Four, distribute the work more evenly.
Solution one: In essence, this amounts to dividing humanity into workers and
drones. It differs from Communism in that in Communism, everyone was supposed
to work and nobody did, while here, only a few people will be supposed to
work. I don't think this solution, or any other which involves dividing
humanity, will be stable. The class conflict will destroy it, and the "Why
should *I* work?" problem will still be present.
How to make it work: Instead of dividing humanity, abolish *all* work. Go
the rest of the way from one billion supporting six billion to zero supporting
six billion. Build arcologies which require only a few thousand people,
*total*, to sustain the entire planet at a minimal standard of living. Then
paste the current economy on top of that. If this is stable, the rate of
progress to Singularity will slow down, but we'll have all the time necessary
to get it right the first time.
Solution two: Partition humanity into six hundred economies. This will
falter on limited resources, brains as well as arable land, of which we have
enough to support five or even twenty economies, but not hundreds. For this
planet to support six billion, we need technological economies of scale.
How to make it work: I don't know yet.
Solution three: Raise the standard of living. Why hasn't this happened
already? That is, how can we be running out of work when people are still
poor and hungry? The answer is that these people are unemployed, hence do not
have any money, hence cannot pay for supplies, and hence do not add to demand.
First let's see us raise the demand to match the current supply, much less
the supply available if everyone worked!
How to make it work: This gets kind of complicated, and requires massive
computing power. Essentially, you have to get rid of money and replace it
with a barter system. We *can* do this now because we have enough computing
power to institute "complex barter" systems. This takes some explaining. A
simple barter transaction is one in which A gives B a pineapple, and B gives A a
coconut. A complex barter is where A gives B a pineapple, B gives C a
coconut, and C gives A a watermelon. Note the triple cycle. Complex barter
is too hard to keep track of - without computers, heh heh heh - which is why
we have money, to transform every transaction into a simple barter. By
"money", we here mean something with artificial value, rather than universal
units of exchange. A complex barter system can still have standardized units
of valuation.
What would be accomplished by this? Well, one of the economic design flaws
leading to our current problem is - more than I can finish this sentence with.
Here's what happens. Person A, who builds widgets, builds 100 widgets. B
through K also build 100 widgets. Unfortunately, there is only a market for
1,000 widgets total. But 1,100 have been built.
The standard capitalistic reply at this point is that G, who isn't very good
at this, goes out of business. There are only two problems: First, G can't
get another job, and more importantly, that's NOT what actually happens!
Instead, the price of widgets drops. Like a brick. They trade widgets for
cigar bands on a one-to-one basis. The result is that not even E, who's the
best, can make a living - much less G.
In an ideal world, this overproduction would simply raise standards of living.
Really. Think about it. There are more widgets for everyone! Hooray!
Again, ideally, food would be overproduced. So the price of food would drop
as well. The price of everything would drop, and widgets would be dirt cheap,
but so would be everything else, and A and E and G and K would all be rich.
The problem is that production of food is out of sync with production of
widgets. The current economy was BUILT ON SCARCITY. Everyone, everything, is
centered on the idea that supply is always slightly less than demand. If one
- and only one - field shifts to a massive-supply model, that field suffers.
It's like being the only cooperator in a world of defectors. You need a
massive shift.
And again, the existence of money is the problem. Not private property! I'm
not saying that. The problem is money instead of barter. See, in a barter
system, you can establish valuations that can outlast supply and demand, or at
least weather the turbulence created by switching to an Economy of Plenty.
Suppose that we declare one widget to equal one hamburger in value. Now, even
if there are too many widgets, anyone who's made a widget can still get a
hamburger in exchange. G's widgets might rot, but that's better than all the
widgets rotting. We'll assume, actually, the existence of some buffer, some
temporary charitable system, so that G remains in business for at least a
month or two - while the production of hamburgers increases also.
We can also move up to establish a futures market - even better than barter -
for absolutely everything on the planet. Then even G's widgets don't rot.
Instead, everyone sells 91 widgets instead of 100 - which may sound bad, but
that's on a G-just-went-into-business model. If you assume a technological
overproduction, 91 widgets is just what they sold last year. It's
technological overproduction that's the problem, after all - population
expansion is self-correcting. So everyone has an extra 9 widgets to trade for
whatever they want. Hopefully, someone besides widget-makers is
overproducing, so the surplus can all be traded around.
Eventually, everyone overproduces, and those who produce the most wind up with
the most of the overproduction. So the system has two parts, really: One, a
futures market (applied to EVERYTHING) so that overproduction doesn't cause
prices to drop, and two, a complex barter system - whose necessity is kind of
hard to put into words, but... The value of a dollar is arbitrary. If a
widget-future involves $5 a widget, the value of the contract is affected by
fluctuations in the money supply as well as the supply of widgets. A
widget-future which involves trading one hundred hamburgers for one hundred
widgets may fluctuate all over the place in THEORY, but in practice, one
hundred hamburgers are what the widget-maker will eat this year. In other
words, the universal futures economy is very vulnerable to inflation, which a
complex-barter system will stop.
Solution four: Distribute the work more evenly. How can I say that we're
running out of work, when everyone WITH a job is spending every waking hour on
their job? The problem is that ***
[End of work in progress.]
To summarize #4, you'd need to entirely revamp the structure of a corporation
along certain lines. This would be even more complex than the complex-barter
bit, which itself should really have diagrams. The solution to number four
requires complex barter as a *prerequisite*, and something called
collaborative filtering, and even substantial alterations to American law.
Anyway, that's why I stopped there - I wasn't sure I had it all worked out.
--
sentience@pobox.com Eliezer S. Yudkowsky
http://tezcat.com/~eliezer/singularity.html
http://tezcat.com/~eliezer/algernon.html
Disclaimer: Unless otherwise specified, I'm not telling you
everything I think I know.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:44:52 MST