[ExI] Economics 2.0
Anders Sandberg
anders at aleph.se
Wed Oct 10 22:18:47 UTC 2012
Yes, I found the 4% quote program downright disturbing. Especially since
I saw the story while listening to a colleauge who runs a research
project looking at using software proof techniques to understand the
behavior of economic software agents... and he was explicitly claiming
(and defending!) that they were rational and would produce equilibria
behaviors!
OK, his rationality argument was not 100% stupid: irrational agents will
lose money and will be turned off. But as we have seen in finance,
systems can make money most of the time and then occasionally fail with
super-losses that outweigh the previous profit. So it hinges on the
long-term rationality of owners, which is empirically suspect thanks to
past history and drunk traders. Plus that we have no reason to think the
system is stable since new agents and agent-agent behaviors occur.
Shades of the delightful amazon.com agents that tried to sell a used
book for a few million dollars. Or try to win the Turing test:
http://carlos.bueno.org/2012/02/bots-seized-control.html
This is actually an area where even I, a crazy libertarian, think it
would be smart to add friction to the system to get the quoting down.
But the amazon bots are even closer to the Economics 2.0 in the novel.
As Carlos Bueno put it: "So with “Turing Test” we have a delightful
futuristic absurdity: a computer program, pretending to be human,
hawking a book about computers pretending to be human, while other
computer programs pretend to have used copies of it. A book that was
never actually written, much less printed and read."
In many ways that is more sinister. A stock market crash is just
economic's way of telling you to be more cautious and invest in
something else. These things on the other hand are spamming our markets
and intellectual worlds with pseudo-information of little or no value.
--
Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University
More information about the extropy-chat
mailing list