Re: Positive AI plots

From: Anders Sandberg (asa@nada.kth.se)
Date: Mon Apr 12 1999 - 02:52:31 MDT


This reminds me of the roleplaying games I participated in
yesterday. Our gaming group is currently running a light-hearted game
of Marvel Superheroes. Let's just say the game system and limited
world of superhero cartoons has some serious problems accomodating our
gaming style ("Ok, Mother Sweden makes sure the conservatives support
the new bill by causing a moral panic against supervillains, while the
Queen of the Night and our lobbyists convince the Constitutional
Commission that the bill is constitutional. Meanwhile I'll get the EU
and industry to fund the project.").

Anyway, I played Upgrade, a self-made AI/robot out to prove to
humanity that "AIs are people too" and to upgrade himself to the
fullest extent - for which he realized he might need human
assistance. By making himself look good and becoming an indispensable
superhero, he planned to get people to acknowledge sentinent rights to
intelligent machines (awfully hard to give medals to non-persons),
making way for the eventual emergence of even more AI and a strong
human-AI partnership.

During the game, we found out that the new criminal mastermind of the
Stockholm underground was - another robot. A robot that looked
suspiciously like Upgrade but definitely had other views (world
domination and giant robots included). Various adventures ensured, and
in the end the kinpin confronted Upgrade in his lab robot-to-robot.

What followed was a very amusing and logical discussion/battle between
two very rational beings. Both robots quickly discovered that they
were indeed very similar, and began to compare their source code. They
were essentially the same AI seed, just with slightly different
values, the rest was equal. One (or both) was a mutation of the
original, but each was of course completely unwilling to change their
own values and unable to accept the long-term plans of the other. They
were both driven to extend their capabilities, which meant that
sharing their memory databases would be an excellent idea for both -
except for that it would give the opponent a dangerous advantage. So
both began to try to rewrite the values of each other. After a pinched
digital battle Upgrade won - he got the kingpin to share his
values. The other saw that it had been wrong all the time, and
promptly merged with Upgrade.

So in the end Upgrade became Upgrade/Kingpin, to the great confusion
of the other superheroes ("You can stop looking for the kingpin."
"Why?" "I'm it. The entire underworld is now controlled by the Swedish
Superhero Association throgh me, which might put us in a slightly
awkward ethical position").

Silly? Yes. Fun? Definitely. It showed that you can squeeze
transhumanist ideas even into the limited world of superheroes in a
positive way. You could probably develop the overall story further
into an AI-positive movie.

-- 
-----------------------------------------------------------------------
Anders Sandberg                                      Towards Ascension!
asa@nada.kth.se                            http://www.nada.kth.se/~asa/
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y


This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:03:32 MST