A Challenge

by Lyle Burkhead <LYBRHED@delphi.com>

This is an idea that goes back to 1987. I have told a couple of people about it, but never really announced it. I think it is an idea whose time has come.

I want to introduce a new game, called Checkmate (aka hyperchess). It is the same as chess, with two exceptions:

1. There is no such thing as a draw or a resignation in Checkmate. You always play the game to a conclusion. Either it ends with a mate, or it is a null game, in which mate cannot be forced. A game is not declared to be null until it is obviously and provably null.

2. The players are teams, consisting of humans and computers working together. There can be as many humans as you want -- as many as you can get to work together without getting in each other's way. Coordinating their efforts is part of the challenge of hyperchess. A team can either be led by a computer, which makes the moves with advice from the humans, who are in turn using other computers with specialized functions; or a team may be led by a human, who plays with the advice of other humans, all of whom are using computers. Another possibility is that a team may not have a leader at all. And of course it is possible to have a minimum team of an individual human and one computer -- with either the human or the computer acting as lead player.

In other words, your task, as a player of hyperchess, is to design and build a SuperIntelligence. How you do it is up to you.

I want to issue a challenge to the Extropians: starting a couple of years from now, let's have an annual Checkmate tournament.

The interesting thing about this is that designing a SuperIntelligence requires designing a mind. How do you do that? This is basically a philosophical problem. You (the Extropian team) can use Ayn Rand's metaphysics, Ian Goddard's logic, Max More's philosophy of mind, Marvin Minsky's insights into AI, Tim Freeman's expertise in computer science, and so forth. There are many people on the list who could make theoretical contributions, and many programmers who would enjoy writing the code that will have to be written.

Meanwhile I will also be designing a SuperIntelligence, using my metaphysics, my logic, and so forth, and recruiting a team to help me write the code and play the game.

I want to extend this challenge to other memes as well. For example, I'm going to go to meetings of some Campus Christian groups, at various colleges, and tell them about this. They can design a SuperIntelligence based on their metaphysics (monotheist), their logic (Thomist), their philosophy of mind, and so forth, and they can use Christian chessplayers to play the game and Christian programmers to write the code. (Christian programmers do exist.) This idea is not new to them -- according to the Bible, the church itself is a kind of SuperIntelligence, with Christ as the head and the church as the body.

There are various other memes that might want to enter teams. IBM can enter Big Blue, assisted by human players.

May the best SI win. May the best philosophy win.


Up to the AI Page

Up to the Transhuman Page

Anders Main Page

Anders Sandberg / asa@nada.kth.se