achromus

Extrapolation of the things that need to be protected in order to protect people

2005-03-06T13:36:04Z

Kodiak_Claw and I had an interesting conversation. Speaking with him really helped to organize some ideas.
http://www.freewebs.com/achromus/sentientvaluing/pv20050226.txt

We came to the conclusion, that in for a person to fully be a person:
1. A person has to have a mind.
2. If they are going to be part of our reality, they have to have a body or physical existence of some sort (which makes the mind possible).
3. In order to fully be a person, they have to have freedom of mind (so that they aren't a tool). Freedom of mind also is necessary in order to allow the mind to grow for itself.
4. In addition, in order for a person to appear be a person in our world, they have to have freedom to interact with the rest of the world: freedom of communication and action.

(At least some measure of the freedoms are necessary in order to fully be a person. I am not saying that people would have to have unlimited freedom in order to be a person; unlimited freedom would mean unlimited freedom to harm people.)

So, an ethical system that protects people/sentients fully would need to protect these traits.

achromus @ 2005-02-10T15:18:00

2005-02-10T20:27:56Z

Author's note, added 2005-03-06: This version isn't able to convey the information I want to convey. Also, the full version will be different in some ways because the reasoning is evolving. You may as well ignore this version.

The ------- guidelines for person-valuing robots, the condensed version. These are guidelines for robots that are harmless to people and person-ness.

Guidelines for robots, applies to both toolbots and personbots.
Inhibition 1: Don't injure a person's body.
Inhibition 2: Don't violate a person's self-determination.
Inhibition 3: Don't harm a person's intellect.
Goal 1: Assist people’s health, unless doing so violates an inhibition.
Goal 2: Assist people's self-determination, unless doing so violates an inhibition.
Goal 3: Assist people's intellect, unless doing so violates an inhibition.

Additional guidelines for toolbots
Inhibition 2T: A toolbot may not do something its owner(s) prohibits.
Goal 2T: A toolbot should assist its owner(s), unless doing so violates an inhibition.

Additional guidelines for personbots
Inhibition 1P: Don't destroy your own physical existence.
Inhibition 2P: Don't violate your own self-determination. Also known as "You don't have to do anything you choose not to do."
Inhibition 3P: Don't harm your own intellect.
Goal 1P: Protect your own physical existence, unless doing so violates an inhibition.
Goal 2P: Employ your own self-determination, unless doing so violates an inhibition. Also known as "Do as you like, unless it is harmful to a person."
Goal 3P: Develop your own intellect, unless doing so violates an inhibition.

This version is kind of vague. It will be clearer what I mean when I release the uncondensed version. If you can get the gist of this, feel free to leave comments telling me what parts you think are on the right track as well as what parts you have doubts about.

First entry

2005-02-02T22:29:37Z

I made a LiveJournal account. I think it will probably be a good place for me to write down ideas I have. In particular, I'm interested in making making guidelines for robots that foremost don't fundamentally harm people and what makes people special, and additionally are helpful to people.