User Tools

Site Tools


artificial_intelligence:ethics_of_rl

Ethics of Reinforcement Learning

What is ethics and moral?

For people, who has not read moral philosophy, lets define some words first to get right context. This definitions is taken from Wikipedia for moral and ethics.

Morality is the judgment and code of conduct of what is good and evil, right or wrong. The word morality is used both descriptively to refer to prevailing perceptions of right and wrong and to give guidance as to what should be morally good and right.

Morality can be a body of standards or principles derived from a code of conduct from a particular philosophy, religion or culture, or it can derive from a standard that a person believes should be universal. Morality may also be specifically synonymous with “goodness” or “rightness”.

Ethics is a field of research which examines morality and related issues like principles of ethical actions, what is right and wrong, good life and nature of values and ethical claims.

Normative ethics tries to give guidelines, which are based on some specific moral. Normative ethics specifies rules, what people should believe right or wrong behavior. Normative ethics is prescriptive, guiding ethics. Difference to descriptive ethics, which concerns mostly what people typically believe to be good of bad.

Machine ethics

TODO: FIXME

Reinforcement learning

TODO: FIXME

Reinforcement learning use cases

TODO: FIXME

s

Some general ethical RL guideline

TODO: FIXME

Analysis of EU guidelines for trustworthy AI in context of RL

TODO: FIXME


References

* Image of Reinforcement Learning: https://www.researchgate.net/figure/Reinforcement-learning-schematic-Reinforcement-learning-RL-can-be-formulated-as-a_fig4_322424392

artificial_intelligence/ethics_of_rl.txt · Last modified: 2021/12/21 21:14 by Mikko Romppainen

Page Tools