2005
DOI: 10.1007/11553090_75
|View full text |Cite
|
Sign up to set email alerts
|

All Else Being Equal Be Empowered

Abstract: The original publication is available at www.springerlink.com . Copyright Springer DOI : 10.1007/11553090_75The classical approach to using utility functions suffers from the drawback of having to design and tweak the functions on a case by case basis. Inspired by examples from the animal kingdom, social sciences and games we propose empowerment, a rather universal function, defined as the information-theoretic capacity of an agent???s actuation channel. The concept applies to any sensorimotoric apparatus. Emp… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
117
0

Year Published

2006
2006
2010
2010

Publication Types

Select...
4
3
1

Relationship

3
5

Authors

Journals

citations
Cited by 98 publications
(117 citation statements)
references
References 8 publications
0
117
0
Order By: Relevance
“…A measure of such a capacity, referred to as empowerment, has been recently introduced in the framework of information theory as a good candidate for a universal utility in the case of single agent systems (see [6] and [7]). An advantage of this measure is that it can be used as a dense feedback for driving behaviour instead of a sparse feedback such as artificial evolution and reinforcement learning.…”
Section: Introductionmentioning
confidence: 99%
“…A measure of such a capacity, referred to as empowerment, has been recently introduced in the framework of information theory as a good candidate for a universal utility in the case of single agent systems (see [6] and [7]). An advantage of this measure is that it can be used as a dense feedback for driving behaviour instead of a sparse feedback such as artificial evolution and reinforcement learning.…”
Section: Introductionmentioning
confidence: 99%
“…Agents will first consider simple situations before they move on to more complex ones because they choose the situation in which they have the highest chance to increase their predictive success. Other approaches to autonomous learning such as homeokinesis [10] and empowerment [11] have been investigated recently. Here we use another principle also based on intrinsic motivation, originally identified by the humanistic psychologist Csikszenmihalyi, and based on studying the activities of painters, rock climbers, surgeons, and other people who were observed to be deeply involved in some very complex activity for the sake of doing it, i.e.…”
Section: The Autotelic Principlementioning
confidence: 99%
“…Such a methodology may go beyond computational aspects, including sensing, actuation, and networking in distributed systems, co-evolving under multiple design/selection pressures. We intend to follow the proposal on information-driven evolutionary design which suggested to use fitness functions according to generic information-theoretic criteria [21,14]. The identification of possible intrinsic fitness criteria is also related to the work of Der et al on self-organization of agent behaviors from domain-invariant principles, e.g., homeokinesis [3].…”
Section: Background and Motivationmentioning
confidence: 99%
“…An example of a selection pressure is the acquisition of information from the environment: there is some evidence that pushing the information flow to the information-theoretic limit (i.e., maximization of information transfer in perception-action loops) can give rise to intricate behaviour, induce a necessary structure in the system, and ultimately be responsible for adaptively reshaping the system [13,14]. Other important selection pressures applicable to distributed systems include stability of self-organizing hierarchies [19]; efficiency of multi-cellular communication topologies [20]; efficiency of locomotion and distributed actuation [21,23].…”
Section: Background and Motivationmentioning
confidence: 99%