academic course: 01 self-awarenesss and computational self-awareness

23
Designed by Peter Lewis Self-awarenesss and Computational Self-awareness Introduction to the principles of self-awareness

Upload: fet-aware-project-self-awareness-in-autonomic-systems

Post on 12-Jan-2015

269 views

Category:

Technology


0 download

DESCRIPTION

Introduction to the principles of self-awareness by Peter Lewis

TRANSCRIPT

Page 1: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Self-awarenesss andComputational Self-awareness

Introduction to the principles of self-awareness

Page 2: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Overview of this lecture

•Some notions of self-awareness.

•Types of self-awareness.

•A framework for computational self-awareness.

Page 3: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Self-awareness

•Term first appeared around the start of the 20th century.•Emerging field within psychology, 1960s-1980s.•Various definitions, e.g.

“The capacity to become the object of one's own attention” (Morin, 2006).“...knowledge of oneself as a defined entity, independent of other individuals...” (Legrain et al., 2010).

Page 4: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Implicit and explicit self-awareness

•Implicit self-awareness:The self as “I”.The self is the “subject of experience”.Ability to perceive that there is something unique about what was experienced, which differentiates it from the surrounding environment.Things experienced are uniquely situated in relation to oneself.Can explore experiences as responses to actions, explore interactions over time.See Rochat (2003) and Legrain et al. (2010) for more.But some psychologists (e.g. Morin, 2006) argue that very simple implicit self-awareness to be consciousness but not “genuine” self-awareness.

Page 5: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Implicit and explicit self-awareness

•Explicit self-awareness:A mental representation of “me”.The self is the object of ones attention.The individual is aware that it is awake and experiencing specific mental events, emitting behaviours and possessing unique characteristics.Knowledge of oneself as a defined entity, independent of others.Knowledge that oneself remains stable and consistent over time.See Morin (2006) and Legrain et al. (2010) for more.

Page 6: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Private and public self-awareness

•Private self-awareness:Knowledge of and based on phenomena internal to oneself.May involve awareness of values, goals or behaviour.E.g. I am hungry, my feet hurt.

•Public self-awareness:Knowledge of and based on phenomena external to oneself.Ones perception of experiences in the environment.Subjective.E.g. awareness of relationships to others, how others perceive me.

See Duval & Wicklund (1972) and Goukens et al. (2007) for more.

Page 7: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Levels of self-awareness

Self-awareness is not an “on or off” capability!

•The self-awareness capabilities of an individual can be described as being at one or more levels.

•These levels range from very simplistic capabilities to highly complex ones.

Page 8: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Neisser's levels of self-awareness1. Ecological self

(Awareness of internal or external stimuli).2. Interpersonal self

(Awareness of interactions with others).3. Extended self

(Awareness of time: past and/or future).4. Private self

(Awareness of owns own thoughts, feelings, intentions).5. Conceptual self

(Awareness of ones own self-awareness, possession of an abstract model of oneself).

The conceptual self has the capacity for “meta-self-awareness”, being aware that one is self-aware.

See Neisser (1997).

Page 9: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Emergence of self-awareness

•In collective systems, the entire system can appear self-aware,•Though constituent parts may exhibit less self-awareness themselves.•Self-information is distributed about the system, and not present at any single point.•See Mitchell (2005).

Page 10: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Computational frameworkWe would like to to take some of these ideas, and translate and apply them to the design of computing systems.

Why?

•Provide a common understanding and language for self-aware computing.

•Relate computing concepts to psychological basis – draw inspiration from

natural systems.

•Enable the principled engineering of self-aware systems by identifying

common features and how to build them.

Page 11: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Private and public computational self-awareness

•Private self-awareness: Concerned with knowledge of internal things.E.g.

Internal state / measurements,Current performance, overheads, progress towards ones own goals,Own capabilities, available algorithms.

•Public self-awareness: Concerned with knowledge of external things.E.g.

State of the observed physical environment,Social environment and interactions (i.e. other agents, components, nodes...)Collective goals, others' goals, and progress towards them.

Page 12: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Levels of ComputationalSelf-awareness

•Ecological self → Stimulus awareness•Interpersonal self → Interaction awareness•Extended self → Time awareness•Private self → Goal awareness•Conceptual self → Meta-self-awareness

Page 13: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Stimulus awareness

•Awareness of experiencing stimuli.•Enables the ability to respond to events.•No knowledge of reasons for stimuli.•No knowledge of past / future stimuli.•Stimuli can be internal or external.•E.g. simple reflex agent.

Page 14: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Interaction awareness

•Knowledge that stimuli and actions form interactions.•Knowledge of interactions with others and the environment.•Knowledge that actions can provoke, generate or cause specific reactions.•Simple interaction awareness may enable reasoning about individual interactions.•More advanced interaction awareness may involve knowledge of social structures such as communities or networks.

Page 15: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Time awareness

•Knowledge of past or likely future phenomena.•May involve explicit memory, time series modelling or anticipation.•Relates to both internal and external experiences.•E.g. model-based reflex agent.

Page 16: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Goal awareness

•Knowledge of goals (including progress towards goals)•E.g. goal states, objective functions, utility functions.•May include knowledge of preferences, values, constraints.•Permits acknowledgement and adaptation to changes in goals.•E.g. utility-based / goal-based agent.

Page 17: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Meta-self-awareness

•Knowledge of ones own self-awareness capabilities.•May include knowledge of which levels are present, and how they are implemented.•Permits the ability to reason about the benefits and costs of engaging in certain levels of self-awareness.•Enables online / runtime adaptation of self-awareness capabilities and how they are implemented.

Page 18: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Computational self-awareness taxonomy

Page 19: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Emergent self-awareness:implications for system design

•Systems can exhibit behaviour which appears globally self-aware,•No single component is required to possess system-wide self-knowledge.•Need not require that a self-aware system possesses a global controller!•Sufficient for components just to have local knowledge, of relevant parts.

Page 20: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Multi-level self-aware systems

“Self” is a concept,

not a box.

Page 21: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Computational self-awareness

•To be self-aware, a system should:Possess knowledge of its internal state (private self-awareness),Possess knowledge about its environment (public self-awareness).

•Optionally, it might also:Possess knowledge of its interactions with others and the wider system (interaction awareness),Possess knowledge of time, e.g. past and likely future experiences (time awareness),Possess knowledge of its goals e.g. objectives, preferences, constraints (goal awareness),Select what is and is not relevant knowledge (meta-self-awareness).

Page 22: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Key considerations in computational self-awareness capabilities

•Where systems differ in terms of their self-awareness, is in what knowledge is available and collected, and how it is represented.•Key questions of a system:

Which level(s) of self-awareness are present?How are its self-awareness capabilities implemented?

Page 23: Academic Course: 01 Self-awarenesss and Computational Self-awareness

Designed by Peter Lewis

Summary

•Some notions of self-awareness in psychology.

•Types of self-awareness:Implicit / explicit self-awareness,Public / private self-awareness,Levels of self-awareness,Emergent self-awareness in collectives.

•A framework for computational self-awareness.Short summary paper on self-awareness in psychology and computing: Lewis et al. (2011)