Mind map
Consciousness is inference (Friston)
Map content
Consciousness is inference (Friston)
[1] We are thrown ito the world [out there] as a PROCESS already in motion
- Processes that matter here: PROCESSES making up COMPLEX SYSTEMS
[2] Knowing LYAPUNOV FUNCTION [LF] for each state of a system --> know flow of the system between states
- Example: water stream through a mountainous landscape
- Lyapunov function: montainous landscape
- Description of evolution of system: movement of stream of water
- Example: water stream through a mountainous landscape
[3] COMPLEX SYSTEMS use their LF FUNCTION to move towards ATTRACTORS (small number of most probable states)
- = repetitive, SELF-ORGANISING behavior
- COMPLEX SYSTEMS are SELF-ORGANISING because they have ATTRACTORS
- ATTRACTORS: cycles of mutually reinforcing states allowing PROCESSES to reach a DYNAMIC EQUILIBRIUM
[4] HUMANS = "strange loops" or "playful loops" (Douglas Hofstader)
- We move through an enormous, high-dimensional state-space of possibilities, but are forced by our ATTRACTORS to move around in confined circles (= pattern of trajectory of an automn leave falling)
- To be ALIVE --> all our PROCESSES must DECREASE the LF --> pushing as to ATTRACTORS
- Our PROCESSES = thoughts & behaviors
- To be ALIVE --> all our PROCESSES must DECREASE the LF --> pushing as to ATTRACTORS
- HUMANS = COMPLEX SYSTEM with a set of ATTRACTING STATES
- We move through an enormous, high-dimensional state-space of possibilities, but are forced by our ATTRACTORS to move around in confined circles (= pattern of trajectory of an automn leave falling)
[5] Interpretations of LF
- I) from INFORMATION TH.
- LF = SURPRISE
- II) from STATISTICS
- LF = (NEGATIVE) EVIDENCE
- = MARGINAL LIKELIHOOD
- = probablity that a given model acounting for that state is correct
- = staying alive --> II) we increase our model evidence (SELF-EVIDENCING) = minimise surprise
- = probablity that a given model acounting for that state is correct
- = MARGINAL LIKELIHOOD
- LF = (NEGATIVE) EVIDENCE
- [6] --> EXISTENTIAL DYNAMICS with PURPOSE & TELEOLOGY
- INFERENCE = maximising evidence for a model of the world (for me!)
- The model is OURSELVES! (i.e. we evaluate what the world might be doing to us)
- INFERENCE: what any COMPLEX PROCESS repeatedly occuppying certain states performs
- [6.1] INFERENCE in NATURAL SELECTION: SELECTING among CREATURES
- Example [evolution in terms of inference]: PHENOTYPES of a population (of a species) = multiple HYPOTESIS OR MODELS about what might 'work'
- EACH INDIVIDUAL = MODEL of what should occupy the corresponding ecological niche
- Example [evolution in terms of inference]: PHENOTYPES of a population (of a species) = multiple HYPOTESIS OR MODELS about what might 'work'
- [6.1] INFERENCE in NATURAL SELECTION: SELECTING among CREATURES
- INFERENCE: process of figuring out the best hypothesis that explains the observed states of the SYSTEM we call the WORLD
- INFERENCE = maximising evidence for a model of the world (for me!)
- I) from INFORMATION TH.
[7] INFERENCE IN CONSCIOUSNESS
- Conscious processing: inferring the causes of sensory states
- --> navigate the wold to elude surprise
- --> SELECTING among different STATES of the same creature (in particular, its brain)
- brain = self-evidencing organ of inference
- Conscious vs non-conscious creatures
- how they make inferences about ACTION & TIME
- To determine a future action, the system must have internal models with TEMPORAL THICKNESS
- --> COUNTERFACTUAL DEPTH
- Future action --> minimisation of EXPECTED surprise following the action
- VIRUSES are unconscious: they don't have a THICK MODEL OF TIME
- To determine a future action, the system must have internal models with TEMPORAL THICKNESS
- how they make inferences about ACTION & TIME
- CONSCIOUSNESS = inference about my future
- Conscious processing: inferring the causes of sensory states
Information
- Description
- Deduction of consciousness as a process of inference in complex systems
- Created by
- Luis de la Pinta
- Views
- 64
- Tags
- Rating