Papers
Topics
Authors
Recent
2000 character limit reached

Why Is Anything Conscious?

Published 22 Sep 2024 in cs.AI | (2409.14545v5)

Abstract: We tackle the hard problem of consciousness taking the naturally selected, embodied organism as our starting point. We provide a formalism describing how biological systems self-organise to hierarchically interpret unlabelled sensory information according to valence. Such interpretations imply behavioural policies which are differentiated from each other only by the qualitative aspect of information processing. Natural selection favours systems that intervene in the world to achieve homeostatic and reproductive goals. Quality is a property arising in such systems to link cause to affect to motivate interventions. This produces interoceptive and exteroceptive classifiers and determines priorities. In formalising the seminal distinction between access and phenomenal consciousness, we claim that access consciousness at the human level requires the ability to hierarchically model i) the self, ii) the world/others and iii) the self as modelled by others, and that this requires phenomenal consciousness. Phenomenal without access consciousness is likely common, but the reverse is implausible. To put it provocatively: death grounds meaning, and Nature does not like zombies. We then describe the multilayered architecture of self-organisation from rocks to Einstein, illustrating how our argument applies. Our proposal lays the foundation of a formal science of consciousness, closer to human fact than zombie fiction.

Citations (2)

Summary

  • The paper presents a formal framework that explains consciousness as emerging from self-organizing biological systems responding to valence-driven sensory input.
  • It utilizes rigorous cybernetic and Bayesian methods to illustrate a progression from basic causal identities to higher-order meta-representational selves.
  • The research challenges philosophical zombie models by demonstrating that phenomenal consciousness naturally arises as a functional, adaptive trait.

Why Is Anything Conscious?

Introduction

The paper "Why Is Anything Conscious?" probes the "hard problem" of consciousness, taking the embodied biological organism as the foundational unit of analysis. It argues for a formalism in which biological systems self-organize to interpret sensory information hierarchically, motivated by valence. This self-organization engenders a hierarchy of consciousness, from first-order causal identities to higher-order self-modeling. By explaining consciousness as a continuum rooted in survival and adaptation, the paper critiques the notion of philosophical zombies, asserting that phenomenal consciousness is intrinsic to functional consciousness.

Biological Foundations and Self-Organization

The authors begin by discussing self-organization, a concept intrinsic to systems manifesting order without central control, widely applied in cybernetics and cognitive science. In biological systems, self-organization helps maintain homeostasis and achieve reproductive goals despite external perturbations. They delineate a framework where the natural selection of embodied organisms cooperatively constructs abstractions necessary for survival.

Biological systems differ from inert systems like snowflakes by virtue of their ability to self-organize in response to environmental pressures to preserve internal order. The authors employ a rigorous formalism to describe these systems, drawing parallels with known cybernetic principles. This is achieved by defining biological organisms as "dissipative systems," energetically open systems that perpetuate themselves by counteracting entropic decay [Friston 2010].

Causal Identities and Psychophysical Causality

In tackling the notion of causality, the authors address causal identity, which relates to how entities influence their environment through intervention. Using the psychophysical principle of causality, the authors eschew the assumption of a priori known causal objects and instead propose that entities are constructed through their cause-and-effect relationships informed by valence.

This causal learning process is illustrated through a Bayesian framework where artificial agents, like Bob, falsely attribute weather changes to interventions like wearing a raincoat. The formalism emphasizes the insufficiency of purely observational metrics without a deeper understanding of the complex dynamics of intervention and interaction with the system [Bennett 2023c].

Scalability of Self and Consciousness

The formalism supports multiple orders of consciousness, illustrating a progression from first-order selves—where an organism develops a sense of reafferent feedback loops essential for subjective experience—to second- and third-order selves capable of meta-representational thought and theory of mind.

In biological terms, a first-order self embodies the capacity to predict the consequences of its own interventions, an echo of what Merker describes as the evolutionary prototype of subjective experience rooted in neural structures common across species [Merker 2005]. A second-order self introduces a theory of mind—an ability to interpret another's perspective, vital for complex social interactions and Gricean communication models [Grice 1969]. The paper argues that such capacities are not merely additions; they restructure the scope of consciousness qualitatively.

Conclusions and Implications

Concluding, the paper integrates these insights into a broader narrative against the notion of philosophical zombies, entities hypothetically mirroring human behavior without subjective experiences. Phenomenal consciousness is posited as functionally indispensable, thus disputing the epistemic foundation of such zombie theses.

This novel approach lays the groundwork for a formal science of consciousness, in which the qualitative experience is tightly coupled with functional capabilities, driven by evolutionarily derived valence interpretations. The research posits that consciousness emerges naturally from adaptive biological mechanisms and reframes the hard problem as a misconception of representation abstracted from its qualitative roots in valence and survival.

In summary, the research compels a re-examination of traditional dualistic views and invites future work on unpacking the scalability of consciousness frameworks within a mathematically formal and evolutionarily grounded paradigm.

Paper to Video (Beta)

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Collections

Sign up for free to add this paper to one or more collections.

Tweets

Sign up for free to view the 28 tweets with 1208 likes about this paper.

HackerNews

  1. Why Is Anything Conscious? (6 points, 1 comment) 
  2. Why Is Anything Conscious? (5 points, 2 comments) 

Reddit

  1. Why is anything conscious? (6 points, 25 comments) 
  2. Why Is Anything Conscious? (1 point, 0 comments)