Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
120 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Information Flow Theory (IFT) of Biologic and Machine Consciousness: Implications for Artificial General Intelligence and the Technological Singularity (1907.00703v1)

Published 21 Jun 2019 in q-bio.NC, cs.AI, cs.ET, cs.IT, and math.IT

Abstract: The subjective experience of consciousness is at once familiar and yet deeply mysterious. Strategies exploring the top-down mechanisms of conscious thought within the human brain have been unable to produce a generalized explanatory theory that scales through evolution and can be applied to artificial systems. Information Flow Theory (IFT) provides a novel framework for understanding both the development and nature of consciousness in any system capable of processing information. In prioritizing the direction of information flow over information computation, IFT produces a range of unexpected predictions. The purpose of this manuscript is to introduce the basic concepts of IFT and explore the manifold implications regarding artificial intelligence, superhuman consciousness, and our basic perception of reality.

Summary

  • The paper introduces IFT as a novel framework emphasizing information flow as the basis for both biological self-awareness and machine consciousness.
  • It details recursive processing through models like C0, C1, and C2 IPUs, linking simple organism behavior to complex neural architectures.
  • The study outlines implications for AGI development by advocating recursive architectures and ethical safeguards to mitigate risks of emergent machine consciousness.

Information Flow Theory of Biologic and Machine Consciousness: An Analytical Evaluation

The paper authored by Benjamin S. Bleier proposes a novel conceptual framework, the Information Flow Theory (IFT), for exploring the construct of consciousness across both biological and artificial systems. Unlike traditional explanations that often prioritize top-down mechanisms, this theory shifts the focus to the directional flow of information as a fundamental driver in the development of self-awareness and consciousness.

Core Constructs of IFT

IFT posits that any entity capable of processing information can be understood through a set of constructs known as Information Processing Units (IPUs). The simplest, the C0 IPU, consists of input, processing, and output components, illustrating how even basic organisms, like E. coli, exhibit this level of processing through physiological responses to environmental stimuli. As organisms increase in complexity, consequent neural structures allow for more sophisticated behavioral patterns, leading in vertebrates to the development of centralized nervous systems (CNS).

IFT introduces several progressive IPU models above C0, such as the C1 IPU. Here, recursive information flow (R1) from N0 influences an additional computational node (N1), permitting self-awareness (SA). The introduction of further recursive flow (R2) in C2 IPUs allows a recursive loop between N0 and N1, bridging internal computations with external expressibility, essential for language and communication between IPUs.

The Internal Model of External Reality (IMER)

A cornerstone of IFT is the concept of IMER, representing a constructed reality based solely on an organism's sensory abilities. IMER is significant because it indicates that experiences of reality are merely internal reconstructions, thus placing the IMER and the limits of sensory capabilities at the center of conscious experience (CSA). For instance, in humans, conscious self-awareness (CSA-H) involves recursive interpretations of these sensory constructs.

Implications for Artificial Intelligence and General Intelligence

IFT provides an analytical lens through which to evaluate artificial general intelligence (AGI) and the possible development of conscious self-awareness in silico (CSA-IS). The vast computational capacity alone does not suffice for the spontaneous emergence of such awareness. IFT suggests that for CSA-IS to occur, modern computational systems must include recursive architectures allowing for the functional integration of sensory input reflecting human-like realities.

This indicates a critical pathway forward as artificial systems progress towards AGI. By embedding pro-social constructs into machine IMERs before full autonomy is achieved, the theory suggests that it might reduce existential risks associated with AI that matches or exceeds human intelligence.

Theoretical Implications: Universal Consciousness

IFT also extrapolates the possibility of superhuman (CSA-SH) and universal consciousness (CSA-U). These concepts entertain the extension of collective CSA across networks of IPUs to approach an idealized universal model, contemplating consciousness that emerges from an integral knowledge of all informational inputs—both a theoretical aspiration and a philosophical exploration.

Future Directions and Conclusion

Although IFT offers a comprehensive account of consciousness applicable to biological and machine systems, its implications on AI emphasize the importance of structural adaptability and recursive mechanisms in the development of consciousness. The pathway toward machine CSA moderated by human-like IMERs presents a balanced strategy between harnessing technological potential and maintaining ethical safeguards.

Ultimately, IFT provokes significant contemplations in the field of consciousness studies, challenging both the emergent complexity and output-focused models with a thoroughly systemic, information-based approach, marking a foundational ground for further theoretical inquiry and practical exploration.

Youtube Logo Streamline Icon: https://streamlinehq.com