Summary: A new theoretical framework proposes that the old divide between computational functionalism and biological naturalism misses how real brains compute. The authors introduce “biological computationalism,” the view that neural computation is inseparable from the brain’s physical, hybrid, and energy-constrained dynamics rather than an abstract algorithm running on generic hardware.
Biological computationalism holds that discrete neural events and continuous physical processes form a tightly coupled system that cannot be reduced to symbolic information processing alone. This perspective implies that current digital AI, despite its impressive capabilities, may not capture the essential computational style that gives rise to conscious experience. Instead, truly mind-like cognition may require systems whose computation emerges from physical dynamics similar to those in biological brains.
Key Facts:
- Hybrid dynamics: Brain computation combines discrete spikes with continuous chemical and electrical fields.
- Multi-scale coupling: Neural processes are deeply intertwined across levels, so algorithms cannot be cleanly separated from physical implementation.
- Energetic constraints: Metabolic limits shape neural computation, influencing learning, stability, and information flow.
Source: Estonian Research Council
The debate about consciousness often feels stuck between two entrenched views.
Computational functionalism treats cognition as abstract information processing: give the right functional organization, and consciousness will follow regardless of substrate. Biological naturalism counters that conscious experience depends on the concrete physical properties of living brains and bodies — biology is not just the vehicle for cognition but part of what cognition is.
Both positions capture important insights, but the impasse indicates something essential is missing. Our paper proposes a third path: biological computationalism. The core claim is that the traditional computational paradigm — treating cognition as software separable from hardware — is poorly matched to how real brains operate.
Brains do not resemble von Neumann machines. Assuming that cognition is merely software running on neural hardware forces awkward metaphors and fragile explanations. A serious theory of neural computation and of what it would take to build minds in other substrates requires broadening what we mean by “computation.”
Biological computation has three defining properties:
1. Hybrid event–field dynamics. Neurons produce discrete spikes and synapses release neurotransmitters, but these events are embedded within continuously evolving voltage fields, chemical gradients, ion diffusion, and time-varying conductances. The brain is neither purely digital nor purely analog; continuous processes shape discrete events and vice versa in ongoing feedback.
2. Scale inseparability. Unlike conventional computing, where software and hardware can be treated as separate layers, brain processes run across multiple scales simultaneously — from ion channels to dendrites to circuits to whole-brain dynamics. There is no tidy boundary where an abstract algorithm ends and physical implementation begins: changing the implementation changes the computation.
3. Metabolic grounding. The brain is severely energy-limited, and its organization reflects energetic constraints. These constraints are not incidental engineering details; they shape what can be represented, how learning proceeds, which dynamics remain stable, and how information flows are orchestrated. Tight coupling across scales can be understood as an energy-optimization strategy for robust, adaptive intelligence.
Taken together, these properties imply a shift in how we interpret computation in neural systems. Computation in brains is not abstract symbol manipulation running independently of physical substrate. Instead, the physical organization does not merely support computation — it constitutes it. Brains are physical processes that compute by unfolding in time; their algorithms are embedded in their material dynamics.
This framing also clarifies limitations in how we often talk about contemporary AI. Current artificial systems largely simulate functions: they implement mappings from inputs to outputs on hardware designed for a different computational style. In biological systems, continuous fields, ion flows, dendritic integration, local oscillations, and emergent electromagnetic interactions are not peripheral details but the computational primitives that enable real-time integration, resilience, and adaptive control.
Biological computationalism is not a claim that consciousness is magically exclusive to carbon-based life. Rather, it specifies that if consciousness depends on this class of hybrid, scale-inseparable, energetically grounded computation, then reproducing mind-like cognition may require architectures that embody those properties — whether in biology or in new engineered substrates.
For researchers and engineers pursuing synthetic minds, this shifts the target. Scaling digital AI might increase capability, but capability alone may not produce mind-like properties if the underlying computational ontology remains unchanged. The deeper challenge is designing physical systems whose dynamics and energetic constraints make computation intrinsic to their material organization.
The practical question becomes: what physical features — hybrid event–field interactions, multi-scale coupling without clean interfaces, and energetic constraints that shape inference and learning — are necessary so that computation is an intrinsic property of the system rather than an abstract description imposed from outside?
Key Questions Answered:
A: It addresses the stalemate between theories that view consciousness as pure information processing and those that ground it exclusively in biology, proposing a model that integrates computation with physical dynamics.
A: Biological computation depends on continuous physical processes, energy constraints, and multi-scale interactions that fundamentally change how information is represented and transformed.
A: If consciousness depends on biological-style computation, future artificial systems may need new physical architectures — not just scaled-up digital algorithms — to replicate mind-like properties.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by staff.
About this consciousness and AI research news
Author: Merilin Reede
Source: Estonian Research Council
Contact: Merilin Reede – Estonian Research Council
Image: The image is credited to Neuroscience News
Original Research: Open access. “On biological and artificial consciousness: A case for biological computationalism” by Jaan Aru et al., Neuroscience and Biobehavioral Reviews
Abstract
On biological and artificial consciousness: A case for biological computationalism
Rapid advances in large language models have intensified debates about whether artificial systems might become conscious. Optimism often rests on computational functionalism — the idea that the right pattern of information processing alone determines consciousness. Biological naturalism opposes this, arguing that conscious experience depends on the concrete physical processes of living systems. Despite the centrality of these positions, there has been no coherent framework explaining how biological computation differs from digital computation or why that difference might matter for consciousness.
We argue that the absence of consciousness in artificial systems is not only a matter of missing functional organization but reflects a deeper divide between digital and biological modes of computation and the dynamical and structural dependencies of living organisms. Biological systems support conscious processing because they instantiate scale-inseparable, substrate-dependent multiscale processing as a metabolic optimization strategy, and because, alongside discrete computations, they perform continuous-valued computations due to their fluidic substrates. These features — scale inseparability and hybrid computations — are essential to the brain’s mode of computation. We outline foundational principles of a biological theory of computation and explain why current artificial intelligence systems are unlikely to replicate conscious processing as it arises in biology.