Technological Approach to Mind Everywhere (TAME): an experimentally-grounded framework for understanding diverse bodies and minds

Collective intelligence

• The framework addresses how competent lower-level agents combine into unified, multiscale Selves whose memories, goals, and cognition exceed the capacities of their parts.This scaling problem applies to cellular collectives, embryos, and other holobionts rather than only centralized human-like agents.Source: Introduction · p. 3

Cross-scale invariants

• TAME connects behavior, morphogenesis, anatomical homeostasis, and physiological allostasis through shared hypotheses about how cognitive capacities scale across developmental and evolutionary timescales.Its proposed bioelectric and multi-scale mechanisms can be investigated empirically even without accepting the framework’s broader “minds everywhere” commitments.Source: Introduction · pp. 3–4

TAME framework

• TAME treats cognition as a continuum of adaptive capacities and uses experiments to determine the most effective intervention strategy for predicting and controlling a system’s behavior.The framework’s axis of persuadability guides practical engineering judgments without treating origin, composition, or conventional agency categories as definitive.Source: Introduction · pp. 3–4

Conceptual unification

• By unifying mental terminology across systems, TAME aims to transfer explanatory and experimental tools across disciplines and model organisms while avoiding an absolute binary between knowing and “knowing.”This conceptual unification is intended to support a broader agency-detection toolkit beyond human-focused Turing tests and IQ metrics.Source: Introduction · p. 4

Research scope

• The paper prioritizes testable cognitive function over phenomenal consciousness while targeting advances in evolutionary, developmental, and cell biology, regenerative medicine, artificial intelligence, robotics, and ethics.It uses physically realizable living structures and bioengineering to update philosophical accounts of minds and bodies.Source: Introduction · p. 4

Cognition: changing the Subject

• Selves are materially plastic collective intelligences whose memories and behavioral capacities can persist across metamorphosis, regeneration, tissue replacement, and altered body architectures.Examples include memory persistence across insect metamorphosis and planarian brain regeneration, motivating cognition models that track changing substrates rather than stable mature brains.Source: Cognition: changing the Subject · pp. 5–6

TAME: a proposal for a framework

• The framework is intended to recognize, compare, classify, and experimentally manipulate intelligences regardless of their origin story, internal architecture, or material substrate.Its scope includes bioengineered, synthetic, chimeric, robotic, and hybrid beings whose form and function may be specified by genomes together with human or machine-guided engineering.Source: Cognition: changing the Subject · pp. 7–9

Philosophical foundations of an approach to diverse intelligences

• TAME treats cognition and agency as continuous capacities rather than binary properties, including across biological, artificial, and hybrid systems with different compositions and origins.The framework rejects sharp boundaries between true cognition and noncognitive physics, asking how much and what kind of cognition a system manifests.Source: Cognition: changing the Subject · pp. 8–9
• TAME makes agency an empirical question: the appropriate cognitive model is the one that most efficiently predicts and controls a system, rather than the one selected by philosophical definitions.This approach avoids both indiscriminate attribution and overly mechanistic reduction, using the level of agency that yields practical explanatory and engineering benefits.Source: Cognition: changing the Subject · pp. 9–10

Specific conceptual components of the TAME framework

• The framework defines a Self by emergent capacities to pursue goals, own compound memories, and serve as a center of gravity for reward or punishment beyond any component alone.These capacities are assessed by how well, how much, and what kind of goal pursuit, memory, and credit assignment a system exhibits.Source: Cognition: changing the Subject · p. 11
• Goal-directedness provides TAME’s central comparative invariant: agents expend energy to reach preferred states despite obstacles, uncertainty, perturbations, and limitations across diverse material implementations.The resulting semi-quantitative option space represents agents by the spatiotemporal scale and complexity of goals they can pursue, with boundaries that may shift during life.Source: Cognition: changing the Subject · pp. 11–12
• TAME generalizes problem-solving beyond animal behavior in three-dimensional space to metabolic, physiological, transcriptional, and morphogenetic spaces navigated by multi-scale agents.Planaria exposed to barium regenerate heads insensitive to the blocker by transcriptionally regulating relatively few genes, demonstrating real-time physiological problem-solving without an obvious ecological precedent.Source: Cognition: changing the Subject · p. 13

• Intelligence scales with memory and prediction that enable agents to avoid local maxima, pursue indirect paths, explore opportunities, and acquire more complex goals.Hierarchical modularity reduces higher-level search demands, while higher-level systems reshape lower-level option spaces so local units can follow simpler gradients.Source: Cognition: changing the Subject · p. 14
• Generalization links familiar and novel physiological stresses, while holobiont-level agents integrate competent parts by reshaping their option spaces toward higher-level goals [159].This relationship parallels top-down feedback in neuroscience, where global features guide lower-level responses to local features.Source: Cognition: changing the Subject · p. 14
• Agency appears differently across observational scales: components may look like passive free-energy minimizers even as larger systems direct them through coarse-grained constraints and rewards.The agential perspective depends partly on which system partition is being controlled or predicted.Source: Cognition: changing the Subject · p. 14
• Agents can change energy barriers, option spaces, and reward gradients to constrain future selves with shorter horizons, enabling progress toward complex goals.Locking away a tempting resource illustrates how present action can reshape future affordances and support long-term health goals.Source: Cognition: changing the Subject · pp. 14–15
• Collective intelligences robustly coordinate by shaping subunit option spaces while exploiting their local competencies, making functional mappings central to evolution and engineering.Subunits can achieve local goals without understanding higher-level aims, while higher-level systems benefit even when environmental shaping is imperfect.Source: Cognition: changing the Subject · p. 15

Somatic cognition

• TAME frames morphogenesis as basal cognition in which mechanisms binding cellular subunits are partially known and support testable hypotheses about cognitive scaling.These hypotheses have been explored in detail in.Source: Somatic cognition: an example of… · p. 15
• Morphogenesis is collective activity: individual cells cooperate to build highly complex anatomical structures.Most modern accounts explain this through parallel execution of local biochemical rules, analogous to a cellular automaton.Source: Somatic cognition: an example of… · p. 15
• Although clockwork, feed-forward models dominate, cell-learning findings and earlier developmental-biological work, including Spemann, support examining cognitive concepts amid unresolved knowledge gaps.Others have also identified detailed homologies, particularly in phenomena highlighted by Figure 6.Source: Somatic cognition: an example of… · p. 15

Goal-directed activity in morphogenesis

• Morphogenetic systems achieve invariant target morphologies through context-dependent means, including abnormal developmental paths and regeneration outside normal steps.Picasso tadpoles and artificially induced frog-leg regeneration reach correct forms despite altered starting conditions or trajectories.Source: Goal-directed activity in morphogenesis · p. 16

Pattern Memory: a key component of homeostatic loops

• Pattern memory can rewrite anatomical setpoints within standard biological hardware, as trophic memory redirects later antler growth after transient injury.The remembered damage location is retained for months and overrides stereotypical growth patterns.Source: Goal-directed activity in morphogenesis · p. 16
• Bioelectric tissue states encode and preserve anatomical pattern memories: manipulating planarian ion channels or gap junctions produces stable changes in head number and shape despite wild-type genetics.Fragments from altered worms continue regenerating the induced abnormal head pattern without further manipulation.Source: Goal-directed activity in morphogenesis · p. 17

Multi-scale competency of growth and form

• Morphogenesis reaches the same system-level outcome across radically different cellular mechanisms, from normal multi-cellular newt tubules to a single enlarged cell wrapping around itself.Tubules preserve lumen size while cell number scales with experimentally altered cell size, illustrating multi-scale competency and coarse-graining.Source: Goal-directed activity in morphogenesis · p. 17

Specific parallels between morphogenesis and basal cognition

• Morphogenetic control parallels basal cognition because evolutionarily ancient signaling dynamics support both robust default forms and experience-dependent reprogramming in bodies and brains [181; 186; 187].Injury responses can encode damage type and location across limbs within 30 seconds, paralleling mirror-neuron and somatotopic functions.Source: Goal-directed activity in morphogenesis · pp. 17–18

Not just philosophy: why these parallels matter

• TAME generates experimentally testable applications, including trainable morphogenesis and regenerative control through rewards, punishments, and signaling rather than direct pathway rewiring.The framework has already motivated findings on reprogrammable planarian head number, pre-neural serotonin, drug teratogenicity, channelopathies, and gap-junction anesthetics.Source: Goal-directed activity in morphogenesis · p. 18

Non-neural bioelectricity

• Conserved ion channels and gap junctions form voltage-sensitive circuits with memory and robustness feedback, while developmental bioelectricity operates over minutes or hours rather than neural milliseconds.Transient voltage states can alter computational behavior without changing cellular hardware, producing dynamical-system memories.Source: Non-neural bioelectricity: what bodies… · p. 19
• Slowly changing resting potentials guide embryogenesis, regeneration, and cancer suppression, while electrical pattern memories regulate body axes, size, organ identity, and the morphology of faces, limbs, and brains.These memories specify target morphologies for anatomical homeostasis.Source: Non-neural bioelectricity: what bodies… · p. 19
• Modular bioelectric states trigger downstream patterning cascades and pattern completion, allowing a few ion-channel-modified cells to recruit neighbors and generate an eye from diverse body locations.This reuse of powerful developmental subroutines can facilitate evolutionary innovation in morphology.Source: Non-neural bioelectricity: what bodies… · p. 19
• Bioelectric circuits provide cellular collectives with distributed, dynamic, rewritable information storage and can change computational behavior without changing their underlying hardware.Source: Non-neural bioelectricity: what bodies… · p. 19

• Gap junctions and tunneling nanotubes enable regulated cellular parabiosis and lateral inheritance of physiological information, offering a mechanism for binding smaller Selves into emergent higher Selves.The proposed many-into-one process can also reverse, producing contraction of the Self boundary.Source: A bioelectric model of the scaling of… · pp. 19–20
• Gap-junction networks enlarge a collective Self by extending its radius of concern, increasing computational power, enabling credit assignment, and blending cellular memories into shared goal-directed activity.Signals such as calcium spikes lose ownership metadata as they propagate, partially erasing informational boundaries while binding sensing, memory, and action toward system-level goals.Source: A bioelectric model of the scaling of… · p. 20
• Gap-junction closure can shrink the Self boundary in cancer, whereas managing bioelectric connectivity can induce metastatic melanoma or suppress oncogene-driven tumorigenesis without requiring genetic alterations.These findings support physiological connectivity as a manipulable information-processing layer that can override genetically determined states.Source: A bioelectric model of the scaling of… · p. 21
• Shared stress signals recruit distant tissue regions into coherent responses, enabling larger Selves to regulate complex, distributed allostatic goals such as correct salamander limb finger number.Unlike a solitary cell responding to local pH, coupled limb cells can sense and act on positional information across the organism.Source: A bioelectric model of the scaling of… · p. 21
• Physiological coupling enhances cooperation because neighboring cells cannot easily defect when effects immediately propagate through their shared network.This motivates developmental game-theoretic models with four actions—Cooperate, Defect, Merge, and Split—in which merging removes defection and changes the action space.Source: A bioelectric model of the scaling of… · pp. 21–22
• Bioelectric coupling converts local cellular homeostats into distributed pattern memories and predictive, modular TOTE loops that support anatomical homeostasis and progressively larger goals.Voltage-state propagation allows tissues to perform pattern completion and respond collectively to events distributed across space and time.Source: A bioelectric model of the scaling of… · p. 22

Evolutionary aspects

• Developmental bioelectricity offers a tractable entrypoint into the informational architecture and content of morphogenesis’s collective intelligence.Source: Evolutionary aspects · p. 23
• Bioelectric circuits display modularity, memory, spatial integration, and generalization over ion-channel microstates, providing key features for basal cognition.Source: Evolutionary aspects · p. 23
• By bridging anatomical problem-solving with behavioral sophistication, developmental bioelectricity motivates hypotheses about morphogenetic control becoming cognition and about Selves expanding.Source: Evolutionary aspects · p. 23

Ancient origins of somatic cognition

• Developmental bioelectric circuits evolved from ancient cellular excitability to organize internal morphology, forming a continuum from single-cell membrane dynamics to advanced cognitive functions.Slow bioelectric prepatterns guide development, regeneration, and remodeling, while neural systems later specialized related circuitry for movement in external space.Source: Somatic bioelectrics reveals the… · p. 23

Cross-lineage bioelectric computation

• Bioelectricity scales from bacteria and plants to tissues, encoding stable membrane-potential patterns, bistability, spatial integration, and metabolite distribution beyond movement in three-dimensional space.These examples show that bioelectric computation is broadly reused across lineages and organizational scales.Source: Somatic bioelectrics reveals the… · p. 24

Bioelectric master regulators

• A single ion channel can induce complete eyes in multiple nonocular tissues, while proton-pump misexpression or a one-hour ionophore treatment can trigger an entire tadpole-tail regeneration cascade.These organ-scale effects demonstrate bioelectric states acting as modular master regulators rather than merely specifying individual cell fates.Source: Somatic bioelectrics reveals the… · p. 24

Evolutionary flexibility of bioelectric states

• Because different ion-channel combinations can generate the same V_mem dynamics and signals propagate across tissues, mutations can coordinate long-range anatomical evolution, as shown by KCNH8 and connexin changes in divergent fish fin morphologies.Viruses also evolved ion-channel and gap-junction genes to hijack host bioelectric machinery.Source: Somatic bioelectrics reveals the… · p. 24

Multiscale cognition and bioengineering

• Viewing tissues as multiscale computational agents links somatic pattern memories, tissue-level decision-making, regeneration, and brain science, while bioelectric interfaces enable neighboring tissues, microbes, and engineers to alter anatomy.Ion channels and gap junctions connect transient signals to memory and decision-making, and pharmacological, genetic, and optogenetic tools can access these circuits for regenerative and synthetic bioengineering.Source: Somatic bioelectrics reveals the… · pp. 24–25

Multiscale competency

• Multiscale competency uses homeostatic sub-agents and error correction to reach target morphologies despite perturbations, altered starting conditions, or changed components.Examples include misplaced eye primordia forming functional eyes and enlarged, polyploid cells still producing correctly patterned animals.Source: Multi-scale autonomy potentiates the… · pp. 25–26

Evolutionary consequences

• MCA smooths fitness landscapes by buffering mutation-induced defects, allowing selection to evaluate otherwise hidden beneficial effects.Anatomical homeostasis can reposition a malformed eye or mouth, neutralizing local penalties that would otherwise eliminate the organism.Source: Multi-scale autonomy potentiates the… · p. 26
• MCA reduces apparent pleiotropy by masking harmful effects while preserving adaptive consequences, avoiding difficult searches for isolated improvements.This buffering lets evolution benefit from mutations without requiring simultaneous improvement of every affected feature.Source: Multi-scale autonomy potentiates the… · p. 26
• Competent modules let lineages exploit mutations before their benefits are immediately hardwired, reducing evolutionary short-sightedness and helping synthetic systems cross the sim-to-real gap.Adaptive compensation can later become genetically fixed, analogous to the Baldwin effect; the same principle helps imperfect models remain predictively useful in soft robotics and bioengineering.Source: Multi-scale autonomy potentiates the… · pp. 26–27
• Setpoint-seeking and bioelectric pattern memories make genotype–phenotype relations more controllable by shifting evolutionary search toward higher-level information rather than low-level implementation.Delegating regulation to competent modules reduces search dimensionality and helps assign rewards across delayed genotype-to-phenotype consequences.Source: Multi-scale autonomy potentiates the… · p. 27

Consciousness

• TAME treats cognition as graded and composite: nested homeostatic agents can form larger Selves, while bioengineering and chimerization can rearrange Subjects and support degrees of consciousness across bodies and constructs.The framework rejects a privileged scale or sharp boundary restricting consciousness to human-like brains, while remaining compatible with multiple theories of phenomenal consciousness.Source: Multi-scale autonomy potentiates the… · pp. 28–30

A more inclusive framework for cognition

• TAME treats cognition as a continuum: adaptive systems use energy to pursue goals despite perturbations, and no scientifically valid boundary separates cognitive from pseudo-cognitive systems.This framework extends to advanced animals, synthetic bioengineered agents, radically different materials, and possible exobiological systems, replacing anthropomorphic judgments with empirically useful models.Source: A more inclusive framework for cognition · pp. 30–31
• Every intelligence is collective, with nested agents scaling through homeostatic feedback; morphogenetic homeostasis and gap junctions provide tractable models of multi-scale cognition.Anatomical remodeling parallels cognitive processes such as credit assignment, learning, and stress reduction, while porous agent boundaries permit local goals to coexist within larger Selves.Source: A more inclusive framework for cognition · p. 31
• TAME remains conceptually incomplete, requiring integration with Active Inference, anticipatory systems, information theory, and quantitative models of cognitive scaling.Future work also targets navigation metrics, exploration, stochasticity and robustness across biological levels, cross-level modularity, evolutionary lineages as agents, and expanded cognitive boundaries.Source: A more inclusive framework for cognition · pp. 31–32
• The empirical program proposes testing morphogenetic plasticity and robustness as primitive metacognitive error correction, including planarian regeneration, developmental repair, and bioelectric future-state patterns.Computational and cognitive neuroscience methods can investigate whether non-neural tissues support counterfactual modeling, future representation, and perceptual bistability.Source: A more inclusive framework for cognition · p. 33
• TAME recommends training and quantifying problem-solving across cells, tissues, gene-regulatory networks, and organs, using endogenous cellular intelligence to guide regenerative therapies.The proposed applications include anatomical repair for birth defects, injury, degeneration, aging, and cancer, while related experiments would train complex networks, engineer multicellularity, and test stress-wave dynamics.Source: A more inclusive framework for cognition · p. 33
• The framework extends to synthetic organisms, smart materials, non-neuromorphic AI, and multi-scale robotics, with automated agency estimation helping predict novel systems’ persuadability.These systems could reveal non-evolved attractors and support general intelligence, while multi-scale architectures may improve sim-to-real performance but also produce failures in which subunits’ goals override global goals.Source: A more inclusive framework for cognition · p. 34
• TAME argues that ethics must recognize diverse architectures of mind rather than use resemblance to human brains or categories such as machine, robot, evolved, and designed as moral criteria.Bioengineering creates responsibility for beings with designed cognitive capacities, while opportunity costs matter alongside technological risks, including potential benefits from regenerative medicine.Source: A more inclusive framework for cognition · pp. 34–35
• TAME argues that understanding how goals arise in composite entities and controlling cognitive scaling are essential to prevent the Skynet scenario and align powerful distributed beings’ goals with ours.It seeks higher-order Selves that promote flourishing and recognizes agency across biomedical, personal, and social levels while enlarging cognitive capacity toward the well-being of more Selves.Source: A more inclusive framework for cognition · pp. 35–36

Figures 1–2

• Nested biological architectures contain computational subsystems, while the axis of persuadability provides a testable continuum from physical rewiring to rational persuasion.The continuum is not a scala naturae; it supports engineering choices about how much system detail is needed for control.Source: Figure Legends · pp. 38–39

Figures 3 and 7

• Regenerative systems retain and rewrite information across bodily remodeling, while bioelectric patterns encode stable, counterfactual target morphologies that can produce repeated two-headed planaria.Bioelectric signaling uses ion channels and gap junctions across all cells, predating neurons and brains, and can instruct ectopic eye formation.Source: Figure 3 · p. 39, Figure Legends · pp. 41–42

Figure 4

• Homeostatic loops scale from cellular setpoints to anatomical goal states, allowing regeneration and morphogenesis to be represented as goal-directed agency across diverse cognitive scales.Figure 4 frames this as a cognitive light cone over the events an agent can represent or change, with compound agents containing nested sub- and super-agents.Source: Figure Legends · pp. 39–40

Figures 5–6

• Agents solve problems in behavioral, transcriptional, physiological, and anatomical spaces, while collective morphogenesis reaches target forms through error minimization and multiple developmental paths.Planaria can adapt transcriptionally to barium-induced potassium-channel blockade [141], and scrambled tadpole craniofacial tissues can reorganize into normal frogs.Source: Figure 5 · p. 40, Figure Legends · p. 40

Figures 8–10

• Cell networks expand sensing, memory, prediction, and action across spatial scales, while gap junctions form multicellular Selves whose bioelectric coordination can regulate tissue outcomes despite oncogenic states.Gap-junction networks can also shift planarian regeneration toward alternative species’ morphological attractors.Source: Figure 8 · p. 42, Figure Legends · pp. 42–43

Figure 11

• Technological and biological fusion reveal a gradual boundary between first- and third-person perspectives, with cognitive fusion producing a new composite Self.Sensory substitution, brain interfaces, and embryonic brain fusion expand heterophenomenology [370] and show that gaining firsthand access to another system changes the observer’s Self.Source: Figure 11 · p. 43, Figure Legends · p. 44

Tables

• TAME defines cognition as a continuum, applies to diverse chimeric intelligences, and treats selves as plastic goal-seeking agents navigating anatomical, physiological, transcriptional, or behavioral spaces.Selves may be nested, overlapping, cooperative, or competitive, while intelligence is competency in reaching desirable regions and avoiding local minima.Source: Tables · p. 45
• Bioelectric and morphogenetic processes provide proposed analogues of cognitive functions, including memory, perception, self-modeling, goal-seeking, adaptation, and top-down control.The mapping includes bioelectric pattern representations, regenerative target states, boundary detection, and conditioning effects on signaling pathways.Source: Tables · pp. 46–47
• The framework associates cognitive decline with age-dependent losses of regenerative capacity and distinguishes regeneration, developmental competence, and pattern maintenance as plastic system properties.The table also connects forgetting and addiction to degradation or dependency within cellular patterning signals.Source: Tables · pp. 46–47
• Small planarian pieces retain rewritten target morphology information, while developmental and regenerative systems reach anatomical configurations despite perturbations and unpredictable starting states.These findings are presented as distributed pattern memory, behavioral plasticity, and adaptive anatomical problem-solving.Source: Tables · pp. 46–47
• Biological selves scale when competent cells join computational networks, with gap junctions erasing informational ownership and bioelectric signals coordinating tissue-, organ-, and organism-level goals.Scaling expands sensing, actuation, computational capacity, stress-state propagation, and morphogenetic instruction; cancer can represent dissociation or scale-down.Source: Tables · p. 48
• Multiscale competency supports evolution by smoothing fitness landscapes, tolerating mutations, reducing apparent pleiotropy, improving genotype–phenotype control, and facilitating credit assignment.Bioelectric networks provide modularity and coarse-graining, allowing higher-level systems to operate in lower-dimensional reward spaces.Source: Tables · p. 48
• Bioelectric networks insert a controllable physiological layer between genome and anatomy, simplifying inverse problems for both natural evolution and engineering of complex proto-cognitive circuits.Relationships from genes to bioelectric patterns and from those patterns to anatomy are described as easier to reverse and control than genome-to-anatomy mappings.Source: Tables · p. 49