TAME

YouTube ... Quora ...Google search ...Google News ...Bing News

• Life~Meaning ... Consciousness ... Creating Consciousness ... Quantum Biology ... Orch-OR ... TAME ... Proteins
• Brain waves that define the limits of you
• Artificial General Intelligence (AGI) to Singularity ... Curious Reasoning ... Emergence ... Moonshots ... Explainable AI ... Automated Learning
• Levin Labs Publications
  • The Computational Boundary of a “Self”: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition - Michael Levin - Tuffs University
• TAME - Materialism - Phylogenetic/Evolutionary | Landscape of Consciousness

A major contemporary influence on Scale-Free Agentialism is developmental biologist Michael Levin (Tufts University), whose work is often summarized under the framework TAME (Technological Approach to Mind Everywhere). TAME encourages researchers to study cognition as a set of capabilities that can appear at multiple scales in living systems: cells, tissues, organs, and brains. Scale-Free Agentialism is a research perspective that shifts attention away from purely metaphysical debates about consciousness (for example, Qualia) and toward a scientific, testable question:

How much agency does a system have?

In this view, “mind” is not a binary property that suddenly appears only in large brains. Instead, agency is treated as a continuum that can, in principle, be measured by observing how systems pursue goals, store information, coordinate actions, and correct errors across time.

Core idea: agency as a continuum - Scale-Free Agentialism proposes that many familiar “mental” capacities exist in simpler forms throughout biology:

• Goal-directed behavior — acting to reach or maintain a target state.
• Memory — persistent internal states that influence future behavior.
• Problem-solving — finding workable paths to a goal under constraints.
• Error correction — detecting deviation from a target and compensating.
• Coordination — distributed parts acting together as a coherent unit.
• Non-local Information Integration — the ability of a collective to sense and make decisions about large-scale anatomical targets that no individual cell can perceive on its own.
• Valence and Stress Regulation — registering states as "good" or "bad" and acting to minimize "stress" (the distance from the target state).
• Predictive Modeling (Anticipation) — using past information to prepare for future states.
• Persuadability (Interaction Level) — the degree to which an agent can be influenced via "messages" and information rather than just physical force.
• Cognitive Light Cone (Goal Scaling) — the spatial and temporal boundary of what a system "cares" about.
• Shared Memory (Anonymization) — the process where individual parts couple via gap junctions to act as a single unified "Self."
• Morphological Plasticity (Equifinality) — the ability to reach the same functional "target state" from different starting points or despite external perturbations.
• Competence — the autonomous intelligence of sub-components that allows the higher-level system to delegate complex tasks (e.g., "build a kidney") without managing every cell migration.

Basal Cognition — intelligence is not limited to brains but exists on a spectrum across all biological levels. In a scale-free view, cognition-like properties can be found at many levels:

• Cells can sense, decide, and act (e.g., migrating, forming boundaries, repairing damage).
• Tissues can coordinate collective actions (e.g., closing wounds, maintaining structural integrity).
• Organs can regulate toward stable functional states (e.g., maintaining physiological variables).
• Brains specialize in high-bandwidth, fast, flexible coordination—especially behavior in complex environments.

From this perspective, the key scientific task is to identify and measure these capacities in different systems, rather than assuming cognition begins (or ends) at the skull.

The mystery of Morphogenesis
A central puzzle for TAME is Morphogenesis: how a single fertilized egg reliably builds a complex 3D anatomy (with correct symmetry, organ placement, boundaries, and “stop” signals that end growth at the right time). TAME frames morphogenesis as a kind of distributed control problem:

Developing and regenerating tissues behave like collectives that can sense deviation from a target structure and coordinate multi-step actions to reduce that deviation.

DNA as “hardware”
DNA is essential, but it is not a literal geometric blueprint for the whole organism.

• DNA primarily provides the parts list (proteins) and many local rules for cell behavior.
• Genes help specify what cells can do, but do not straightforwardly encode the full, high-level geometry of a finished body.

This distinction is often summarized as:

Genes specify parts and local rules; additional control layers help specify global outcomes.

A useful way to read Levin's contribution is: genes specify parts and local rules, but bioelectric networks help specify global outcomes—where “global” means whole-organism geometry (symmetry, organ placement, boundary conditions, and when to stop growing). This frames development and regeneration as a kind of distributed control system: tissues sense deviation from a target and then coordinate multi-step repair.

Bioelectricity as “software” Michael Levin's work emphasizes that many cells (not only neurons) use Ion channels and other mechanisms to generate and respond to electrical signals. These signals can form networks that coordinate behavior across tissues.

In the TAME framing:

• Cells form bioelectric networks that influence growth, patterning, and regeneration.
• These networks can act like a shared information layer that helps tissues coordinate toward a target anatomy.
• Bioelectric signaling can integrate local cellular events into stable, large-scale outcomes (for example, coherent organ boundaries or correct overall proportions).

This “software” metaphor is meant to highlight that:

• The same genetic “hardware” can support different anatomical outcomes depending on the information dynamics of cellular networks.
• Some patterning information may be stored and updated in tissue-level states, not only in DNA.

Bioelectric “pattern memory” (target anatomy) A recurring idea in TAME is that tissues can behave as if they contain a memory of a target structure:

• Tissues can detect when they are “off-target” (missing parts, incorrect boundaries, wrong polarity).
• They can recruit coordinated cellular actions (growth, differentiation, remodeling) to move back toward the target.
• They can stop once the target is achieved (an important feature of healthy development and regeneration).

This supports a practical, engineering-friendly interpretation:

Morphogenesis and Regeneration can be studied as feedback control in living collectives.

Illustrative examples (development and regeneration)

TAME draws attention to experimental and natural phenomena that look like collective problem-solving:

• Regenerative accuracy: Some organisms regenerate complex structures (for example, correct limb shape) and stop at the right size.
• Large-scale pattern changes from bioelectric perturbations: In some experimental contexts (notably in Planarian regeneration studies), altering bioelectric signaling has been associated with dramatic anatomical outcomes (such as altered body polarity), suggesting that pattern-level control is not exclusively genetic.

These examples are used to motivate a broader claim:

Anatomy is not only “built” by chemistry; it is also “steered” by information.

A distributed mind - TAME encourages a view of the body as a cognitive system composed of agential materials—living parts that can coordinate and pursue goals at multiple scales.

Scale-free continuum

In a scale-free view, cognition-like properties can be found at many levels:

• Cells can sense, decide, and act (e.g., migrating, forming boundaries, repairing damage).
• Tissues can coordinate collective actions (e.g., closing wounds, maintaining structural integrity).
• Organs can regulate toward stable functional states (e.g., maintaining physiological variables).
• Brains specialize in high-bandwidth, fast, flexible coordination—especially behavior in complex environments.

The emphasis is not that “a cell thinks like a person,” but that:

Similar organizational principles (information integration, feedback, goal maintenance) can appear across scales.

The brain as a specialist, not the sole source

In this framework, the brain is understood as a powerful specialization of a broader biological capacity:

• It expands the range of goals and the speed/complexity of coordination.
• It does not erase the agency present in non-neural tissues, which also regulate, coordinate, and pursue target states.

This framing is intended to be biologically unifying:

The same general idea—distributed, goal-directed regulation—can connect development, healing, physiology, and behavior.

Why this matters TAME and Scale-Free Agentialism are often seen as impactful because they:

• Provide a testable research program (measure agency rather than argue only in philosophy).
• Highlight actionable control layers in biology (especially Bioelectricity) relevant to:
  • regenerative medicine
  • birth defect repair (in principle)
  • tissue engineering and synthetic morphology
  • robust self-repairing systems inspired by biology

The “blueprint” problem (and why DNA alone is not enough)

A useful way to sharpen the morphogenesis question is to separate molecular ingredients from organizational information. Standard biology often talks as if the genome “codes for” anatomy, but the genome primarily encodes protein sequences and regulatory elements—not an explicit file that says “five fingers,” “two eyes,” or “bilateral symmetry.”

A common thought experiment makes the paradox intuitive:

• The “salamander soup” paradox: If you reduce a salamander to a soup of molecules, you still have the same chemical ingredients (and in principle, the same DNA), but you have destroyed the organism’s higher-level organization. The missing piece is not “more molecules,” but the information that coordinates them into a coherent body plan.

In Levin's framing, the “blueprint” is better described as a dynamic, distributed state maintained in the communication among cells—especially via bioelectric networks—that helps tissues converge on robust anatomical outcomes.

Key clarifier points:

• Not a picture: The blueprint is not an internal image of the final body, but a set of constraints, goal-states, and control policies that can be reached by more than one route.
• Not only chemical gradients: Morphogen gradients matter, but Levin emphasizes that stable electrical states and tissue connectivity can store and update pattern information across time.
• Not fragile: Development and regeneration can be surprisingly robust—tissues can compensate, reroute, and still converge on a correct form.

The bioelectric code (mechanism-level details)

Many (non-neural) cells maintain a steady resting membrane voltage, often written as <math>V_{mem}</math>. In the bioelectric view, these stable voltages are not “background noise”—they can function as state variables that influence gene expression, cell behavior, and tissue-level patterning. Levin refers to this as a kind of bioelectric code—a computational layer that sits between genes and anatomy. <ref>Levin M. “The bioelectric code: An ancient computational medium for dynamic control of growth and form.” (2017) Biosystems. https://pmc.ncbi.nlm.nih.gov/articles/PMC10464596/</ref>

Core components:

• Ion channels and pumps help establish and modify <math>V_{mem}</math> by controlling which ions enter/leave the cell.
• Gap junctions couple neighboring cells, allowing ion flow and small signaling molecules to pass, which enables tissues to behave like coordinated networks.
• network patterns matter: The important variable is often not one cell’s voltage, but the spatial pattern across a tissue and the way cells are electrically coupled.

This supports a “software-like” interpretation:

• Genes provide the parts list (hardware).
• Bioelectric circuits provide a higher-level control layer (software).
• Changing the tissue’s bioelectric state can change outcomes without changing the DNA.

Measuring and visualizing bioelectric patterns

Bioelectric states are measurable, and multiple complementary tools are commonly used:

• Voltage-sensitive dyes to visualize tissue-level voltage domains during development.
• Electrophysiology (e.g., microelectrodes) to directly measure membrane potentials.
• Genetically encoded voltage indicators (GEVIs) to report voltage changes in living tissues over time.

This measurement toolbox is important because it keeps the concept of “bioelectric pattern memory” anchored to observable, falsifiable signals.

Voltage domains, boundaries, and “prepatterns”

In developing tissues, bioelectric patterns can form voltage domains—regions with distinct electrical states. These domains can behave like “territories,” and the boundaries between domains can function like anatomical borders that help guide where structures form.

A helpful intuition:

• Prepatterns are early electrical layouts that can predict later anatomy—like a “map” that appears before the “city” is built.
• Boundaries between domains can help stabilize where different tissues or structures should emerge.

This perspective reinforces the idea that anatomy is not only assembled from parts, but also steered by large-scale informational constraints.

Target morphology as an attractor (and why growth stops)

The concept of target morphology can be usefully described in dynamical-systems language as an attractor state:

• Many small perturbations still “roll downhill” into the same final form.
• Larger perturbations can push the system into a different stable outcome.
• Changing bioelectric networks parameters can shift which attractor is “home.”

This framing also makes “knowing when to stop” more intuitive:

• Unchecked growth leads to tumors or malformations.
• A target-morphology model provides a natural stopping rule: as the tissue approaches its target, the “error” signal decreases and growth/remodeling taper off.

For planarian regeneration and stable changes in body patterning after brief bioelectric perturbations, see:

• <ref>Durant F. et al. “Long-Term, Stochastic Editing of Regenerative Anatomy via Targeting Endogenous Bioelectric Gradients.” (2017) https://pmc.ncbi.nlm.nih.gov/articles/PMC5443973/</ref>
• <ref>Beane W.S. et al. “Bioelectric signaling regulates head and organ size during planarian regeneration.” (2013) Development. https://pubmed.ncbi.nlm.nih.gov/23250205/</ref>

“Agential materials” (a simple, non-mystical spectrum)

TAME often motivates clearer scientific questions by treating tissues as agents (in a precise, non-mystical sense). One simple spectrum is:

• Passive matter (e.g., a rock): must be forced into shape.
• Active matter (e.g., a wind-up toy): contains energy, but no goal.
• Agential matter (e.g., a cell, tissue, animal): has preferred states and can act flexibly to reach them.

In this usage, agency does not require human-like consciousness. It means:

• the system has a preferred state,
• it can sense deviation,
• it can act to reduce deviation,
• it can do so via multiple strategies (flexibility).

This turns vague questions into actionable ones:

• What is the goal state?
• What counts as “error,” and how is it sensed?
• What actuators exist (growth, migration, apoptosis, remodeling, etc.)?
• What strategies (policies) does the system use under constraints?

Example — “Picasso tadpoles” and anatomical competence

A well-known example used by Levin involves “Picasso tadpoles,” in which craniofacial features begin in abnormal positions. During development and metamorphosis, these animals can still end up with largely normal frog faces, because the organs move along novel paths and stop when they reach correct anatomical relationships. <ref>Levin M. “Technological Approach to Mind Everywhere.” (2022) https://pmc.ncbi.nlm.nih.gov/articles/PMC8988303/</ref>

The conservative takeaway worth emphasizing is:

• The system’s behavior looks less like a hardcoded script (“move organ X by distance Y”) and more like goal-seeking navigation in morphospace (reach a target configuration from varied starting states).

Note: Popular retellings sometimes blur details; the durable scientific point is the demonstration of developmental plasticity and the usefulness of describing it as problem-solving toward a stable anatomical goal.

The “Cognitive Light Cone” (boundaries of self and collective goals)

Levin uses the metaphor of a cognitive light cone to describe the spatial and temporal scope of a system’s goals—what it can “care about” and regulate over space and time. <ref>Levin M. “The Computational Boundary of a ‘Self’: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition.” (2019) https://pmc.ncbi.nlm.nih.gov/articles/PMC6923654/</ref>

Examples (as a conceptual ladder):

• A single cell regulates nearby conditions over short time horizons.
• A tissue can regulate geometry (e.g., wound closure or limb pattern) over longer horizons.
• An organism regulates survival and behavior over years.

Mechanistic bridge to collective intelligence:

• When cells couple via Gap junctions, information flow can be fast and dense enough that “self vs neighbor” boundaries soften.
• Strong coupling can expand the effective cognitive light cone, enabling higher-level goals (e.g., “build and maintain a limb”).

Cancer through a network lens (carefully framed):

• One interpretation emphasizes that reduced coupling and altered bioelectric states can isolate cells from tissue-level constraints, shrinking the scope of collective goals and pushing behavior toward single-cell priorities.
• This complements (not replaces) genetic and immune perspectives, and motivates research into restoring normal constraint signals as one potential therapeutic angle.

Xenobots and programmable living collectives

Xenobots are small living constructs made from Xenopus (frog) cells, developed by teams including Levin and Josh Bongard. They are notable because they demonstrate that cellular collectives can form novel, stable bodies and behaviors when placed under new constraints. <ref>Kriegman S., Blackiston D., Levin M., Bongard J. “Kinematic self-replication in reconfigurable organisms.” (2021) PNAS 118(49):e2112672118. https://krorgs.github.io/</ref>

Why this matters to morphogenic intelligence:

• Cells are competent: they can self-organize into functional forms beyond their usual developmental context.
• Form follows constraints: changing geometry, coupling, and environment yields new solutions.
• Top-down goals, bottom-up implementation: you can specify a goal (e.g., locomote or aggregate cells) while the living material discovers how to implement it.

Ethics/clarifier note (useful on public-facing pages):

• Xenobots are not metal robots; they are living cell collectives.
• The main scientific takeaway is “programmable living matter” and the importance of careful goals and safety constraints.

Practical implications (beyond philosophy)

These ideas are not only conceptual; they suggest actionable directions:

• Regenerative medicine as setpoint repair: restoring or editing pattern memory (target morphology), not only adding cells.
• Birth defects and repair: some defects may be framed as broken patterning control loops or mis-specified tissue states.
• Cancer and governance: complementing genetic/immune approaches with strategies that restore coupling and constraint signals.
• Living materials and bioengineering: programming tissue goals and letting self-construction do the hard work, rather than “machining” anatomy.

Third State

The concept of a Third State challenges the traditional biological boundary between life and death. Death is not a binary switch, but a process where certain cells can transition into new multicellular forms with autonomous behaviors. While an organism (like a human or a frog) may be clinically dead, its individual cells often remain viable for hours or days. Under specific laboratory conditions, these cells do not simply stay "alive"—they reorganize into entirely new biological entities.

Comparison of Biological States

Feature	Life (Organism)	Death (Decay)	The Third State
Control	Centralized (Brain/NS)	None (System failure)	Decentralized (Cellular)
Movement	Skeletal/Muscular	Stationary	Cilia/Self-propulsion
Goal	Reproduction/Survival	Breakdown	Reorganization/Function

== Scientific and Ethical Implications == This discovery has profound implications for how we view the end of life and the future of medicine:

• Medical Robotics: Since Anthrobots are made of human tissue, they could be used to deliver drugs or clear plaque from arteries without triggering an immune response.

• Redefining Death: If our cells can "reboot" into a new form of life after we die, it complicates the legal and philosophical definition of when an individual truly ceases to exist.

• Synthetic Biology: It suggests that the "instructions" for life are more flexible than previously thought, allowing us to create bio-machines from existing cellular building blocks.

Common misunderstandings (quick clarifiers)

• “Cells are intelligent” does not mean “cells are conscious like humans.” It means they can store information, pursue goals, and adapt strategies.
• “Bioelectric fields” does not mean mystical energy. It refers to measurable voltage patterns and ion flows across membranes and tissues.
• “Not the genome” does not mean “genes don’t matter.” It means genes are necessary but not sufficient to specify high-level geometry by themselves.
• "Life Builds Meaning" does "not" mean “mystical destiny” or “evolutionary fate.” It means that agency is scalable: simple systems with local goals (like maintaining pH) cooperate to build larger systems capable of pursuing complex, long-term goals (like building an organ).

Glossary

Agential Materials

Living parts (cells, tissues) that possess preferred states and can act flexibly to reach them. Unlike "passive matter" (which must be forced) or "active matter" (which has energy but no goal), agential matter can sense deviations and employ multiple strategies to correct them.

Agency (Continuum)

In the TAME framework, agency is not a binary property but a measurable continuum. It is defined by a system's ability to pursue goals, store information, coordinate actions, and correct errors across time.

Anthrobots

Multicellular biological robots constructed from human lung cells. They act as "repair bots" capable of autonomous movement and have been observed healing damaged neural tissue in vitro.

Basal Cognition

The scientific view that intelligence and cognition-like properties (sensing, deciding, acting) are not limited to brains but exist on a spectrum across all biological levels, including cells, tissues, and organs.

Bioelectric Code

A computational layer of information residing in the stable resting membrane voltages ($V_{mem}$) of cells. It functions as a "software" layer between genetic "hardware" and anatomical shape, guiding growth and form.

Bioelectric Networks

Networks formed by cells coupled via gap junctions. These networks process information collectively, acting as a shared information layer that coordinates morphogenesis and regeneration toward specific anatomical targets.

Cognitive Light Cone (Goal Scaling)

A metaphor defining the spatial and temporal boundaries of what a system "cares" about. A single cell has a small light cone (local environment), while strongly coupled tissues have a larger light cone (organ-level geometry).

Competence

The autonomous intelligence of sub-components (like cells) that allows a higher-level system to delegate complex tasks (e.g., "build an eye") without micromanaging the movement of every individual part.

Coordination

The capacity of distributed parts to act together as a coherent unit to achieve a collective outcome.

Equifinality (Morphological Plasticity)

The ability of a system to reach the same functional "target state" (e.g., a normal face) despite starting from different configurations (e.g., scrambled organs) or facing external perturbations.

Error Correction

The ability of a system to detect when it has deviated from a target state (such as a wound or deformity) and initiate compensatory actions to restore the target.

Gap Junctions

Channels that electrically and chemically couple neighboring cells, allowing them to share information. This coupling is the mechanism that "softens" individual boundaries, enabling the emergence of a unified, larger-scale self.

Goal-directed Behavior

Activity undertaken by a system specifically to reach or maintain a defined target state.

Memory

Persistent internal states within a system (neural or non-neural) that influence its future behavior.

Morphogenesis

The process by which a single fertilized egg builds a complex 3D anatomy. TAME frames this as a distributed control problem where tissues sense and reduce deviation from a target structure.

Non-local Information Integration

The ability of a cellular collective to sense and make decisions about large-scale targets (like the length of a limb) that no single cell can perceive individually.

Persuadability

A measure of agency based on interaction. High-agency systems can be "persuaded" (influenced by information/signals), whereas low-agency systems must be "pushed" (influenced by physical/chemical force).

Picasso Tadpoles

An experimental example where frog embryos with scrambled craniofacial organs still develop into normal frogs. This demonstrates that development is a goal-seeking process, not a hardcoded script.

Predictive Modeling (Anticipation)

The capacity of a system to use past information to prepare for future states (e.g., a tissue preparing for metabolic shifts).

Problem-solving

The ability to find workable paths to a goal state even when operating under novel constraints or barriers.

Scale-Free Agentialism

A research perspective that studies cognition as a set of capabilities (goal-seeking, memory, etc.) that can appear at any scale of biology, shifting focus from metaphysical consciousness to measurable agency.

Shared Memory (Anonymization)

The process where strong coupling allows individual parts to "forget" their separate identities and act as a single "Self" with shared goals.

TAME (Technological Approach to Mind Everywhere)

A framework by Michael Levin encouraging the study of mind as a continuum. It treats diverse systems (cells, bots, brains) as cognitive agents to improve bioengineering and regenerative medicine.

Target Morphology (Attractor)

The specific anatomical shape a tissue collective "remembers" and strives to maintain. In dynamical systems terms, it is an attractor state that the system naturally converges toward.

The Third State

A biological phenomenon where cells from a deceased organism reorganize into new, functional multicellular forms (like Xenobots) with autonomous behaviors, distinct from both typical life and cellular death.

Valence and Stress Regulation

The ability of a system to "feel" states as favorable or unfavorable. "Stress" is defined as the informational distance from a target state, which drives the system to act.

Xenobots

Synthetic living organisms created from frog skin cells. They demonstrate that cells have the competence to self-organize into new life forms with novel behaviors (like kinematic self-replication) when released from their usual anatomical constraints.