Endless forms most beautiful 2.0: teleonomy and the bioengineering of chimaeric and synthetic organisms

What to take away

1. 1. Xenobots — proto-organisms derived solely from dissociated Xenopus frog skin cells with no transgenic modification — self-assemble within 48 hours into motile, self-repairing spherical constructs that discover kinematic self-replication, a reproductive mode not observed in any other known organism (Kriegman et al., 2020, 2021).
2. 2. Fankhauser (1945a, b) showed that polyploid salamander embryos maintain normal kidney tubule cross-section by reducing participating cell number as individual cell size increases, with extreme ploidy causing a single cell to wrap around itself — demonstrating that morphological goal states are enforced by whatever molecular mechanism is available, not by a fixed cellular program.
3. 3. Tadpoles with primary eyes removed and eye primordia grafted onto their tails nonetheless achieve light-mediated visual learning, with ectopic optic nerves connecting to the spinal cord and the brain integrating the aberrant input into behavior (Blackiston & Levin, 2013).
4. 4. Planarian flatworms are naturally mixoploid — different cells within one individual carry different chromosome numbers — yet are champions of regeneration, rebuilding anatomically correct worms from small fragments, demonstrating that high anatomical fidelity is decoupled from genomic uniformity.
5. 5. Bioelectric circuit manipulation, not genomic editing, produces permanently two-headed planarian strains that propagate the altered body plan across regeneration events (Oviedo et al., 2010; Durant et al., 2017), showing that anatomical set points are stored in physiological rather than genomic memory.
6. 6. Willett et al. (2021) achieved high-performance brain-to-text communication in a paralyzed patient by decoding imagined handwriting from intracortical microelectrode array recordings of the precentral gyrus, producing written sentences in real time at speeds exceeding prior brain-computer interface approaches.
7. 7. The multi-scale competency architecture (MSCA) is introduced as the paper's central explanatory construct: each subsystem from molecular networks to organs maintains homeostatic loops pursuing goals in its own problem space, enabling collectives to reach target morphologies by diverse means and accelerating evolutionary search by providing reliable, goal-directed components that higher levels can exploit without micromanagement.
8. 8. To replicate the core experimental logic, researchers can dissociate epidermal cells from stage-matched Xenopus laevis embryos, culture them in isolation from organismal instructive signals, and assay for emergent motility, damage repair, and population-level kinematic self-replication within 48 hours, as described in Blackiston et al. (2021).
9. 9. The paper raises the open question of whether the behavioral and anatomical goals of Xenobots — which were never directly selected for by evolution — are best explained by generic laws of form, mathematical constraints, or emergent autopoiesis, and explicitly states that current developmental genetics models cannot predict that anatomically coherent outcomes would arise from such genomically unmodified, contextually novel cellular collectives.
10. 10. Because teleonomy is substrate-independent and scales continuously from simple homeostatic circuits to reflective cognition, the paper predicts that legal and ethical frameworks built on crisp species or 'human' categories will be functionally inadequate within decades, and proposes that moral consideration should instead be indexed to the spatiotemporal scale and competency of an agent's goal-directed behavior.

Peer brief — for seminar discussion

Clawson and Levin's 2023 paper in the Biological Journal of the Linnean Society argues that teleonomy — goal-directed, problem-solving behavior in diverse problem spaces — is the substrate-independent invariant that should replace genomic, phylogenetic, and origin-based categories as the organizing framework for understanding, engineering, and ethically evaluating living agents. The paper is theoretical and synthetic rather than reporting a single new experiment, but it is grounded throughout in specific empirical demonstrations. Xenobots (Kriegman et al., 2020, 2021), derived from wild-type Xenopus frog skin cells with zero genomic modification, self-assemble into motile proto-organisms in 48 hours and discover kinematic self-replication — a mode of reproduction absent from the rest of the known biosphere. Fankhauser's classic polyploid salamander work shows that when cell size is artificially expanded by increasing ploidy, fewer cells participate in tubulogenesis to preserve normal tubule cross-section, with extreme cases producing a single cell that wraps around itself. Blackiston and Levin's 2013 ectopic-eye tadpole experiments show that the brain integrates visual input arriving via the spinal cord, and Willett et al.'s 2021 intracortical MEA work on imagined handwriting decoding in a paralyzed patient demonstrates the same principle in an engineered chimaeric context. The multi-scale competency architecture (MSCA) is the paper's named theoretical construct: it holds that every subsystem from molecular networks to organs pursues goals in its own problem space via homeostatic and allostatic loops, and that this nested goal-directedness simultaneously accelerates evolutionary search and enables robust response to novel perturbations without requiring aeons of selection on specific configurations. The load-bearing implication is that current categories — organism vs. machine, evolved vs. designed, cognitive vs. non-cognitive — are contingent artifacts of technological limitation rather than natural kinds, and that teleonomy parameterizes a continuous spectrum on which Xenobots, hybrots like the MEART robotic-arm drawing system controlled by rat cortical tissue on a microelectrode array, and brain-computer interface cyborgs all sit without principled discontinuity. The paper explicitly predicts that rational bioengineering, guided by MSCA principles of top-down goal-state manipulation rather than bottom-up molecular micromanagement, will outperform evolution in generating coherent novel organisms, and that ethical and legal frameworks indexed to species or human-IQ cutoffs will require complete replacement. An alternative methodological framing the paper could have used is agent-based modeling of morphogenetic goal-state navigation, which would operationalize MSCA predictions quantitatively rather than relying on analogical synthesis across disparate experimental systems. The most pointed criticism a careful reader should raise is that teleonomy as deployed here risks being unfalsifiable: by defining goal-directedness as detectable by an observer with sufficient intelligence, the framework can be extended to virtually any dynamical system, and the paper offers no criterion for what would count as a system that genuinely lacks teleonomy rather than merely having teleonomy that is too subtle to detect. This conflation of apparent goal-directedness with a real invariant property of life could undermine the framework's empirical traction precisely in the hardest cases — minimal synthetic life, early proto-biological chemistry — where the distinction matters most for both regenerative medicine and the ethics the authors want to derive from it.

Methods (5)

• Closed-loop techniques
  Feedback-based coupling between neural cultures and robots/virtual environments.
• Immunofluorescence imaging
  Imaging method used to visualize neural synapses and hyphal bodies.
• Microelectrode array (MEA)
  Device to record and stimulate electrical activity of neural cultures.
• Principal components analysis (PCA)
  Statistical method used to analyze neural activity data.
• Three-dimensional reconstruction techniques
  Methods for visualizing fungal networks in ants.

Frameworks (5)

• Bioelectric Networks
  Ancient mechanism using ion channels, gap junctions, and neurotransmitters as 'cognitive glue' enabling collective intelligence across morphogenesis and behavior.
• Diverse Intelligence
  Research program studying intelligence at multiple scales and substrates; proposed as relevant to implications of mnemonic improvisation.
• Guided self-assembly
  Approach of nudging components with agendas to achieve desired outcomes rather than micromanaging.
• Multiscale Competency Architecture
  A framework originating from Levin that formalizes how hierarchical biological systems—from cells to tissues to organs—exhibit integrated problem-solving and adaptive plasticity across multiple levels of organization (metabolic, transcriptional, physiological, anatomical). It models system-level behaviors as emergent from competition and cooperation among heterogeneous subunits within composite agents, explaining how goals and regulations scale across biological scales.
• Synthetic Morphology
  Framework for bioengineering via active and agential matter; cited as relevant unification domain.

Findings (22)

• Enlarged polyploid cells produce normal-sized kidney tubules by adjusting cell number and using cytoskeletal bending

  Shows diverse molecular mechanisms serve higher-level anatomical specification despite radical changes in cell size and quantity.

• Single newt cell can wrap around itself to form a kidney tubule when cell size is artificially increased.

  From Fankhauser (1945), demonstrates diverse molecular mechanisms serving a higher-level anatomical specification.

• Planarians maintain high regenerative fidelity despite genetic heterogeneity and chromosomal variation
• Brain-to-text communication via handwriting achieved 90 characters per minute in a paralyzed person.

  From Willett et al. (2021), shows high-performance artificial chimaera.

• MEART created drawings with neural control and showed learning when training stimulus was updated.

  From Bakkum et al. (2007b), demonstrates closed-loop learning in a hybrot.

• Tadpoles with displaced craniofacial organs can still develop normal face through organ movement.

  From Vandenberg et al. (2012) and Pinet et al. (2019), reveals regulative morphogenesis.

• Tadpoles with ectopic eyes on tail can see and integrate sensory input from aberrant location

  Demonstrates neural plasticity: brain adapts behavioral programs to sensory input from abnormal anatomical locations within single organism lifetime.

• Ectopic eyes on tadpole tails support visual learning despite connecting to the spinal cord.

  From Blackiston & Levin (2013), shows plasticity of brain and body.

• Ophiocordyceps unilateralis forms a fungal network invading ant muscles but not the brain.

  From Fredericksen et al. (2017), demonstrates how a parasite can control host behavior without neural takeover.

• Wild-type frog skin cells form novel proto-organisms (Xenobots) without genomic editing.

  From Blackiston et al. (2021) and Kriegman et al. (2020), reveals emergent goals from cellular collectives.

Claims (24)

• Significant gaps exist in our understanding of the origin of anatomical novelty, of the relationship between genome and form, and of strategies for control of large-scale structure and function.

  Author assertion about current knowledge gaps.

• Cybernetic perspectives on chimaeric morphogenesis erase artificial distinctions established by past limitations of technology and imagination.

  Author argument for cybernetic framework.

• Evolution does not simply make hardwired machines that execute a predetermined set of steps; instead, it produces problem-solving hardware.

  Key insight about the nature of evolved systems.

• The life vs. machine distinction is relevant only to the lowest class of machines available in prior decades and is not fundamental.

  Claim that the organism–machine dichotomy is outdated.

• The option space of beings enables us truly to see life for the first time, in the absence of standard phylogenetic relationships.

  Epistemological claim about the role of novel organisms.

• Teleonomic behavior in diverse problem spaces is a powerful invariant across possible beings regardless of composition or origin.

  Central claim proposing teleonomy as a unifying parameter.

• Analysis of living forms that have never existed before is necessary to reveal deep design principles of life as it can be.

  Central premise of the paper.

• Chimaerism is a type of conceptual universal acid, dissolving existing terminology that is not based on deep concepts.

  Argument that hybrid systems break outdated categories.

• Problems of cognition (mind) and morphogenesis (body) share fundamental commonality in competent problem-solving.

  Authors propose deep principle unifying neuroscience and developmental biology through goal-directed behavior as common thread.

• A multi-scale competency architecture facilitates evolution of robust problem-solving, living machines.

  Key proposal linking multi-scale agency to evolutionary advantage.

Hypotheses (7)

• Example of extreme unconventional agent combining mammalian neural culture, insect neurons, robotic swarm body, and social media-based environment
• Life exploits multi-scale competency architecture enabling adaptation to novel circumstances much faster than evolution alone.

  Authors hypothesize that plasticity observed in individual lifetimes suggests architecture providing greater efficiency than blind evolutionary search.

• Addition of neural tissue to standard brains will likely result in increased processing capacity due to adaptive design.

  Prediction about the plasticity of neural systems.

• Xenobots’ behavior reveals baseline geodesics through option space that are normally masked by larger collectives.

  Hypothesis about the origin of novel goals in synthetic organisms.

• Multi-scale competency increases the apparent IQ of the evolutionary process, enabling better generalization.

  Hypothesis linking competency to evolutionary learning efficiency.

• Teleonomy provides the ratchet that drives the great transitions of cognitive capacity along the continuum.

  Hypothesis about the role of goal-directedness in major evolutionary transitions.

• Morphogenesis is a collective intelligence problem, not merely genetic execution

Questions (15)

• How does evolution capitalize on the laws of physics and computation to generalize so well from specific examples to highly diverse possible instantiations?

  Question about the deep principles enabling evolvability.

• What is the relationship between the genome and anatomy, and what mechanisms allow biology to exhibit robustness and plasticity simultaneously?

  Driving question for the research program.

• What is the relationship or overlap between the sets demarcated by ‘life’ and ‘cognition’?

  Probes the boundary between life and mind.

• How do organisms achieve robustness and plasticity simultaneously in novel circumstances?

  Core question motivating bioengineering studies: how can biological systems adapt to unexpected perturbations while maintaining coherence?

• Lack of predictive science for controlling morphogenesis in regenerative medicine
• Can we make new autonomous robotics via a multi-scale competency architecture?

  Practical question about engineering robust robots.

• Where do the specific goal states come from? What determines the set points?

  Question about the origin of teleonomic targets.

• How can we develop a semiotics of synthetic agents and their Umwelten?

  Question about communication and meaning in novel systems.

• Could there have been very early life forms that were not teleonomic?

  Question about minimal life and goal-directedness.

• Hybrid frog-axolotl embryos defy prediction from constituent genomes

Related work— refs + corpus + external arXiv

Cited / in-corpus / arXiv badges show which signals surfaced each row. Multi-source rows weighted higher.

• Darwin's agential materials: evolutionary implications of multiscale competency in developmental biology

  in corpus

  2023
  ≈ 91%
• The computational boundary of a 'self': developmental bioelectricity drives multicellularity and scale-free cognition

  cited

  in corpus

  2019
  ≈ 86%
• Collective intelligence: A unifying concept for integrating biology across scales and substrates

  in corpus

  2024
  ≈ 91%
• Technological Approach to Mind Everywhere: An Experimentally-Grounded Framework for Understanding Diverse Bodies and Minds

  cited

  in corpus

  2022
  ≈ 83%
• A cellular platform for the development of synthetic living machines

  cited

  2021
  ≈ 86%
• The scaling of goals via homeostasis: an evolutionary simulation, experiment and analysis

  Johanna Bischof, Jennifer V. LaPalme, and Michael Levin Leo Pio-Lopez

  2022
  ≈ 85%
• Developmental Bioelectricity: the cognitive glue enabling evolutionary scaling from physiology to mind

  in corpus

  2023
  ≈ 84%
• The collective intelligence of evolution and development

  in corpus

  2023
  ≈ 84%
• Empowered Neural Cellular Automata

  Caitlin Grasso and Josh Bongard

  2024
  ≈ 83%
• Re-membering the body: applications of computational neuroscience to the top-down control of regeneration of limbs and other complex organs

  cited

  2015
  ≈ 83%
• Generalizing frameworks for sentience beyond natural species

  in corpus
  ≈ 83%
• The biogenic approach to cognition

  in corpus

  2005
  ≈ 83%
• Remapping and navigation of an embedding space via error minimization: a fundamental organizational principle of cognition in natural and artificial systems

  L\'eo Pio-Lopez, Chris Fields, Michael Levin Benedikt Hartl

  2026
  ≈ 82%
• Microelectronic Morphogenesis: Progress towards Artificial Organisms

  Daniil Karnaushenko, Minshen Zhu and Oliver G. Schmidt John S. McCaskill

  2024
  ≈ 82%
• AI: a Bridge toward Diverse Intelligence and Humanity’s Future

  in corpus

  2024
  ≈ 82%
• Neural cellular automata: applications to biology and beyond classical AI

  Michael Levin, L\'eo Pio-Lopez Benedikt Hartl

  2025
  ≈ 82%
• Competency of the Developmental Layer Alters Evolutionary Dynamics in an Artificial Embryogeny Model of Morphogenesis

  Lakshwin Shreesha and Michael Levin

  2022
  ≈ 82%
• Morphological Coordination: A Common Ancestral Function Unifying Neural and Non-Neural Signaling

  cited

  2019
  ≈ 82%
• Collective control of modular soft robots via embodied Spiking Neural Cellular Automata

  Eric Medvet, Stefano Nichele, Sidney Pontes-Filho Giorgia Nadizar

  2022
  ≈ 81%
• Bootstrapping Life-Inspired Machine Intelligence: The Biological Route from Chemistry to Cognition and Creativity

  Michael Levin Giovanni Pezzulo

  2026
  ≈ 81%
• Self-Improvising Memory: A Perspective on Memories as Agential, Dynamically Reinterpreting Cognitive Glue

  in corpus

  2024
  ≈ 81%
• Cellular Plasticity Model for Bottom-Up Robotic Design

  Thomas J. Smith, Nicholas S. Szczecinski, Sergiy Yakovenko, and Yu Gu Trevor R. Smith

  2024
  ≈ 81%
• Swarms, Phase Transitions, and Collective Intelligence

  LANL and Santa Fe Institute) Mark M. Millonas (Center for Nonlinear Studies and Theoretical Division

  2008
  ≈ 81%
• Are Biological Systems More Intelligent Than Artificial Intelligence?

  Michael Timothy Bennett

  2026
  ≈ 81%
• Basic principles drive self-organization of brain-like connectivity structure

  Valeriy A. Makarov Slizneva, and Cees van Leeuwen Carlos Calvo Tapia

  2024
  ≈ 80%
• Biology, Buddhism, and AI: Care as the Driver of Intelligence

  in corpus

  2022
  ≈ 80%
• Synthetic living machines: a new window on life

  cited

  2021
  ≈ 80%
• Artificial Collective Intelligence Engineering: a Survey of Concepts and Perspectives

  Roberto Casadei

  2023
  ≈ 80%
• Knowing one's place: a free-energy approach to pattern regulation

  cited

  2015
  ≈ 80%
• Multiple ways to implement and infer sentience

  in corpus
  ≈ 80%

+27 more

Similar preprints — Semantic Scholar

Cited by (11)

• Developmental Bioelectricity: the cognitive glue enabling evolutionary scaling from physiology to mind

  Developmental bioelectricity—the network of ion channels, gap junctions, and neurotransmitter-mediated Vmem (membrane potential) dynamics operating across all body cells, not just neurons—functions as

• Darwin's agential materials: evolutionary implications of multiscale competency in developmental biology

  Cellular collectives operating between the genotype and anatomical phenotype constitute an agential substrate that fundamentally reshapes the evolutionary search process—this is the central claim of L

• Self-Improvising Memory: A Perspective on Memories as Agential, Dynamically Reinterpreting Cognitive Glue

  Memory's function is not fidelity but salience preservation through active reinterpretation—a claim Levin defends by unifying phenomena from caterpillar-to-butterfly metamorphosis (where training-indu

• Biology, Buddhism, and AI: Care as the Driver of Intelligence

  Care—defined as concern for the alleviation of stress (the delta between current and optimal states)—is proposed as the substrate-independent invariant that unifies biology, Buddhist philosophy, and a

• Toward an ethics of autopoietic technology: Stress, care, and intelligence

  The Stress-Care-Intelligence (SCI) loop — a substrate-neutral feedback model in which homeostatic stress engenders care, care activates the latent capacity of intelligence, and successful problem-reso

• Stress sharing as cognitive glue for collective intelligences: A computational model of stress as a coordinator for morphogenesis

  Stress sharing between cells accelerates collective morphogenesis and expands individual cells' cognitive light cone, as demonstrated in a multiscale agent-based computational model of embryogenesis p

• Multiple ways to implement and infer sentience

  Rouleau and Levin argue that the evidential standard applied to inferring sentience is inconsistently applied: behavioural signatures used to attribute subjective experience to non-human animals — goa

• AI: a Bridge toward Diverse Intelligence and Humanity’s Future

  Current AI debates are importantly incomplete because they fixate on large language models while ignoring the broader space of impending minds — including cyborgs, hybrots, genetically augmented human

• The Causally Emergent Alignment Hypothesis: Causal Emergence Aligns with and Predicts Final Reward in Reinforcement Learning Agents

  Causal emergence in the latent-space representations of reinforcement learning agents is consistently predictive of final reward and aligns dynamically with reward improvement across training — a find

• Brains and where else? Mapping theories of consciousness to unconventional embodiments

  Rouleau and Levin's 2026 analysis in *Philosophical Transactions of the Royal Society A* (384, issue 2320) demonstrates that the functional and operational principles underlying most contemporary theo

• Cognitive glues are shared models of relative scarcities: the economics of collective intelligence

  Levin and Lyons argue that the price system in market economies is not merely a resource-allocation device but the canonical instantiation of what they term a *cognitive glue* — a coordinating afforda