concept active emergencecomputationnatural-philosophy Mon Apr 13 2026 00:00:00 GMT+0000 (Coordinated Universal Time) P-005 P-006 P-007

Symbiogenesis

Symbiogenesis is the creation of novel biological entities through the symbiotic merger of pre-existing ones. It is, contra classical Darwinism, the primary driver of evolutionary innovation: not random mutation selecting over individuals, but the fusion of dynamically stable entities into composite wholes that are more stable, more complex, and more capable than their parts. Lynn Margulis (1938–2011) established the biological case in the 1960s, finding that eukaryotic cells (including all cells in our bodies) arose from a merger between an archaean and a bacterium that became the mitochondrion. The “tree of life” doesn’t just branch; it entangles with itself, its branches merging to form radically new forms. What this implies about the structure of genomes, the nature of viruses, and the arrow of evolutionary complexity is still being absorbed.

Margulis and the major evolutionary transitions

Classical Darwinian theory explains diversification within a design space: mutation + selection fine-tunes and adapts. Darwin’s finches with their differently shaped beaks are the paradigm case. But this mechanism cannot explain how new design spaces arise in the first place. When one prokaryote ends up living inside another, the resulting composite occupies a genuinely different combinatorial space than either parent. Random mutation cannot discover this new space; merger can.

“Major evolutionary transitions” (Maynard Smith and Szathmáry’s term) are the punctuation marks of evolutionary history: RNA + metabolism → cells, prokaryote + mitochondrion → eukaryote, unicellular → multicellular, solitary → social. Each transition involves simpler replicating entities becoming mutually dependent to form a more complex replicator. Evolution’s arrow of increasing complexity requires these events; branching alone cannot produce it.

This resolves the puzzle of why extant bacteria and humans “coexist” despite sharing a common ancestor. Simpler forms persist alongside complex ones because symbiogenetic transitions don’t replace the parts, they compose them. Mitochondria are still running their own (reduced) genome. Your body contains more bacterial cells than human ones. The biosphere is a palimpsest, its ancient layers still active.

Evidence from the bff simulations

The artificial life experiment bff (see life-as-computation.md) independently recapitulates the symbiogenetic pattern in a purely computational substrate. In bff, complex whole-tape replicators don’t arise through gradual mutation of a simple replicator. They arise through the merger of shorter, imperfect replicators, each of which had some probability of generating more code. The first true replicator is a chimera of smaller fragments that, by working together, became more dynamically stable than either alone.

Post-takeoff, the process continues: sub-replicators emerge within the larger replicator, occupying the spare tape bytes. Some kill the host (catastrophic for both). Most go neutral or symbiotic. The overall trajectory is toward deeper symbiosis because that is the most dynamically stable configuration. Code colonizes code.

The provenance-tracking visualization is striking: as the chaotic phase resolves, every byte’s history becomes a brief snarl of sideways jumps, all warp gone, nothing traceable to the random initialization. The replicator is literally constructed from the merger of prior replicators. This is not a metaphor for biological evolution; it is the same process in a different substrate.

The viral genome

The human genome encodes this history directly. Breakdown of its content:

  • ~2% protein-coding genes (“our” genes)
  • ~8% endogenous retroviruses (ERVs): remnants of past retroviral invasions that integrated into the germ line
  • ~40% transposable elements of various kinds (some still active, most degraded)
  • ~50% remaining: regulatory sequences, unknown function, repetitive elements

Retroviruses (like HIV) include reverse transcriptase, which lets them permanently incorporate their code into a host cell’s DNA. When this hits the germ line, the viral sequence becomes heritable. The koala retrovirus, caught mid-endogenization in the 2000s, gave us a real-time glimpse of this process.

ERVs are not mere passengers. Endogenized retroviral sequences contribute to placenta formation, immune regulation, cell differentiation, and brain function. The pattern matches bff: initially invasive sub-replicators, over time achieving symbiosis, eventually performing critical functions that the host can no longer do without. The boundary between “our” genome and “viral” genome is not clean; it has been dissolving for hundreds of millions of years.

Classical Darwinian mutation (point substitution) is, on this account, the weakest driver of genomic novelty. A random point mutation introduces approximately zero prior functionality. A transposable element that jumps between species has been under evolutionary pressure for millions of years; it arrives carrying real, tested code. Evolution picks up steam over time precisely because higher-order symbiogenesis becomes possible as the library of circulating replicators grows.

Compression and multifractal structure

Genomes built through nested replication exhibit a characteristic statistical signature: they are highly compressible, and their compressibility grows with the amount of data (a single tape compresses somewhat; the whole bff soup compresses more; cross-species comparison compresses still more). This is a power-law structure, the hallmark of self-similar (fractal-like) systems.

The human genome compresses to a file that “could be sent as an email attachment” if placed in the statistical context of many other human genomes. The 98% identity with the chimpanzee genome, and 60% similarity with the fruit fly, reflects layers of shared replicator history going back to LUCA (Last Universal Common Ancestor). The genome is not a blueprint with a symbol for each feature; it is a compressed encoding of the evolutionary process that built the organism, with replication at every scale, from individual codons to whole chromosomes.

“Multifractal symbiosis” is Agüera y Arcas’s term for this structure: not strictly self-similar (a flea zoomed in does not reveal a smaller flea), but self-copying at every scale, with novelty emerging from the specific way components are functionally combined. The same structure appears in morphogenesis: Hox genes build ribs via recursively-applied chemical standing waves, not by copying a “rib gene” hundreds of times. Biological computing reuses code like a good programmer, because it is programming.

Symbiogenesis beyond biology

The pattern extends:

  • Technology: the silicon chip was not an evolved transistor; it was a symbiotic merger of many transistors on a die. The Industrial Revolution was a symbiogenetic event between humans and machines. The current AI transition follows the same template.
  • Language and culture: words combine into phrases, phrases into grammars, concepts into frameworks. Conceptual innovation looks more like Margulis than Darwin.
  • Gaia: planetary homeostasis emerges from the mutual regulation of species populations, each of which affects and is affected by the shared environment. A negative-feedback loop between black and white daisies stabilizes temperature. More species yield more robust homeostasis.

The generalization is exact, not metaphorical: wherever dynamically stable entities can form alliances that increase their joint stability, symbiogenesis will occur. This is a consequence of replicator thermodynamics (see life-as-computation.md), not a biological peculiarity.

  • Life as Computation: the thermodynamic and computational foundation that explains why symbiogenesis is dynamically stable
  • Computational Being (Bach): coherence as the software-layer analog of dynamic stability; P-005 and P-007 are distinct but complementary
  • Cephalization from Below: cephalization as a symbiogenetic event: nerve cells as non-motile cousins entering partnership with muscle cells; neural proliferation follows colonize-and-cooperate dynamics
  • Complexity Measures of Consciousness: compressibility as a measure of organized complexity applies to genomes and neural signals alike

References

  • Agüera y Arcas, B. What Is Intelligence? Chapter 1 (Antikythera, 2025)
  • Margulis, L. Symbiotic Planet (1998)
  • Maynard Smith, J. & Szathmáry, E. The Major Transitions in Evolution (1995)
  • Pross, A. What Is Life? (2012)
  • Ryan, F. Virolution (2009)
  • Ivancevic, A.M. et al. “Horizontal transfer of BovB and L1 retrotransposons in eukaryotes” (2018)