The Mycelial Record — Fungi as Evolutionary Driver (and its Deflations)
conceptMoney & Finance · Biotech & Transhumanism · Defense & Military-Industrial
Fungi made life on land possible — that part is real and huge. The famous 'trees talking through fungus' story is mostly not. Both are here, together.
Who they are
The fungal kingdom, and the scientists who both built and deflated its reputation — including a 2023 review that took the popular network story apart, and a January 2026 paper that reclassified the biggest land organism of its era.
What they do
What holds: plants could not colonize dry land without fungal partners, preserved in fossil beds and still written into plant genes today. A parasitic fungus hijacks ants' bodies without ever touching their brains — it wires the muscles directly. Fungal electrical pulses are genuinely complex, though the researcher himself allows they might mean nothing.
How it works
What broke: the tree-network story is mapped in two species out of about 73,300, showed benefit in five of twenty-eight controlled experiments, and has no published field evidence for mother trees feeding their own seedlings. The eight-metre giant long called a fungus turns out to have the wrong body, the wrong chemistry, and none of the material fungi are made of. And the theory that mushrooms grew the human mind has no mechanism and no fossil trace.
Why it matters
Because the honest version is better than either sales pitch: fungi are one of the great forces in the history of life, and the story people repeat about them is a documented case of a claim outrunning its evidence — inside peer review, not outside it.
The engine's record — word for word
[Report #183] The engine's first fungal surface — the documented record with its corrections attached, minted because fungi had ZERO canon presence while carrying one of the largest documented roles in the history of life. WHAT HOLDS AT TIER: land plants could not colonize the continents without fungal partners — the 407-Myr Rhynie chert preserves arbuscule-like structures 'morphologically identical to those of living arbuscular mycorrhizae' (Remy et al., PNAS 1994), and the symbiosis-signalling gene toolkit is functionally conserved in liverworts, demonstrating maintenance since the most recent common ancestor of land plants (PNAS 2024) [caveat: the pathway is repeatedly LOST in lineages that abandoned symbiosis — 'all land plants have it' would be wrong]. PARASITIC PUPPETRY, DOCUMENTED: Ophiocordyceps unilateralis controls ant behaviour with 'no fungal cells observed inside the brain' — hyphae encircle the mandibular muscles and drive the death-grip by hypercontraction (PNAS 2017 for the 3D reconstruction; the hypercontraction/sarcomere damage is Mangold et al., J Exp Biol 2019, which replaced the earlier atrophy account). Peripheral biochemical control, not neurological possession. ELECTRICAL SIGNALLING, MEASURED AND UNRESOLVED: Adamatzky (R. Soc. Open Sci. 2022) recorded spike trains in four species (durations 1-21 h, amplitudes 0.03-2.1 mV) clustering into a lexicon of up to ~50 'words' whose length distribution tracks human languages [note: the published abstract's own typo reads 'Liz-Zempel' for Lempel-Ziv complexity — quoted faithfully wherever cited]. Adamatzky's own hedge: 'There is also another option — they are saying nothing.' Dan Bebber (Exeter): interpretation as language 'seems somewhat overenthusiastic.' Physiology measured; cognition not demonstrated. THE DEFLATIONS THE POPULAR STORY OMITS: Karst, Jones & Hoeksema (Nature Ecology & Evolution, 2023) found common mycorrhizal networks rigorously field-mapped in only TWO of ~73,300 tree species; in the best-controlled experiments only 5 of 28 (18%) showed positive network effects not offset by negative root effects; and the signature claim — mother trees preferentially feeding kin — has 'no peer-reviewed, published evidence.' Unsupported citation rates roughly doubled over 25 years, reaching ~50% on function claims: a documented citation-cascade, and a live instance of the engine's epistemic-drift pattern operating inside peer-reviewed science. PROTOTAXITES RECLASSIFIED: the 8-metre Devonian giant, long read as a massive fungus on Boyce et al. (2007) carbon-isotope evidence (~13 permil variance inconsistent with photosynthesis), was found by Science Advances (21 Jan 2026) to have anatomy matching no known fungus, lignin-like wall chemistry and NO chitin while true fungi in the same chert have it — an extinct eukaryotic lineage of its own; Boyce concedes 'there is no good place to put Prototaxites in the fungal phylogeny.' Both readings carried; 'the biggest land organism was a mushroom' no longer holds as a flat statement. STONED APE: FALSIFIED as mechanism — no primary evidence links psilocybin to hominid neurogenesis, and the theory's strongest peer-reviewed sympathizers (Rodríguez Arce & Winkelman 2021) concede direct evidence 'is lacking' and call McKenna's strong version 'most certainly false.' APPLIED FUNGAL ENGINEERING, REAL: NASA's 'Mycotecture Off Planet' (Rothschild, Ames) reached NIAC Phase III in 2024 (~$2M) toward grown mycelium habitats; Cornell's Science Robotics (Aug 2024) work drove robots from live mycelial electrical spikes, including full override of the native signal; Chernobyl's melanized fungi are documented (Zhdanova 2004) while 'radiosynthesis' (Dadachova & Casadevall 2007) remains unreplicated and the 2022 ISS shielding result (~2%) may be ordinary mass shielding. THE ENGINE READ: fungi are a documented macroevolutionary driver AND the popular 'intelligent network' story is a documented over-claim — both true, held together. No fungal-NHI wiring; resemblance is not evidence.
Follow the trail
Walk this on the live map →