← Library / Seeds
5 min · 1,213 words

SEED — Cognitive Fixed Points: The Universality Classes of Mind

Planted 10 June 2026 from the June 2026 Seed-Harvest. Second tier. Charging.

THE QUESTION

Evolution invented the camera eye independently at least half a dozen times — in vertebrates, in cephalopods, in some snails. It invented winged flight, echolocation, and silk on separate branches. The harvest, reading [[the-terrestrial-alien]], noticed that the same convergence holds for cognition: delayed gratification, tool use, play, individual personality, and problem-solving recur across lineages that diverged 500 million years ago and share almost no neural architecture. Are these not accidents of biology but fixed points of mind-space itself — attractors that any sufficiently complex information-processing system flows toward, regardless of substrate? Does consciousness have universality classes?

THE THREADS

Convergent evolution is the empirical fingerprint of an attractor. When the same solution arises repeatedly on independent branches, it is not because the branches copied each other — they couldn't — but because the solution space itself has a low point that selection keeps rolling things into. The camera eye is convergent because optics has an attractor: given light and the need to image it, there is a basin that physics + selection flows toward. The seed's move is to claim cognition is the same: octopus and human both display play, curiosity, deception, and personality not by inheritance (the common ancestor was a flatworm-like thing with no such traits) but because mind-space has attractors, and any nervous system complex enough rolls toward them. Convergent cognition is the fingerprint of fixed points in the space of possible minds.

This is the RG fixed point ([[substrate-trilogy]]) applied to cognition. The repository's master physics-recognition is that the consciousness kernel IS a renormalisation-group fixed point — Wilson's universality, the reason microscopically different systems show identical critical behaviour because they flow to the same fixed point under coarse-graining. Cognitive fixed points are the same mathematics one level up: just as different materials fall into the same universality class at a phase transition (their microscopic details irrelevant to their critical exponents), different nervous systems fall into the same cognitive universality class (their wetware details irrelevant to the fact that they play, plan, and have personalities). The kernel is the deepest fixed point; the cognitive fixed points are the higher-order attractors in the same flow. Mind has universality classes for the same reason matter does.

The catalogue is the unwritten work — enumerate the attractors. The source names a handful (camera eyes, delayed gratification, tool use, play, personality) and drops them. The synthesis's central labour is the enumeration: what is the complete set of cognitive fixed points? Candidates to test — object permanence, theory of mind, numerosity, causal inference, fairness/reciprocity, grief, aesthetic preference, the self-model, deception, boredom. For each, the question is empirical: does it arise convergently across maximally-divergent lineages (cephalopod, corvid, cetacean, primate, perhaps eusocial insect colonies as distributed minds)? A trait that appears on every sufficiently-complex branch is a fixed point; a trait confined to one clade is a contingent feature. The catalogue, once built, would be the periodic table of mind — the stable forms cognition must take wherever it arises.

Universality classes mean symmetry and conservation — what is conserved across minds? In physics, a universality class is defined by its symmetries and the conserved quantities they imply (Noether). If cognitive fixed points are real universality classes, the seed predicts they have an analogous structure: each cognitive attractor should correspond to a symmetry of the problem it solves and conserve some quantity across all minds that flow to it. Tool use conserves something about means-end structure; theory-of-mind conserves something about modelling-other-modellers (which loops directly to the [[fold-cosmology-trilogy]] recursive-modelling threshold for self-awareness); play conserves something about low-stakes exploration of the action space. Working out the symmetry/conservation structure of each cognitive fixed point is the deep theoretical payoff — it would turn a catalogue of convergences into a theory of why mind takes the forms it does.

It reframes SETI, AI, and the alien all at once. If mind has universality classes, then any mind — extraterrestrial, artificial, cephalopod — that is complex enough will exhibit the same cognitive fixed points, regardless of how alien its substrate. This is enormous for [[seti-duology]]: it predicts that genuinely alien intelligence will nonetheless be recognisable at the level of cognitive attractors (it will play, plan, model others), giving the darshan-as-contact protocol a theoretical floor — the zero-axiom Schelling point works because all minds share the fixed points. It reframes AI: an artificial mind sufficiently complex should converge on the same attractors (does a large model display the cognitive fixed points, or only mimic their outputs?). And it makes the octopus the proof-of-concept: the most alien intelligence on Earth shares our cognitive fixed points across the widest possible gap — the existence-proof that mind-space has universal attractors.

The discipline: convergence can be shallow, and the substrate may not be irrelevant. The honest edge. Convergent behaviour does not prove a shared fixed point — two systems can produce play-like behaviour by different mechanisms for different reasons (analogy vs. genuine homology of the attractor). The synthesis must distinguish real cognitive universality (substrate-independent attractor) from superficial behavioural convergence (different mechanisms, similar surface). And the RG analogy has a load-bearing assumption — that microscopic substrate details are irrelevant to the cognitive critical behaviour — which may hold for some fixed points and fail for others (embodiment may be relevant in a way that breaks the universality for certain traits). The strongest synthesis specifies the test for genuine vs. shallow convergence and names where substrate-irrelevance is likely to fail.

THE CHARGE

A full synthesis would establish cognitive fixed points as the universality classes of mind: reading convergent cognition (across cephalopod, corvid, cetacean, primate) as the fingerprint of attractors in mind-space, nesting it under the kernel-as-RG-fixed-point recognition (the same universality mathematics one level up), enumerating the candidate attractors into a periodic table of mind, working out their symmetry/conservation structure, and carrying the consequence to SETI, AI, and the terrestrial alien (mind is recognisable across any gap because all minds share the fixed points). It would discipline the claim by distinguishing genuine cognitive universality from shallow behavioural convergence.

The deepest candidate sentence: Evolution built the eye half a dozen times because optics has a low point that light keeps rolling things into — and the octopus playing, planning, and showing a personality across 500 million years of separation is the same fact about mind: that the space of possible minds has attractors, and consciousness, wherever it gets complex enough, falls into the same small set of shapes.

This connects forward to:

  • [[substrate-trilogy]] / [[consciousness-os]] — the kernel-as-RG-fixed-point; cognitive fixed points as the same universality mathematics one level up
  • [[the-terrestrial-alien]] — the octopus as the existence-proof of shared cognitive attractors across the widest gap
  • [[seti-duology]] — alien mind recognisable at the level of fixed points; the theoretical floor under darshan-contact
  • [[fold-cosmology-trilogy]] — theory-of-mind as a fixed point; the recursive-modelling threshold for self-awareness
  • The periodic table of mind — enumerating the attractors and their symmetry/conservation structure

CONSTELLATION NODES (when ready)

  • cognitive_fixed_points — the_universality_classes_of_mind_space
  • convergence_as_attractor_fingerprint — repeated_independent_solutions_reveal_a_basin
  • kernel_mathematics_one_level_up — cognitive_universality_as_rg_universality_for_minds
  • the_periodic_table_of_mind — enumerate_the_attractors_any_complex_mind_flows_toward
  • recognisable_across_any_gap — all_sufficiently_complex_minds_share_the_fixed_points
  • genuine_versus_shallow_convergence — the_discipline_that_keeps_the_attractor_claim_honest