Pith. sign in

REVIEW 4 major objections 5 minor 37 references

Hamiltonian Formalism for Comparing Quantum and Classical Intelligence

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper proposes that the fundamental difference between classical and quantum artificial general intelligence is an algebraic one: in a classical agent every Hamiltonian generator commutes, while in a quantum agent they do not.

desk verdict A suggestive but flawed framework: the central commutativity dichotomy is false in the paper's own toy model, and the quantum induction generator is a scalar; still worth a serious look as a conceptual proposal. read the letter →

arxiv 2506.14456 v1 pith:SLBKMLP6 submitted 2025-06-17 quant-ph cs.AI

Elija Perrier This is my paper · ORCID
classification quant-phcs.AI PACS 03.65.-w03.67.-a
keywords HamiltonianformalismquantumAGIclassicalcommutationrelationsgeneratordecompositionagent-environmentinteractioncontextualityinformation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes a Hamiltonian-based language for comparing classical and quantum artificial general intelligence (AGI) on equal footing. It splits an agent's total dynamics into a sum of Hamiltonian generators, one for each core function such as induction, reasoning, recursion, learning, sensing, and memory, and writes a classical and a quantum version of each generator. The central claim is that the qualitative difference between classical and quantum agents is algebraic: every classical generator commutes with every other, while quantum generators generally do not, so that in a quantum agent sensing can disturb learning, logical truth becomes context-dependent, and computational histories can exist in superposition. This matters because it turns the question of what a quantum substrate changes about intelligence into a precise mathematical condition that can be examined generator by generator.

What carries the argument

The central object is the total Hamiltonian $H = \sum_k H_k$, decomposed into functional generators for induction ($H_{\mathrm{ind}}$), reasoning ($H_{\mathrm{reas}}$), recursion ($H_{\mathrm{rec}}$), learning ($H_{\mathrm{learn}}$), and sensing/environment ($H_{\mathrm{sens}}, H_{\mathrm{env}}$). Classical generators are smooth functions on the cotangent bundle of configuration space, evolving via Poisson brackets $\dot f = \{f, H_C\}_{PB}$; quantum generators are operators on a Hilbert space evolving via the commutator $i\hbar \dot\rho = [H_Q, \rho]$. The machinery's work is to convert each cognitive capability into an energy term and to let the vanishing or non-vanishing of the brackets/commutators among those terms carry the explanation of behavioral difference: zero brackets give simultaneous determinability and order independence, nonzero commutators give back-action, contextuality, and superposition.

What would settle it

Build a classical agent whose sensing and learning updates are coupled by a deterministic non-commuting rule and check whether it reproduces the measurement-induced disturbance of learning that the paper attributes to non-commuting quantum generators; if the classical system shows the same qualitative effect, commutation fails as the distinguishing criterion. Alternatively, on a two-qubit NISQ implementation of the toy Hamiltonian $H_Q = \kappa(|1\rangle\langle 0|_m \otimes Z_E + \mathrm{H.c.}) + \mu(I-\Pi_\alpha) + gX_A + J Z_A Z_{A'}$, measure whether enabling the sensing term increases learning error at the predicted rate $\gamma\sim\kappa^2$ and compare to a classical simulation.

Watch

Extended reading notes

Core claim

The paper argues that the commutation relations among Hamiltonian generators are the fundamental distinction between classical and quantum agents. For each cognitive function, it constructs a classical generator as a function on phase space with vanishing Poisson brackets against the other generators, and a quantum generator as a self-adjoint operator with generally nonzero commutators. Through explicit toy Hamiltonians, it shows that a classical agent interfacing with a quantum environment must rely on commuting classical-to-classical, classical-to-quantum, and quantum-to-classical channels, whereas a quantum agent can additionally use non-commuting quantum-to-quantum interactions. The result is that measurement back-action, contextual logical reasoning, and superposition-based recursion are features unique to the non-commuting quantum regime, whereas the classical regime is characterized by order-independent, independently evaluable subprocesses.

Load-bearing premise

The framework assumes that an agent's cognitive dynamics can be faithfully written as a sum of Hamiltonian generators, one per function, and that the commutation properties of those energy operators capture the real differences between classical and quantum intelligence; the paper asserts this mapping without proving its domain of validity.

Editorial extensions

If this is right

  • Within this formalism, a classical AGI cannot exhibit disturbance of its own learning by its own sensing; any such self-disturbance implies non-commuting generators and hence a quantum substrate.
  • A quantum AGI's logical reasoning is inherently contextual: the order in which it evaluates logical clauses can change the truth values it assigns, a direct consequence of non-commuting projectors in $H_{\mathrm{reas}}$.
  • The sensing generator predicts a concrete, testable decoherence rate $\gamma \sim \kappa^2$ for the environment when a quantum agent reads a pointer, linking the formalism to standard open-quantum-system dynamics.
  • The decomposition into commuting vs non-commuting generators extends naturally to open and many-body environments and to embedding alignment constraints as additional generators, which the paper lists as future work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the commutation criterion is accepted, a necessary condition for genuine quantum advantage in agency is that at least one pair of functional generators fails to commute; agents whose generators all commute are operationally classical even if run on quantum hardware.
  • The formalism suggests an experimental route: measure the back-action of sensing on learning in a small quantum agent and compare the disturbance to the classical commuting simulation; agreement with the quantum prediction would validate the generator mapping, and agreement with the classical one would falsify it.
  • The treatment of recursion via a history state implies that self-inspection in a quantum agent collapses the superposed computational history, a structural obstacle to quantum self-modification that the paper notes but does not develop; this could bear on the safety of self-improving agents.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a Hamiltonian-based framework for comparing classical and quantum artificial general intelligence (CAGI and QAGI). It decomposes an agent's dynamics into Hamiltonian generators associated with induction, reasoning, recursion, learning, sensing, and environment interaction, giving a classical phase-space realization and a quantum operator realization for each. The central thesis is that CAGI generator terms commute completely while QAGI terms do not, and that this difference produces qualitatively different agent-environment dynamics, including measurement back-action and contextual reasoning. The framework is illustrated with a toy example in which a classical agent and a quantum agent act on a qubit environment, and the paper concludes that non-commutativity is the fundamental algebraic distinction between the two.

Significance. If the proposed framework were correct, it would provide a unified Hamiltonian language for comparing classical and quantum AGI and would identify non-commutativity as a single algebraic marker of quantum-agent behavior. The paper deserves credit for making the comparison concrete: it gives explicit generators, connects them to channel types (CTC, CTQ, QTC, QTQ), and correctly notes that quantum measurement back-action and contextuality have no classical analogue in the usual sense. However, the paper provides no machine-checked proofs, reproducible code, or falsifiable predictions, and the central commutativity dichotomy is contradicted by the paper's own equations. The significance of the proposal is therefore not established by the manuscript as it stands.

major comments (4)
  1. [Example Comparison Hamiltonian, Eq. (10); Appendix B, Eq. (12)] The central claim that CAGI terms commute completely is false under the paper's own Poisson-bracket definition. The learning term p_theta^2/(2m) + lambda|theta| and the logic term mu delta((1-q_m)q_A - 1) both depend on theta through q_A = 1{theta > 0}, so their Poisson bracket is not identically zero. More generally, even a single classical learning Hamiltonian fails the claimed property: {p_theta^2/(2m), lambda|theta|} = -lambda p_theta sign(theta)/m, which is nonzero. In addition, the final term eta(t) q_A Z_E contains the quantum operator Z_E, so Eq. (10) is not a real-valued phase-space function and the Poisson bracket of Eq. (12) does not apply to it. The claimed complete commutation is therefore either false or ill-defined, and the qualitative CAGI/QAGI dichotomy is left without a valid demonstration.
  2. [Section 2, Induction, Eq. (2)] The quantum induction generator H^Q_ind = k_B T S(rho_D || rho_theta) is a scalar number, not an operator on the agent Hilbert space, because relative entropy is a trace. Hence [H^Q_ind, rho_D] = 0 identically, and the claim that non-commuting rho_D and rho_theta imply [H^Q_ind, rho_D] is nonzero is incorrect. The induction generator as written generates no Schrodinger evolution; an operator-valued quantity such as a term involving log(rho_D) - log(rho_theta) would be needed.
  3. [Section 2, Sensing, Eq. (7)] The statement that {q_env, H^C_sens} = 0 shows the environment is not disturbed is incomplete. With H^C_sens = kappa P delta(q_sens - q_env), Hamilton's equation gives p_dot_env = -partial H^C_sens / partial q_env = kappa P delta'(q_sens - q_env), which is generically nonzero. A coordinate-coupled classical sensor therefore does exert back-action on the environment, so the model does not support the claim that CTC sensing leaves the environment untouched.
  4. [Section 2, Generator Decomposition Analysis] The framework assumes without argument that cognitive functions such as induction, reasoning, recursion, learning, and sensing can be faithfully represented by the specific Hamiltonian generators in Eqs. (1)-(10), and that Poisson or operator commutativity of these generators captures the cognitive difference. No derivation, uniqueness, or domain of validity is provided; the footnote that follow-on work detailing bounds and limitations is in preparation confirms this gap. Because the central comparison rests on this mapping, the paper is currently a set of definitions rather than a theory with falsifiable content.
minor comments (5)
  1. [Abstract and title] The manuscript contains many spacing and OCR-like artifacts, such as 'Theprospect' and 'HamiltonianFormalism' in the abstract; these should be corrected in a published version.
  2. [Section 2, Induction and Learning] Eq. (1) and the later H^C_learn both contain a kinetic term and a loss term, but the formal distinction between induction and learning is not stated; please clarify how these two generators differ.
  3. [Eq. (6)] The clock Hilbert space has basis |t> for t = 0,...,L, but the sum runs over t = 0,...,L-1; please state the clock dimension explicitly and define the halt projector Pi_halt and the initial state |psi_0>.
  4. [Table 1] The symmetric logarithmic derivative L_i in the QAGI momentum row is not defined in the table or in the text; please add a definition or a reference.
  5. [Notation] The symbol delta is used both as a Dirac delta (e.g., Eqs. (3) and (7)) and as a Kronecker-like indicator (e.g., Eq. (10)); these uses should be disambiguated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is a framework proposal whose commutation-based classical/quantum contrast is a stated modeling premise, not a derived prediction.

full rationale

I walked the claimed derivation chain and found no step where a result reduces to its own input. The paper fits no parameters to data and makes no empirical predictions; the example Hamiltonians are explicitly illustrative, and the author even flags the recursion ansatz as 'open to debate.' The central contrast between commuting classical observables and non-commuting quantum operators is standard Hamiltonian and operator-algebra content, imported as the formalism's starting point rather than derived from it. The only self-citation is [23], listed among geometry references in the background section; nothing load-bearing depends on it, and no uniqueness theorem from the author's prior work is invoked. The skeptical concern that the CAGI terms in Eq. (10) may not Poisson-commute is a potential internal-consistency or correctness issue in the example, not circularity, because it does not show that an output was equivalent to an input by construction. Accordingly, I identify no circular step and assign score 0.

Assumptions & free parameters 6 free parameters · 6 assumptions · 2 invented entities

The central claim rests on a set of ad hoc modelling choices (the specific Hamiltonians) plus standard mathematical background. The main free parameters are coupling strengths and energy scales chosen by hand; they are not fitted to data because the paper makes no quantitative predictions. The core axiom, that cognitive functions correspond to Hamiltonian generators, is asserted without independent justification. No new physical entities are introduced beyond the generator decomposition itself.

free parameters (6)
  • k_B T (temperature scaling) = not specified, tunable
    Rescales relative entropy to energy units in Eq. (2); chosen by hand, no physical basis provided.
  • mu (penalty weight) = not specified, tunable
    Weights logical inconsistency penalties in H^C_reas and H^Q_reas (Eqs. 3-4).
  • lambda (loss-to-energy scale) = not specified, tunable
    Sets the scale of the loss term in H^C_learn and the CAGI toy Hamiltonian (Eq. 10).
  • kappa (sensing coupling) = not specified, tunable
    Coupling strength between sensor pointer and environment in H^C_sens and H^Q_sens (Eqs. 7-8).
  • g, J (transverse field and Ising coupling) = not specified, tunable
    Define the quantum learning Hamiltonian H^Q_learn (Ising with transverse field).
  • m, ms, m_ell (effective masses) = not specified, tunable
    Mass-like inertia parameters in the induction, learning, and recursion Hamiltonians control the rate of parameter or stack-depth change.
assumptions (6)
  • standard math Hamiltonian mechanics governs the evolution of classical systems
    Used throughout Section 2 and Appendix B to define classical AGI dynamics via Poisson brackets.
  • standard math Quantum systems evolve by the Schrödinger-von Neumann equation
    Used in Section 1 and Appendix B.1 to define quantum AGI dynamics.
  • ad hoc to paper AGI cognitive functions can be represented as Hamiltonian generators
    This is the central modelling assumption introduced in the Introduction and Section 2; no justification is given beyond plausibility.
  • domain assumption Classical states are diagonal density operators in a fixed computational basis
    Introduced in the 'Classical and Quantum Information Processing' section; standard in quantum information, but it restricts classical AGI to a specific basis-dependent notion.
  • standard math Lindblad dynamics describes realistic open agents
    Used in Appendix B.2 and in the measurement model to introduce decoherence; standard but adds structure not derived from the Hamiltonians.
  • ad hoc to paper Commutation relations of generators determine the agent's capabilities
    This interpretive claim is the core of the comparison and is asserted rather than derived.
invented entities (2)
  • Generator decomposition of AGI functions (H_ind, H_reas, H_rec, H_learn, H_sens, H_env)
    purpose: To assign a Hamiltonian term to each cognitive function and compare classical/quantum agents via commutation.
    This is a conceptual entity, not a measurable physical object; no falsifiable handle is provided beyond the paper's own definitions.
  • CAGI/QAGI agent model
    purpose: To formalize agents as either classical (commuting) or quantum (non-commuting) Hamiltonian systems.
    The mapping from real AGI systems to these models is not validated against empirical data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hamiltonian Formalism for Comparing Quantum and Classical Intelligence." pith.science (2026). https://pith.science/paper/SLBKMLP6

@misc{pith2026250614456,
  author       = {Pith},
  title        = {Pith review of: Hamiltonian Formalism for Comparing Quantum and Classical Intelligence},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SLBKMLP6}},
  note         = {Machine review of arXiv:2506.14456}
}
read the original abstract

The prospect of AGI instantiated on quantum substrates motivates the development of mathematical frameworks that enable direct comparison of their operation in classical and quantum environments. To this end, we introduce a Hamiltonian formalism for describing classical and quantum AGI tasks as a means of contrasting their interaction with the environment. We propose a decomposition of AGI dynamics into Hamiltonian generators for core functions such as induction, reasoning, recursion, learning, measurement, and memory. This formalism aims to contribute to the development of a precise mathematical language for how quantum and classical agents differ via environmental interaction.

Figures

Figures reproduced from arXiv: 2506.14456 by the authors.

Figure 1
Figure 1. Classical agent (CAGI) interacting via CTC, CTQ or QTC maps with classical EC or quantum EQ environments QAGI EC EQ QTC CTQ QTQ QTQ CTQ QTC [PITH_FULL_IMAGE:figures/full_fig_p013_1.png] view at source ↗
Figure 2
Figure 2. Quantum agent (QAGI) interacting via QTC, CTQ or QTQ maps. B Hamiltonian Dynamics Classical Evolution Evolution is described by Hamilton’s equations: q˙i = ∂HC ∂pi , p˙i = − ∂HC ∂qi . (11) This can be expressed more abstractly using the Poisson bracket. For two ob￾servables f, g, their Poisson bracket is: {f, g}P B = Xn i=1  ∂f ∂qi ∂g ∂pi − ∂f ∂pi ∂g ∂qi  . (12) [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 34 canonical work pages

  1. [1]

    Cambridge University Press (2013)

    Aaronson, S.: Quantum computing since Democritus. Cambridge University Press (2013)

    work page 2013

  2. [2]

    Physics Physique Fizika1, 195–200 (1964)

    Bell, J.S.: On the Einstein Podolsky Rosen Paradox. Physics Physique Fizika1, 195–200 (1964)

    work page 1964

  3. [3]

    Cambridge University Press, Cambridge, 2nd edn

    Bell, J.: Speakable and Unspeakable in Quantum Mechanics. Cambridge University Press, Cambridge, 2nd edn. (2004)

    work page 2004

  4. [4]

    In: Artificial General In- telligence

    Bennett, M.T., Maruyama, Y.: The artificial scientist: Logicist, emergentist, and universalist approaches to artificial general intelligence. In: Artificial General In- telligence. Springer (2022)

    work page 2022

  5. [5]

    Phys- ical review letters101(6), 060401 (2008)

    Chiribella, G., D’Ariano, G.M., Perinotti, P.: Quantum circuit architecture. Phys- ical review letters101(6), 060401 (2008)

    work page 2008

  6. [6]

    Springer (2004)

    Chruściński, D., Jamiołkowski, A.: Geometric phases in classical and quantum mechanics. Springer (2004)

    work page 2004

  7. [7]

    International Journal of The- oretical Physics 21(6), 467–488 (Jun 1982)

    Feynman, R.P.: Simulating physics with computers. International Journal of The- oretical Physics 21(6), 467–488 (Jun 1982)

    work page 1982

  8. [8]

    Cambridge University Press (2011)

    Frankel, T.: The Geometry of Physics: An Introduction. Cambridge University Press (2011)

    work page 2011

Show all 37 references
  1. [9]

    Journal of Artificial General Intelligence5(1), 1 (2014)

    Goertzel, B.: Artificial general intelligence: concept, state of the art, and future prospects. Journal of Artificial General Intelligence5(1), 1 (2014)

    2014

  2. [10]

    Goertzel, B.: The general theory of general intelligence: A pragmatic patternist perspective (2021)

    2021

  3. [11]

    Goertzel, B., Bogdanov, V., Duncan, M., Duong, D., Goertzel, Z., Horlings, J., Ikle’, M., Meredith, L.G., Potapov, A., de Senna, A.L., Suarez, H.S.A., Vander- vorst, A., Werko, R.: Opencog hyperon: A framework for agi at the human level and beyond (2023)

    2023

  4. [12]

    Pearson Education (Sep 2002)

    Goldstein, H.: Classical Mechanics. Pearson Education (Sep 2002)

    2002

  5. [13]

    Springer (2013)

    Hall, B.C.: Quantum theory for mathematicians. Springer (2013)

    2013

  6. [14]

    ISSN, Elsevier Science (1979)

    Helgason, S.: Differential Geometry, Lie Groups, and Symmetric Spaces. ISSN, Elsevier Science (1979)

    1979

  7. [15]

    Springer Science & Business Media (2004)

    Hutter, M.: Universal artificial intelligence: Sequential decisions based on algorith- mic probability. Springer Science & Business Media (2004)

    2004

  8. [16]

    CRC Press (2024)

    Hutter, M., Quarel, D., Catt, E.: An Introduction to Universal Artificial Intelli- gence. CRC Press (2024)

    2024

  9. [17]

    Knapp, A.W., Knapp, A.W.: Lie groups beyond an introduction, vol. 140. Springer (1996)

    1996

  10. [18]

    Journal of Mathematics and Mechanics17(1), 59–87 (1967)

    Kochen, S., Specker, E.P.: The problem of hidden variables in quantum mechanics. Journal of Mathematics and Mechanics17(1), 59–87 (1967)

    1967

  11. [19]

    Manin, vol

    Manin, I.I.: Mathematics as metaphor: Selected essays of Yuri I. Manin, vol. 20. American Mathematical Soc. (2007)

    2007

  12. [20]

    Communications Biology 7(1), 378 (Mar 2024)

    McMillen, P., Levin, M.: Collective intelligence: A unifying concept for integrating biology across scales and substrates. Communications Biology 7(1), 378 (Mar 2024)

    2024

  13. [21]

    Cambridge University Press, 10th anniversary edn

    Nielsen, M.A., Chuang, I.L.: Quantum Computation and Quantum Information. Cambridge University Press, 10th anniversary edn. (2010)

    2010

  14. [22]

    Özkural, E.: What is it like to be a brain simulation? In: International Conference on Artificial General Intelligence. pp. 232–241. Springer (2012)

    2012

  15. [23]

    arXiv preprint arXiv:2409.04955 (2024) 12 Elija Perrier

    Perrier, E.: Quantum geometric machine learning. arXiv preprint arXiv:2409.04955 (2024) 12 Elija Perrier

    2024 arXiv

  16. [24]

    In: International Conference on Artificial General Intelligence

    Potapov, A., Rodionov, S.: Making universal induction efficient by specializa- tion. In: International Conference on Artificial General Intelligence. pp. 133–142. Springer (2014)

    2014

  17. [25]

    arXiv:2106.10522 (2021)

    Preskill, J.: Quantum computing 40 years later. arXiv:2106.10522 (2021)

    2021 arXiv

  18. [26]

    Pearson, 4th edn

    Russell, S., Norvig, P.: Artificial Intelligence: A Modern Approach. Pearson, 4th edn. (2020)

    2020

  19. [27]

    Cambridge University Press (2020)

    Sakurai, J.J., Napolitano, J.: Modern quantum mechanics. Cambridge University Press (2020)

    2020

  20. [28]

    arXiv preprint cs/0309048 (2003)

    Schmidhuber, J.: Gödel machines: Self-referential optimal universal problem solvers. arXiv preprint cs/0309048 (2003)

    2003 arXiv

  21. [29]

    part i & ii

    Solomonoff, R.J.: A formal theory of inductive inference. part i & ii. Information and Control 7(1–2), 1–22, 224–254 (1964)

    1964

  22. [30]

    Philosophical Trans- actions of the Royal Society B: Biological Sciences374(1774), 20190040 (2019)

    Solé, R., Moses, M., Forrest, S.: Liquid brains, solid brains. Philosophical Trans- actions of the Royal Society B: Biological Sciences374(1774), 20190040 (2019)

    2019

  23. [31]

    In: Artificial General Intelligence: 4th International Conference, AGI 2011, Moun- tainView,CA,USA,August3-6,2011.Proceedings4.pp.275–280.Springer(2011)

    Steunebrink, B.R., Schmidhuber, J.: A family of gödel machine implementations. In: Artificial General Intelligence: 4th International Conference, AGI 2011, Moun- tainView,CA,USA,August3-6,2011.Proceedings4.pp.275–280.Springer(2011)

    2011

  24. [32]

    In: Artificial General Intelligence: 5th In- ternational Conference, AGI 2012, Oxford, UK, December 8-11, 2012

    Sunehag, P., Hutter, M.: Optimistic aixi. In: Artificial General Intelligence: 5th In- ternational Conference, AGI 2012, Oxford, UK, December 8-11, 2012. Proceedings

    2012

  25. [33]

    pp. 312–321. Springer (2012)

    2012

  26. [34]

    In: International Conference on Artificial General Intelligence

    Veness, J., Sunehag, P., Hutter, M.: On ensemble techniques for aixi approxima- tion. In: International Conference on Artificial General Intelligence. pp. 341–351. Springer (2012)

    2012

  27. [35]

    Cambridge University Press (2018)

    Watrous, J.: The Theory of Quantum Information. Cambridge University Press (2018)

    2018

  28. [36]

    Nature299, 802–803 (Oct 1982)

    Wootters, W.K., Zurek, W.H.: A single quantum cannot be cloned. Nature299, 802–803 (Oct 1982)

    1982

  29. [37]

    Reviews of Modern Physics75(3), 715–775 (2003) A

    Zurek, W.H.: Decoherence, einselection, and the quantum origins of the classical. Reviews of Modern Physics75(3), 715–775 (2003) A. DIAGRAMS 13 Technical Appendices A Diagrams CAGI E C E Q CTC CTC CTQ QTC CTQ QTC Fig. 1. Classical agent (CAGI) interacting via CTC, CTQ or QTC m...

    2003

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.