pith. sign in

arxiv: 2604.10505 · v1 · submitted 2026-04-12 · 💻 cs.AI · cs.MA

Cooperation in Human and Machine Agents: Promise Theory Considerations

M. Burgess This is my paper

Pith reviewed 2026-05-10 15:54 UTC · model grok-4.3

classification 💻 cs.AI cs.MA
keywords cooperationautonomous agentshuman-machine systemssignalingtrustriskfeedbackartificial intelligence
0
0 comments X

The pith

A promise-based perspective unifies cooperation design across human and machine agent systems.

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

The paper argues that cooperation in systems mixing humans and machines can be understood consistently by focusing on the abstract properties of autonomous agents. It revisits principles of signaling intentions, mutual comprehension, building trust, handling risk, and exchanging feedback to explain how such systems stay aligned with intended purposes. A sympathetic reader would care because this suggests practical ways to organize semi-automated teams, hardware, software, and AI without always depending on traditional management structures. The view applies the same lens to pure human efforts and fully automated setups alike.

Core claim

Promise Theory represents the fundamentals of signalling, comprehension, trust, risk, and feedback between agents, and offers some lessons about success and failure in cooperation for human, machine, and hybrid systems.

What carries the argument

Promise Theory, which models interactions between autonomous agents by tracking what each promises to do and how those promises are communicated and upheld.

If this is right

  • Cooperation principles apply uniformly to human-only teams, hardware systems, software, and artificial intelligence setups.
  • Organization and functional design of semi-automated efforts can proceed without centralized management in many cases.
  • Success or failure in agent interactions can be assessed through the same elements of signaling, trust, and feedback.
  • Lessons from promise exchanges inform design choices that reduce misalignment in mixed human-machine environments.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Design tools could be built to help humans and machines negotiate and track promises more explicitly in daily work.
  • Experiments with real teams might identify where human cognitive limits create different promise-keeping patterns than in software agents.
  • The approach could inform rules for AI systems that emphasize verifiable commitments over opaque goal optimization.

Load-bearing premise

The abstract properties of autonomous agents such as signaling, comprehension, trust, risk, and feedback transfer directly from computer systems to human efforts and mixed teams without material differences.

What would settle it

A documented case of a human-machine team where promise-based signaling and feedback fail to predict or improve cooperation outcomes compared with conventional management methods.

Figures

Figures reproduced from arXiv: 2604.10505 by M. Burgess.

Figure 1
Figure 1. Figure 1: Autonomous agents each have their own internal language. There is no authority that calibrates these to be the same without the agents’ consent. This means that, between any two agents there is a subset of language that overlaps and is comprehensible to both, but may still mean different things to each. What happens where these meet? where α = α ′ . In general, however, one language may need more words in … view at source ↗
Figure 2
Figure 2. Figure 2: Overlapping patches of language require all agents in a patch to use a compatible language, and for (at least some) agents in each patch to comprehend a faithful translation of a neighbouring language, in order to bring long range order. language, the sender can go on in the hope that eventually the right intention will be communicated. However, at some point saying too much could make things worse. Agents… view at source ↗
Figure 4
Figure 4. Figure 4: Cooperation in parallel and in series presents different challenges, each with their own uncertainty. In the serial case of figure 4, which represents a relay station or supply chain: S +bS −−→ I I −bS′ −−−→ S I +bS′′ | bS′ −−−−−−−→ R I +bS′ −−−→ R R −bS′′ −−−→ I (25) The intermediate agent becomes the chief suspect when transmission of intent from S to R fails. How can we interpret the promise body sets: … view at source ↗
Figure 3
Figure 3. Figure 3: The aggregation of promises to assemble a new (tensor) promise is dependent on the overlap between each pair of agents, which involves language uncertainty LiR and measure uncertainty b (+) i ∩b (−) Ri . These configurations each lead to their own uncertainties. Figure figure 3 combines a number of inputs into a single promise dependent on all. If the inputs are redundant sources, then the dependent promis… view at source ↗
Figure 5
Figure 5. Figure 5: A potential difference (assessed trustworthiness) acts like an incentive to begin kinetic sampling of the other party, and thus drives work. Uncertainty thus drives increased kinetic sampling, which costs work-energy like the square of the sampling rate. The function of trust is thus to reduce the overhead of managing a promise dependency. If the receiver invests only a lower rate of checking, then it effe… view at source ↗
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Assured delivery on a promise through a proxy. delivery. the client end-state E: ‘You will have package’ and this is conditional on ‘if the delivery agent promises πP ’. The Server also promises to use a promise to delivery the package from a proxy, where P is ‘Deliver package to Client’. πE : Server +E|P −−−−→ Client πP : Server −P −−→ Client    ≡ Server +E(P ) −−−−→ Client. (32) i.e. a shorthand for ‘… view at source ↗
Figure 8
Figure 8. Figure 8: figure 8 [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 8
Figure 8. Figure 8: Graphical view of a three intermediary promise chain. the promises in rectangular boxes along the chain move down the chain from server S to proxy P3 handing over the service itself. The oval promises are directed back up the chain providing delivery assurances. Finally, each agent promises independently with the end-point or client that it is an authorized agent for delivering the promise from the source … view at source ↗
Figure 9
Figure 9. Figure 9: When we deal with agents that operate on a high level, like humans in the workplace, most singular promises are made in an ambient environment that’s already filled with promises and perceived obligations (tacit impositions) as a matter of ’common knowledge’. This can make it difficult to resolve exactly what an agent should be trying to do. E. Normative teams and swarms Collaborations in which several age… view at source ↗
read the original abstract

Agent based systems are more common than we may think. A Promise Theory perspective on cooperation, in systems of human-machine agents, offers a unified perspective on organization and functional design with semi-automated efforts, in terms of the abstract properties of autonomous agents, This applies to human efforts, hardware systems, software, and artificial intelligence, with and without management. One may ask how does a reasoning system of components keep to an intended purpose? As the agent paradigm is now being revived, in connection with artificial intelligence agents, I revisit established principles of agent cooperation, as applied to humans, machines, and their mutual interactions. Promise Theory represents the fundamentals of signalling, comprehension, trust, risk, and feedback between agents, and offers some lessons about success and failure.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper claims that Promise Theory supplies a unified perspective on cooperation in systems of human and machine agents (including AI) by focusing on the abstract properties of autonomous agents—signaling, comprehension, trust, risk, and feedback. It revisits established principles of agent cooperation, applies them across human efforts, hardware, software, and artificial intelligence with or without management, and draws lessons about success and failure in maintaining intended purpose within semi-automated organizations.

Significance. If the central analogy holds, the work could provide a common conceptual language for designing mixed human-AI systems and semi-automated organizations. The manuscript correctly notes the revival of agent paradigms in connection with AI and synthesizes Promise Theory principles for this context. However, as a conceptual revisit without new empirical data, formal proofs, quantitative validation, or reproducible mappings, its significance is limited to perspective synthesis rather than advancing testable predictions or novel derivations.

major comments (2)
  1. [Abstract and main discussion of agent properties] The central claim of a unified perspective (Abstract; main body on agent properties) rests on the assumption that the abstract properties of signaling, comprehension, trust, risk, and feedback transfer directly from computational agents to human agents and mixed teams without material differences. No section supplies a concrete mapping, counter-example analysis, or additional primitives to handle human-specific factors such as subjective interpretation, power asymmetries, or cultural context that the IT-originated framework abstracts away; this makes the unification appear metaphorical and undermines the 'unified perspective' for functional design.
  2. [Section on lessons about success and failure] The lessons on success and failure (section revisiting principles of agent cooperation) are drawn directly from the author's prior Promise Theory framework without external benchmarks, independent derivations, or falsifiable predictions, reducing the contribution to re-application rather than a new unified view.
minor comments (1)
  1. [Abstract] Abstract contains a grammatical issue ('agents, This applies' uses incorrect capitalization after a comma) that should be corrected for clarity.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive review and the opportunity to respond. We address each major comment below, clarifying the manuscript's conceptual scope while noting where revisions can strengthen the presentation.

read point-by-point responses

  1. Referee: The central claim of a unified perspective (Abstract; main body on agent properties) rests on the assumption that the abstract properties of signaling, comprehension, trust, risk, and feedback transfer directly from computational agents to human agents and mixed teams without material differences. No section supplies a concrete mapping, counter-example analysis, or additional primitives to handle human-specific factors such as subjective interpretation, power asymmetries, or cultural context that the IT-originated framework abstracts away; this makes the unification appear metaphorical and undermines the 'unified perspective' for functional design.

    Authors: Promise Theory is formulated as an abstraction to capture common principles of autonomous intention and interaction across agent types. The manuscript applies these to human-machine contexts by design, without asserting identical transfer of all details. We agree that the unification would be clearer with explicit discussion of how human-specific elements (e.g., subjective interpretation) can be accommodated within promise signaling and feedback. We will revise the sections on agent properties to include brief illustrative examples and note the framework's intentional abstraction of certain contextual factors. revision: partial

  2. Referee: The lessons on success and failure (section revisiting principles of agent cooperation) are drawn directly from the author's prior Promise Theory framework without external benchmarks, independent derivations, or falsifiable predictions, reducing the contribution to re-application rather than a new unified view.

    Authors: The lessons derive from the established Promise Theory framework, as the paper explicitly revisits these principles to synthesize their application in mixed human-AI and semi-automated systems. The contribution is the extension of this lens to contemporary agent paradigms in AI, highlighting parallels for functional design that have not been systematically connected in this way. We do not introduce new benchmarks or derivations, consistent with the work's conceptual nature rather than an empirical or formal study. revision: no

standing simulated objections not resolved
  • Providing new empirical data, quantitative validation, formal proofs, or falsifiable predictions, as these would require a fundamentally different study beyond the current conceptual synthesis.

Circularity Check

1 steps flagged

Unified perspective reduces to reapplication of author's own Promise Theory framework without independent mapping or external validation

specific steps
  1. self citation load bearing [Abstract]
    "A Promise Theory perspective on cooperation, in systems of human-machine agents, offers a unified perspective on organization and functional design with semi-automated efforts, in terms of the abstract properties of autonomous agents... Promise Theory represents the fundamentals of signalling, comprehension, trust, risk, and feedback between agents, and offers some lessons about success and failure."

    The 'unified perspective' and 'lessons' are presented as derived from Promise Theory, yet the paper supplies no independent derivation, mapping, or falsifiable test; the unification is achieved simply by re-applying the same author-developed primitives to human agents, making the central claim equivalent to its own premise by construction.

full rationale

The manuscript is a conceptual revisit with no new formal results, equations, or empirical tests. Its central claim—that Promise Theory supplies a unified view of cooperation via signalling, comprehension, trust, risk, and feedback applicable to humans, hardware, software, and AI—rests entirely on re-stating the author's prior theory and assuming direct transfer of its abstract primitives. No derivation chain, counter-example analysis, or additional primitives are supplied to justify the transfer; the 'lessons about success and failure' are therefore drawn from the same self-referential source.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that Promise Theory's abstract agent properties apply uniformly across human and machine domains.

axioms (1)
  • domain assumption Promise Theory principles of signaling, comprehension, trust, risk, and feedback apply equally to human efforts, hardware, software, and AI agents.
    Invoked to justify the unified perspective on cooperation in mixed systems.

pith-pipeline@v0.9.0 · 5412 in / 1045 out tokens · 19237 ms · 2026-05-10T15:54:23.557755+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    M. Burgess. A site configuration engine.Computing systems (MIT Press: Cambridge MA), 8:309, 1995

    work page 1995

  2. [2]

    An approach to understanding policy based on autonomy and voluntary cooperation

    Mark Burgess. An approach to understanding policy based on autonomy and voluntary cooperation. InIFIP/IEEE 16th international workshop on distributed systems operations and management (DSOM), in LNCS 3775, pages 97–108, 2005

    work page 2005

  3. [3]

    Bergstra and M

    J.A. Bergstra and M. Burgess.Promise Theory: Principles and Applications (second edition).χtAxisPress, 2014,2019

    work page 2014

  4. [4]

    Marriott and M

    D. Marriott and M. Sloman. Implementation of a management agent for interpreting obligation policy.Implementation of a management agent for interpreting obligation policy, IFIP/IEEE 7th international workshop on distributed systems operations and management (DSOM), 1996

    work page 1996

  5. [5]

    Prakken and M

    H. Prakken and M. Sergot. Dyadic deontic logic and contrary-to-duty obligations. InDefeasible Deontic logic: Essays in Nonmonotonic Normative Reasoning, volume 263 ofSynthese library. Kluwer Academic Publisher, 1997

    work page 1997

  6. [6]

    Daskalopulu and M

    A. Daskalopulu and M. Sergot. The representation of legal contracts.AI & Society, 11:6–17, 1997

    work page 1997

  7. [7]

    Hellerstein, Y

    J.L. Hellerstein, Y . Diao, S. Parekh, and D.M. Tilbury.Feedback Control of Computing Systems. IEEE Press/Wiley Interscience, 2004

    work page 2004

  8. [8]

    M. Burgess. Automated system administration with feedback regulation. Software practice and experience, 28:1519, 1998

    work page 1998

  9. [9]

    Diao, J.L

    Y . Diao, J.L. Hellerstein, and S. Parekh. Optimizing quality of service using fuzzy control.IFIP/IEEE 13th International Workshop on Distributed Systems: Operations and Management (DSOM 2002), page 42, 2002

    work page 2002

  10. [10]

    M. Burgess. Configurable immunity model of evolving configuration management.Science of Computer Programming, 51:197, 2004

    work page 2004

  11. [11]

    Burgess and S

    M. Burgess and S. Fagernes. Norms and swarms.Lecture Notes on Computer Science, 4543 (Proceedings of the first International Conference on Autonomous Infrastructure and Security (AIMS)):107–118, 2007

    work page 2007

  12. [12]

    M. Burgess. Spacetimes with semantics (ii).arXiv.org:1505.01716, 2015

    work page Pith review arXiv 2015

  13. [13]

    Shannon and W

    C.E. Shannon and W. Weaver.The Mathematical Theory of Communica- tion. University of Illinois Press, Urbana, 1949

    work page 1949

  14. [14]

    Cover and J.A

    T.M. Cover and J.A. Thomas.Elements of Information Theory. (J.Wiley & Sons., New York), 1991

    work page 1991

  15. [15]

    Lewis and C

    H. Lewis and C. Papadimitriou.Elements of the Theory of Computation, Second edition. Prentice Hall, New York, 1997

    work page 1997

  16. [16]

    M. Burgess. Spacetimes with semantics (iii).arXiv:1608.02193, 2016

    work page Pith review arXiv 2016

  17. [17]

    Defining n-ary relations on the semantic web: use with individuals

    W3C. Defining n-ary relations on the semantic web: use with individuals. http://www.w3.org/TR/2004/WD-swbp-n-aryRelations-20040721/

    work page 2004

  18. [18]

    Robertson

    E.L. Robertson. Triadic relations: An algebra for the semantic web. Lecture Notes in Computer Science, 3372:91–108, 2005

    work page 2005

  19. [19]

    Rasmussen

    J. Rasmussen. Skills, rules, and knowledge; signals, signs and symbols, and other distinctions in humans performance models.IEEE Transactions on Systems, Man and Cybernetics, 13:257, 1983

    work page 1983

  20. [20]

    Management and context integration based on ontologies behind the interoperability in autonomic communications

    J.M.Serrano, J.Serrat, J.Strassner, and M.O.Foghl ´u. Management and context integration based on ontologies behind the interoperability in autonomic communications. InSIWN International Conference on Complex Open Distributed Systems (CODS 2007), volume 1, pages 435–442, 2007

    work page 2007

  21. [21]

    Burgess.A Treatise on Systems: Volume 2: Intentional systems with faults, errors, and flaws

    M. Burgess.A Treatise on Systems: Volume 2: Intentional systems with faults, errors, and flaws. in progress, 2004-

    work page 2004

  22. [22]

    M. Burgess. γ(3,4) ‘attention’ in cognitive agents: Ontology-free knowledge representations with promise theoretic semantics, 2026

    work page 2026

  23. [23]

    Kahneman.Thinking, Fast and Slow

    D. Kahneman.Thinking, Fast and Slow. Penguin, London, 2011

    work page 2011

  24. [24]

    Myerson.Game theory: Analysis of Conflict

    R.B. Myerson.Game theory: Analysis of Conflict. (Harvard University Press, Cambridge, MA), 1991

    work page 1991

  25. [25]

    Axelrod.The Complexity of Cooperation: Agent-based Models of Competition and Collaboration

    R. Axelrod.The Complexity of Cooperation: Agent-based Models of Competition and Collaboration. Princeton Studies in Complexity, Princeton, 1997

    work page 1997

  26. [26]

    Axelrod.The Evolution of Co-operation

    R. Axelrod.The Evolution of Co-operation. Penguin Books, 1990 (1984)

    work page 1990

  27. [27]

    M. Burgess. Notes on trust as a causal basis for social science. SSRN Archive, available at http://dx.doi.org/10.2139/ssrn.4252501 (DOI: 10.2139/ssrn.4252501), August 2022

    work page doi:10.2139/ssrn.4252501 2022

  28. [28]

    Burgess and R.I.M Dunbar

    M. Burgess and R.I.M Dunbar. A quantitative model of trust as a predictor of social group sizes and its implications for technology.European Economic Review, 2025

    work page 2025

  29. [29]

    Olsen.The Logic Of Collective Action

    M. Olsen.The Logic Of Collective Action. Harvard Univesity Press, 1965

    work page 1965

  30. [30]

    Dornhaus, S

    A. Dornhaus, S. Powell, and S. Bengston. Group size and its effects on collective organization.Annual review of entomology, 57:123–41, 01 2012

    work page 2012

  31. [31]

    M. Burgess. Authority (i): A promise theoretic for- malization.SSRN: https://ssrn.com/abstract=3855352, http://dx.doi.org/10.2139/ssrn.3855352, 2021

    work page doi:10.2139/ssrn.3855352 2021

  32. [32]

    Marschak and R

    J. Marschak and R. Radner.Economic Theory of Teams. Yale University Press, 1972

    work page 1972

  33. [33]

    Newman.Building Microservices

    S. Newman.Building Microservices. O’Reilly, 2015

    work page 2015

  34. [34]

    Bergstra and M

    J. Bergstra and M. Burgess.Money, Ownership, and Agency. χt-axis Press, 2019

    work page 2019

  35. [35]

    https:://www.openspaceagility.com

    OST. https:://www.openspaceagility.com

    work page

  36. [36]

    Penguin, 2010

    Jane McGonigal.Reality is Broken: Why Games Make Us Better and How They Can Change the World. Penguin, 2010

    work page 2010

  37. [37]

    Minsky.Society of Mind

    M. Minsky.Society of Mind. Simon & Schuster, 1986

    work page 1986

  38. [38]

    W. Atmar. A profoundly repeated pattern.Bulletin of the Ecological Society of America, pages 208–211, 2001

    work page 2001

  39. [39]

    Dunbar.Friends: Understanding the Power of our Most Important Relationships

    R. Dunbar.Friends: Understanding the Power of our Most Important Relationships. Little, Brown, 2022

    work page 2022

  40. [40]

    R.I.M. Dunbar. Constraints on the evolution of social institutions and their implications for information flow.J. Institutional Econ., 7:345–371, 2011

    work page 2011

  41. [41]

    Forster.The Machine Stops

    E.M. Forster.The Machine Stops. Oxford and Cambridge Review, 1909

    work page 1909

  42. [42]

    J.O. Kephart. A biologically inspired immune system for computers. Proceedings of the Fourth International Workshop on the Synthesis and Simulation of Living Systems. MIT Press. Cambridge MA., page 130, 1994

    work page 1994

  43. [43]

    M. Burgess. Computer immunology.Proceedings of the Twelth Systems Administration Conference (LISA XII) (USENIX Association: Berkeley, CA), page 283, 1998

    work page 1998

  44. [44]

    Forrest, S

    S. Forrest, S. Hofmeyr, and A. Somayaji. Computer immunology. Communications of the ACM,40:88–96, 1997

    work page 1997

  45. [45]

    Somayaji, S

    A. Somayaji, S. Hofmeyr, and S. Forrest. Principles of a computer immune system.New Security Paradigms Workshop, ACM, September:75– 82, 1997

    work page 1997

  46. [46]

    Burgess.In Search of Certainty: the science of our information infrastructure

    M. Burgess.In Search of Certainty: the science of our information infrastructure. O’Reilly Media, 2013. 19

    work page 2013