Pith. sign in

REVIEW 3 major objections 5 minor 24 references

Self++: Merging Human and AI for Co-Determined XR Living in the Metaverse

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The Self++ framework proposes that Metaverse AI be designed as a nine-level progression, from basic tutoring to purpose amplification, grounded in the psychological needs of competence, autonomy, and relatedness.

desk verdict A clear, well-organized nine-level design taxonomy for human-AI XR, but the SDT grounding is an imposed structure, not a derived one. read the letter →

arxiv 2507.10967 v1 pith:MJ264NRZ submitted 2025-07-15 cs.HC

Thammathip Piumsomboon This is my paper · ORCID
classification cs.HC
keywords MetaverseSelf-DeterminationTheoryExtendedRealityHuman-AICollaborationCo-determinationCognitivebiasmitigationAutonomyDesignspace
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 position paper argues that extended reality and AI in the Metaverse should be organised around the psychological needs identified by Self-Determination Theory: competence, autonomy, and relatedness. To make that idea actionable, it proposes Self++, a nine-level framework in which AI takes a different role at each level, from Basic Tutor through Choice Architect and Social Facilitator to Purpose Amplifier. The paper is trying to establish that this structure is a usable design space for human-centred XR, not a fixed ladder, and that it can counter technological determinism by prioritising user empowerment. If the framework is right, designers would gain a concrete way to build XR systems that build skills, counter cognitive biases, and deepen social connection rather than erode them.

What carries the argument

The central object is the nine-level Self++ matrix. Each of the three SDT needs is split into three sub-levels, and each sub-level pairs a purpose (e.g., initial sense-making, shared decision-making, meaningful action) with a named AI role (Basic Tutor, Skill-Builder, Performance Coach, Choice Architect, Informed Advisor, Customisable Agent, Contextual Interpreter, Social Facilitator, Purpose Amplifier). The framework's dynamic mechanism is the shift from System 2 to System 1 processing as users advance, supported by debiasing strategies such as Cognitive Bias Modification, and by XR affordances like spatial visualisation, context-aware nudging, and avatar-mediated social presence. This matrix carries the argument by turning abstract SDT needs into checkable design choices for AI behaviour at each level.

What would settle it

A longitudinal study that follows users through the nine levels and measures the three SDT needs before and after; if higher levels do not yield greater autonomy, competence, or relatedness, or if AI assistance reduces autonomy, the proposed progression is not supported.

Watch

Extended reading notes

Core claim

The central claim is that co-determined living in the Metaverse can be operationalised as a 3x3 progression: three psychological needs (competence, autonomy, relatedness), each developed through three AI-supported levels, giving a nine-tier structure. The paper labels the stages Instruct, Empower, and Connect, and assigns each level an AI role and purpose, from Level 1 Guided Familiarisation with AI as Basic Tutor to Level 9 Purpose-Driven Engagement with AI as Purpose Amplifier. It explicitly presents the framework as a flexible design space that adapts to user expertise and context, integrating dual-process theory so that users move from deliberate System 2 thinking toward a balanced interplay with intuitive System 1 thinking. On the paper's own terms, Self++ is a guideline for designers and researchers, and its ultimate goal is to keep the Metaverse human-centred and empowering rather than automation-driven.

Load-bearing premise

The framework assumes a fixed progression from competence to autonomy to relatedness, with three levels per need, and that AI can deliver the coaching, debiasing, and social facilitation at each level without undermining user autonomy.

Editorial extensions

If this is right

  • Designers can map an XR system's AI behaviour to a specific level and purpose, making the nine levels a practical checklist for human-centred Metaverse design.
  • The framework predicts that users should move from deliberate, System 2 learning toward fluent, System 1 interaction as they gain competence, so training systems should fade support accordingly.
  • Debiasing techniques, including Cognitive Bias Modification, become a core design element rather than an afterthought in XR systems.
  • Level 6 implies that XR interfaces should let users set their own AI autonomy and delegate tasks with dashboards and real-time monitoring.
  • The framework implies that relatedness (Levels 7-9) should be designed explicitly, through consequence visualisation, social facilitation, and purpose alignment, not left to chance.

Reading between the lines

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

  • Beyond the paper's claims, the nine-level structure could be converted into an evaluation rubric: assess an XR system by which level it currently supports and whether its AI role matches the stated purpose.
  • A testable extension would compare two XR training programmes, one aligned to the Self++ sequence and one not, on measures of skill retention, self-reported SDT need satisfaction, and cognitive bias.
  • The fixed ordering from competence to autonomy to relatedness is one possible ordering; a reader could extend the framework to a lattice where users enter at different needs, and test whether sequences interact with individual differences.
  • If AI debiasing interventions are imperfect, the framework's empowerment promise could invert, so a natural next study is to measure whether nudging and coaching reduce bias or merely shift it.
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

3 major / 5 minor

Summary. This position paper introduces Self++, a nine-level framework for co-determined human-AI living in extended reality (XR) and the Metaverse. The framework organizes three stages—Instruct, Empower, and Connect—each of which is mapped onto one of Self-Determination Theory's three basic needs (competence, autonomy, relatedness). Each stage contains three levels, with a specified AI role and intended user outcome, from guided familiarization to purpose-driven engagement. The paper claims that the framework is grounded in SDT and dual-process theory, and it presents illustrative examples from prior work. It concludes with research directions, including empirical validation, cognitive bias mitigation, and ethical safeguards.

Significance. If the Self++ framework is a valid operationalization of SDT, it would provide a useful design space for XR systems that prioritize human flourishing over technological determinism. The paper clearly articulates a nine-level taxonomy with named AI roles and purposes, and it explicitly identifies important open questions, such as how to measure flourishing and how to balance user control with AI support. The lack of empirical data is acceptable for a position paper, and the author draws on relevant prior studies as illustrations. However, the central claim that the framework is grounded in SDT is load-bearing and requires careful scrutiny, as discussed in the major comments.

major comments (3)
  1. [Section 2, Table 1] The paper states that Self++ is 'grounded in Self-Determination Theory,' but SDT does not posit a fixed sequential progression from competence to autonomy to relatedness. In basic-needs theory, the three needs are universal and mutually supportive rather than ordered in a linear stage sequence. The 3-3-3 level structure, and especially the stage ordering, is an authorial scaffolding rather than an SDT-derived consequence. This is exacerbated by internal tension: Section 3 calls the framework a 'flexible design space,' yet Sections 2.2 and 2.3 build explicitly on prior stages ('With the solid foundation of user competence established'; 'Building upon established competence and autonomy'). Please either soften the SDT grounding claim to say the framework is 'inspired by' or 'maps onto' SDT with justification for the ordering, or provide a theoretical argument for why competence must precede autonomy and relatedness.
  2. [Section 2.3, Level 7] The description of Level 7, 'Contextual Awareness & Consequence Visualisation,' focuses on systems-level thinking, consequence awareness, and ripple effects across interconnected systems; there is no reference to another person, relationship, or felt interpersonal connection. Labeling this 'Basic Relatedness' is a construct mismatch with SDT's definition of relatedness as feeling connected to and cared for by others. Similarly, Level 9's emphasis on purpose, meaning, and value alignment is closer to eudaimonic meaning than to SDT relatedness. If the labels are inaccurate, the framework does not cleanly operationalize SDT, and designers may be guided by an invalid structure. Please either rename the need dimensions (e.g., 'meaningfulness' for the third stage) or substantially revise the level descriptions to genuinely address interpersonal connection.
  3. [Section 3] The paper honestly calls for 'rigorous testing' to validate the 9-level progression and measure its long-term impact. However, the manuscript provides no empirical data and no formal or theoretical derivation for why three levels per need are necessary or sufficient, nor for why the specified AI roles at each level would produce the intended psychological outcomes. To make the framework actionable and falsifiable, the paper should articulate concrete, testable hypotheses, such as predicting which level transitions are most sensitive to individual differences or how the stage ordering affects user flourishing. Without such specificity, the framework remains a descriptive taxonomy rather than a validated model.
minor comments (5)
  1. [Section 2.1, Level 2] The reference to Vygotsky's scaffolding and Zone of Proximal Development cites the original 1978 translation; citing a secondary overview or a more accessible educational psychology source would make the connection clearer.
  2. [Table 1] Several rows in Table 1 have inconsistent punctuation or capitalization; for example, the Level 4 row reads 'Basic Autonomy Enhances choice with subtle guidance,' which should have a period and consistent capitalization. Please proofread the table.
  3. [Section 2, paragraph 1] The paper states that Self++ differs from 'traditional linear models' but gives no examples of such models; a brief contrast with an existing framework (e.g., the levels of automation) would strengthen the positioning.
  4. [Section 1] The term 'co-determined' is central to the paper's title and thesis but is never explicitly defined. Please provide a working definition early in the introduction, distinguishing it from related concepts such as shared control or human-in-the-loop.
  5. [Author affiliation block] The author email appears to contain an unusual character ('envel⌢pe'), likely a typesetting artifact. Please verify the displayed email address.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Self++ is a position-paper framework whose SDT grounding and prior-work examples are not used to derive the 9-level structure by construction.

full rationale

The paper proposes a 9-level framework for co-determined XR living, grounded in Self-Determination Theory (SDT). There are no equations, fitted parameters, or quantitative predictions whose value is fixed by an earlier fitted input; the framework is presented as a design guideline rather than as an empirical derivation. The load-bearing external inputs (SDT [5], dual-process theory [7], debiasing work [11,12], nudging [13]) are independent literature, not results whose assumptions already contain the 9-level conclusion. The author's own prior publications appear only as illustrative examples mapped onto the levels (Virtual Triplets [8] for Level 1, XR-LIVE [10] for Level 2, Ex-Cit XR [14] for Level 4, and avatar/collaboration work [20-22] for Level 8); these are not invoked as evidence that the framework is correct, so the citations are not load-bearing. The paper's explicit call for 'rigorous testing' to validate the 9-level progression (Section 3) further indicates the framework is offered as a hypothesis rather than a forced conclusion. Whether SDT actually implies the fixed competence-to-autonomy-to-relatedness ordering is a substantive validity concern, not circularity: the framework could be wrong or under-justified without reducing to its own inputs. No step in the paper makes a 'prediction' equivalent by definition to a fitted parameter or to a self-citation chain.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The framework rests on two psychological theories (SDT and dual-process) and on an enabling assumption that AI can debias and coach without harming autonomy. The nine-level ordering is itself an untested design choice, not an empirical result. No free parameters or invented entities are required.

assumptions (4)
  • domain assumption Self-Determination Theory's three basic needs (autonomy, competence, relatedness) are the fundamental drivers of human flourishing.
    The entire framework is built on this psychological theory [5]; no empirical validation is provided for applying it to XR.
  • domain assumption Dual-process theory (System 1 and System 2 thinking) accurately describes human cognition in XR contexts.
    Used to justify the progression in Section 2; cited to [7] but not tested in XR.
  • ad hoc to paper AI can deliver effective debiasing and coaching that enhances rather than undermines user skills and autonomy.
    Core enabling assumption of the framework; no evidence is cited that XR-based AI can do this at scale.
  • ad hoc to paper The specific ordering of stages (competence before autonomy before relatedness) is psychologically optimal.
    Novel claim presented without justification; the paper itself calls for empirical validation of the 9-level progression.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Self++: Merging Human and AI for Co-Determined XR Living in the Metaverse." pith.science (2026). https://pith.science/paper/MJ264NRZ

@misc{pith2026250710967,
  author       = {Pith},
  title        = {Pith review of: Self++: Merging Human and AI for Co-Determined XR Living in the Metaverse},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MJ264NRZ}},
  note         = {Machine review of arXiv:2507.10967}
}
read the original abstract

This position paper introduces Self++, a novel nine-level framework for co-determined living in the Metaverse, grounded in Self-Determination Theory. Self++ prioritises human flourishing by progressively cultivating competence, autonomy, and relatedness through dynamic human-AI collaboration in extended reality (XR). Unlike technologically deterministic approaches, Self++ emphasises user empowerment by enhancing competency, mitigating cognitive biases and leveraging XR's immersive capabilities. Key research directions proposed include exploring the boundaries of user-defined AI autonomy, designing for meaningful social connection in XR, and establishing proactive ethical safeguards. Ultimately, Self++ offers a roadmap for creating a human-centred, AI-enhanced Metaverse where technology amplifies, rather than diminishes, human potential.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 20 canonical work pages

  1. [1]

    Ball, The metaverse: and how it will revolutionize everything, Liveright Publishing, 2022

    M. Ball, The metaverse: and how it will revolutionize everything, Liveright Publishing, 2022

    work page 2022

  2. [2]

    Parasuraman, T

    R. Parasuraman, T. B. Sheridan, C. D. Wickens, A model for types and levels of human interaction with automation, IEEE Transactions on systems, man, and cybernetics-Part A: Systems and Humans 30 (2000) 286–297

    work page 2000

  3. [3]

    H.-P. H. Lee, A. Sarkar, L. Tankelevitch, I. Drosos, S. Rintel, R. Banks, N. Wilson, The impact of generative ai on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers, in: To appear in the Proceedings of the 2025 CHI conference on human factors in computing systems, 2025

    work page 2025

  4. [4]

    Kruger, D

    J. Kruger, D. Dunning, Unskilled and unaware of it: how difficulties in recognizing one’s own incompetence lead to inflated self-assessments., Journal of personality and social psychology 77 (1999) 1121

    work page 1999

  5. [5]

    E. L. Deci, R. M. Ryan, Self-determination theory, Handbook of theories of social psychology 1 (2012) 416–436

    work page 2012

  6. [6]

    Z. Dong, B. Han, J. Zhang, R. Wen, B. Ens, A. Clark, T. Piumsomboon, An exploratory study on AI-driven visualisation techniques on decision making in extended reality, in: To appear in the Proceedings of the 2024 OzCHI Conference on Australian Computer-Human Interaction, ACM, 2024

    work page 2024

  7. [7]

    J. S. B. Evans, K. E. Stanovich, Dual-process theories of higher cognition: Advancing the debate, Perspectives on psychological science 8 (2013) 223–241

    work page 2013

  8. [8]

    Zhang, B

    J. Zhang, B. Han, Z. Dong, R. Wen, G. A. Lee, S. Hoermann, W. Zhang, T. Piumsomboon, Virtual triplets: a mixed modal synchronous and asynchronous collaboration with human-agent inter- action in virtual reality, in: Extended Abstracts of the CHI Conference on Human Factors in Computing Systems, 2024, pp. 1–8

    work page 2024

Show all 24 references
  1. [9]

    L. S. Vygotsky, M. Cole, Mind in society: Development of higher psychological processes, Harvard university press, 1978

    1978

  2. [10]

    Thanyadit, P

    S. Thanyadit, P. Punpongsanon, T. Piumsomboon, T.-C. Pong, Xr-live: Enhancing asynchronous shared-space demonstrations with spatial-temporal assistive toolsets for effective learning in immersive virtual laboratories, Proceedings of the ACM on Human-Computer Interaction 6 (2022) 1–23

    2022

  3. [11]

    C. K. Morewedge, H. Yoon, I. Scopelliti, C. W. Symborski, J. H. Korris, K. S. Kassam, Debiasing decisions: Improved decision making with a single training intervention, Policy Insights from the Behavioral and Brain Sciences 2 (2015) 129–140

    2015

  4. [12]

    P. T. Hertel, A. Mathews, Cognitive bias modification: Past perspectives, current findings, and future applications, Perspectives on Psychological Science 6 (2011) 521–536

    2011

  5. [13]

    R. H. Thaler, C. R. Sunstein, Nudge: Improving decisions about health, wealth, and happiness, Yale University Press, 2008

    2008

  6. [14]

    Piumsomboon, G

    T. Piumsomboon, G. Ong, C. Urban, B. Ens, J. Topliss, X. Bai, S. Hoermann, Ex-cit xr: Expert- elicitation and validation of extended reality visualisation and interaction techniques for disengag- ing and transitioning users from immersive virtual environments, Frontiers in Vir...

    2022

  7. [15]

    Han, Triadic human-agent dynamics in collaborative decision-making within extended reality, Master’s thesis, School of Product Design, 2024

    B. Han, Triadic human-agent dynamics in collaborative decision-making within extended reality, Master’s thesis, School of Product Design, 2024

    2024

  8. [16]

    an ideal human

    R. Zhang, N. J. McNeese, G. Freeman, G. Musick, "an ideal human" expectations of ai teammates in human-ai teaming, Proceedings of the ACM on Human-Computer Interaction 4 (2021) 1–25

    2021

  9. [17]

    D. B. Acharya, K. Kuppan, B. Divya, Agentic ai: Autonomous intelligence for complex goals–a comprehensive survey, IEEE Access (2025)

    2025

  10. [18]

    J. Yang, R. Tan, Q. Wu, R. Zheng, B. Peng, Y. Liang, Y. Gu, M. Cai, S. Ye, J. Jang, Y. Deng, L. Liden, J. Gao, Magma: A foundation model for multimodal ai agents, 2025. URL: https://arxiv.org/abs/ 2502.13130. arXiv:2502.13130

    2025 arXiv

  11. [19]

    R. Jain, J. Shi, A. Benton, M. Rasheed, H. Doh, S. Chidambaram, K. Ramani, Visualizing causality in mixed reality for manual task learning: A study, IEEE Transactions on Visualization and Computer Graphics (2025)

    2025

  12. [20]

    Piumsomboon, G

    T. Piumsomboon, G. A. Lee, J. D. Hart, B. Ens, R. W. Lindeman, B. H. Thomas, M. Billinghurst, Mini-me: An adaptive avatar for mixed reality remote collaboration, in: Proceedings of the 2018 CHI conference on human factors in computing systems, 2018, pp. 1–13

    2018

  13. [21]

    Piumsomboon, G

    T. Piumsomboon, G. A. Lee, A. Irlitti, B. Ens, B. H. Thomas, M. Billinghurst, On the Shoulder of the Giant: A Multi-Scale Mixed Reality Collaboration with 360 Video Sharing and Tangible Interaction, in: Proceedings of the 2019 CHI Conference on Human Factors in Computing Syste...

    2019

  14. [22]

    J. D. Hart, T. Piumsomboon, G. A. Lee, R. T. Smith, M. Billinghurst, Manipulating avatars for enhanced communication in extended reality, in: 2021 IEEE International Conference on Intelligent Reality (ICIR), IEEE, 2021, pp. 9–16

    2021

  15. [23]

    Yousefi, S

    M. Yousefi, S. E. Crowe, S. Hoermann, M. Sharifi, A. Romera, A. Shahi, T. Piumsomboon, Advancing prosociality in extended reality: systematic review of the use of embodied virtual agents to trigger prosocial behaviour in extended reality, Frontiers in Virtual Reality 5 (2024) 1386460

    2024

  16. [24]

    Slater, C

    M. Slater, C. Gonzalez-Liencres, P. Haggard, C. Vinkers, R. Gregory-Clarke, S. Jelley, Z. Watson, G. Breen, R. Schwarz, W. Steptoe, D. Szostak, S. Halan, D. Fox, J. Silver, The ethics of realism in virtual and augmented reality, Frontiers in Virtual Reality 1 (2020). URL: http...

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.