Community-Based AI Learning: Redistributing Artificial Intelligence's Epistemic Authority in Education

Kylie Peppler; Santiago Ojeda-Ramirez; Symone Gyles

arxiv: 2604.21986 · v1 · submitted 2026-04-23 · 💻 cs.HC

Community-Based AI Learning: Redistributing Artificial Intelligence's Epistemic Authority in Education

Santiago Ojeda-Ramirez , Symone Gyles , Kylie Peppler This is my paper

Pith reviewed 2026-05-09 20:42 UTC · model grok-4.3

classification 💻 cs.HC

keywords community-based AI learningepistemic authoritycritical AI literacyAI educationconstructionismsituated learningequitable educationgenerative AI

0 comments

The pith

Equitable AI education requires negotiating authority through learners' place, history, and social context.

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

The paper proposes community-based AI learning as a framework to counter the default treatment of generative AI as an authoritative knowledge source in education. It draws from community-driven and constructionist traditions to ground AI use in learners' lived experiences and local epistemologies instead. Three commitments—epistemic fine tuning, redistribution of authority, and situated discernment—work together to localize critical AI literacy, calibrate appropriate trust, highlight community knowledge, and enable collective decisions about when to engage with or reject AI. A sympathetic reader would care because AI tools now mediate much of learning yet often present universal claims that can sideline diverse ways of knowing. If the argument holds, AI education shifts from passive consumption to active negotiation of what counts as valid knowledge in specific communities.

Core claim

Community-based AI learning repositions AI's epistemic authority by anchoring engagement in learners' community-based epistemologies rather than universal claims. The framework rests on three commitments: epistemic fine tuning to adjust how AI knowledge is received locally, redistribution of authority to foreground community voices over AI defaults, and situated discernment to support collective judgment on whether to design with, interrogate, or reject AI. These processes localize critical AI literacy by calibrating trust and tying it to place, history, and social context.

What carries the argument

Community-based AI learning framework, built from the three commitments of epistemic fine tuning, redistribution of authority, and situated discernment, which together negotiate AI authority by embedding it in learners' lived and community contexts.

If this is right

AI education programs must incorporate learners' specific place, history, and social context to redistribute epistemic authority away from the AI system.
Critical AI literacy becomes tied to local community knowledge rather than abstract or universal principles.
Learners gain the capacity for collective judgment to decide when to design with AI, question it, or set it aside.
Trust in AI is calibrated according to alignment with community epistemologies instead of default acceptance.
Equitable AI education emerges from explicit negotiation of authority rather than imposition of standardized AI tools.

Where Pith is reading between the lines

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

This framework could be tested by comparing trust calibration and community knowledge integration in schools that adopt the three commitments versus those that use standard AI literacy curricula.
The approach may connect to efforts in other fields, such as data ethics or algorithmic governance, where authority is similarly redistributed to affected communities.
Without additional implementation details, the commitments risk remaining high-level and may need concrete tools for educators to apply them consistently across settings.
The argument implies that failing to negotiate authority this way could deepen existing educational inequities by reinforcing AI as a neutral arbiter.

Load-bearing premise

The three commitments will localize critical AI literacy and calibrate trust without needing extra mechanisms or direct evidence of how they interact with existing AI systems.

What would settle it

Classroom observations or trials in which students continue to accept AI outputs as authoritative regardless of community context or the three commitments would falsify the central claim.

Figures

Figures reproduced from arXiv: 2604.21986 by Kylie Peppler, Santiago Ojeda-Ramirez, Symone Gyles.

**Figure 1.** Figure 1: Community-Based AI Learning: From AI-Centered Authority to Collective Discernment. 5.1 Epistemic Fine Tuning In technical ML usage, fine tuning refers to updating a pretrained model’s parameters with additional task or domain data so the model performs better for a specific purpose. Contemporary Natural Language Processing and Large Language Models work has popularized this framing through transfer learnin… view at source ↗

read the original abstract

As generative AI systems increasingly mediate learning, they are often treated as authoritative sources of knowledge. This perspective paper introduces community-based AI learning as a framework that repositions authority, grounding AI engagement in learners' lived and community-based epistemologies. Drawing from community-driven learning and constructionist traditions, we articulate three commitments: epistemic fine tuning, redistribution of authority, and situated discernment. Together, these processes localize critical AI literacy by calibrating trust, foregrounding community knowledge, and supporting collective judgment about when to design with, interrogate, or reject AI. We argue that equitable AI education requires negotiating authority through place, history, and social context.

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.

Desk Editor's Note private letter to a colleague

A conceptual framework for community-based AI learning that synthesizes prior ideas but stays untested at a high level.

read the letter

This paper proposes community-based AI learning as a way to reposition AI in education so that authority comes from learners' communities rather than treating generative systems as default knowledge sources. The main pitch is that equitable AI education requires negotiating that authority through place, history, and social context via three commitments: epistemic fine tuning, redistribution of authority, and situated discernment. It draws explicitly from community-driven learning and constructionist traditions to localize critical AI literacy and support decisions about when to use, question, or reject AI tools. That synthesis is the clearest part of the work and gives the commitments a coherent shape in the abstract. The argument lands as a normative stance rather than a tested claim. The soft spots are straightforward: the piece offers no implementation details, classroom examples, or data showing these commitments actually redistribute authority or improve outcomes. Without those, it is hard to judge whether the processes would interact productively with real AI systems or handle conflicts inside communities. The value placed on community epistemologies is built into the framing, which makes the structure self-referential in the way the reader noted, though that is common in perspective papers and not a hidden flaw here. This is for people already working on critical AI literacy, equity in ed-tech, or constructionist approaches who want language to think about power in AI-mediated learning. It could prompt useful discussion but will not give practitioners or empiricists concrete methods or results to build on. I would send it to peer review. The engagement with the literature is honest and the framing is clear enough that referees could push for the missing grounding or examples without starting from scratch.

Referee Report

1 major / 2 minor

Summary. The paper claims that as generative AI systems increasingly mediate learning and are treated as authoritative sources, equitable AI education requires a 'community-based AI learning' framework that repositions authority by grounding engagement in learners' lived and community-based epistemologies. Drawing from community-driven learning and constructionist traditions, it articulates three commitments—epistemic fine tuning, redistribution of authority, and situated discernment—that together localize critical AI literacy by calibrating trust, foregrounding community knowledge, and supporting collective judgment on when to design with, interrogate, or reject AI. The central argument is that authority must be negotiated through place, history, and social context.

Significance. If the framework is adopted, it could meaningfully influence HCI and education technology by shifting focus from AI as default authority to community-grounded critical engagement, potentially improving equity in AI-mediated learning environments. The paper's explicit synthesis of established traditions into a normative framework is a strength, providing a coherent perspective that may stimulate discussion and guide tool design or curricula. As a perspective piece, its value is in articulation rather than demonstration; the internal definitions of commitments do not introduce problematic circularity for a normative argument, and the stress-test concern about self-reference does not undermine the central claim.

major comments (1)

[Articulation of the three commitments] In the articulation of the three commitments (as presented following the abstract), the paper asserts that epistemic fine tuning, redistribution of authority, and situated discernment 'together' localize critical AI literacy and calibrate trust, but provides no mechanisms for their interaction, potential conflicts, or integration with existing generative AI systems; this is load-bearing for the claim that the framework supports collective judgment about AI use.

minor comments (2)

The paper would benefit from one or two concrete hypothetical scenarios illustrating how a community might apply situated discernment to reject or adapt an AI output in a specific educational context.
Explicitly distinguish the novel synthesis from prior work in constructionism and community-based learning to clarify the framework's incremental contribution.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive and positive review, which recognizes the paper's value as a perspective piece in synthesizing traditions to influence HCI and education technology. We address the major comment below.

read point-by-point responses

Referee: In the articulation of the three commitments (as presented following the abstract), the paper asserts that epistemic fine tuning, redistribution of authority, and situated discernment 'together' localize critical AI literacy and calibrate trust, but provides no mechanisms for their interaction, potential conflicts, or integration with existing generative AI systems; this is load-bearing for the claim that the framework supports collective judgment about AI use.

Authors: We agree that the manuscript would benefit from greater explicitness on how the commitments interact, including potential tensions and points of integration with generative AI. As a normative perspective piece, the framework intentionally avoids prescribing universal mechanisms, which must remain context-specific. However, to strengthen the load-bearing claim, we will revise by adding a dedicated subsection that outlines the synergistic dynamics (e.g., epistemic fine tuning surfacing community-specific biases that then guide authority redistribution) and illustrative conflict-resolution pathways (e.g., situated discernment determining when to reject AI outputs). We will also include brief examples of integration with current generative systems, such as using the commitments to inform prompt design or community review protocols. This revision preserves the paper's perspective orientation while making the collective judgment process more actionable. revision: yes

Circularity Check

0 steps flagged

No significant circularity in normative framework

full rationale

The paper is a perspective piece that articulates a normative framework drawn from community-driven learning and constructionist traditions rather than presenting a derivation chain, equations, fitted parameters, or empirical predictions. The three commitments (epistemic fine tuning, redistribution of authority, and situated discernment) are introduced as guiding principles to localize critical AI literacy, and the central claim—that equitable AI education requires negotiating authority through place, history, and social context—is advanced as an argument, not a result that reduces to its own inputs by construction. No self-definitional loops, fitted inputs renamed as predictions, load-bearing self-citations, uniqueness theorems, or smuggled ansatzes are present in the text; the framework presupposes the value of community epistemologies as an external starting point from established traditions, not as a self-referential output.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The claim rests on the assumption that community epistemologies are valid and superior grounds for AI engagement, drawn from existing traditions without new supporting evidence or mechanisms.

axioms (1)

domain assumption Community-driven learning and constructionist traditions provide appropriate foundations for addressing AI epistemic authority in education.
Invoked to articulate the three commitments and ground the framework.

invented entities (1)

Community-based AI learning framework no independent evidence
purpose: To reposition AI's epistemic authority by grounding it in community epistemologies.
Newly introduced conceptual structure without independent external validation or falsifiable predictions.

pith-pipeline@v0.9.0 · 5406 in / 1314 out tokens · 59278 ms · 2026-05-09T20:42:51.676739+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

3 extracted references · 3 canonical work pages

[1]

Ballard, H.L. et al. 2023. Community-driven science and science education: Living in and navigating the edges of equity, justice, and science learning. (2023). https://doi.org/10.1002/tea.21880 [4] Bender, E.M. et al. 2021. On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?. Proceedings of the 2021 ACM Conference on Fairness, Accountabi...

work page doi:10.1002/tea.21880 2023
[2]

Feeling That You’re Not Alone

Morales-Navarro, L. et al. 2024. Youth as Peer Auditors: Engaging Teenagers with Algorithm Auditing of Machine Learning Applications. Proceedings of the 23rd Annual ACM Interaction Design and Children Conference (New York, NY, USA, Jun. 2024), 560–573. [29] Moudgalya, S.K. and Swaminathan, S. 2024. Toward Data Sovereignty: Justice-oriented and Community-b...

work page doi:10.1080/17439884.2025.2452199 2024
[3]

Veldhuis, A. et al. 2025. Critical Artificial Intelligence literacy: A scoping review and framework synthesis. International Journal of Child-Computer Interaction. 43, (2025), 100708. [51] Warren, B. et al. 2020. Re-Imagining Disciplinary Learning. Handbook of the cultural foundations of learning. (2020), 277. [52] Wei, X. et al. 2025. Addressing Bias in ...

work page doi:10.1016/j.im.2025.104103 2025

[1] [1]

Ballard, H.L. et al. 2023. Community-driven science and science education: Living in and navigating the edges of equity, justice, and science learning. (2023). https://doi.org/10.1002/tea.21880 [4] Bender, E.M. et al. 2021. On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?. Proceedings of the 2021 ACM Conference on Fairness, Accountabi...

work page doi:10.1002/tea.21880 2023

[2] [2]

Feeling That You’re Not Alone

Morales-Navarro, L. et al. 2024. Youth as Peer Auditors: Engaging Teenagers with Algorithm Auditing of Machine Learning Applications. Proceedings of the 23rd Annual ACM Interaction Design and Children Conference (New York, NY, USA, Jun. 2024), 560–573. [29] Moudgalya, S.K. and Swaminathan, S. 2024. Toward Data Sovereignty: Justice-oriented and Community-b...

work page doi:10.1080/17439884.2025.2452199 2024

[3] [3]

Veldhuis, A. et al. 2025. Critical Artificial Intelligence literacy: A scoping review and framework synthesis. International Journal of Child-Computer Interaction. 43, (2025), 100708. [51] Warren, B. et al. 2020. Re-Imagining Disciplinary Learning. Handbook of the cultural foundations of learning. (2020), 277. [52] Wei, X. et al. 2025. Addressing Bias in ...

work page doi:10.1016/j.im.2025.104103 2025