AI in Education: Rationale, Principles, and Instructional Implications

Eyvind Elstad

REVIEW 3 major objections 4 minor 1 cited by

AI in Education: Rationale, Principles, and Instructional Implications

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper argues that using generative AI as a shortcut to answers blocks learning, and that AI's role in schools should be context-dependent, supplementing rather than replacing students' cognitive effort.

desk verdict A sensible, honest position paper whose practical advice is sound but whose empirical anchor is thinner than its strongest claims imply. read the letter →

arxiv 2412.12116 v1 pith:NP3DHQP6 submitted 2024-12-02 cs.CY cs.AI

Eyvind Elstad This is my paper · ORCID

classification cs.CYcs.AI

keywords generativeAIineducationlargelanguagemodelsdesirabledifficultiescognitiveloadlearningwithunderstandingcriticalsourcecompetenceclassroomecosystemshortcuts

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

The paper argues that generative AI tools, especially large language models like ChatGPT, differ fundamentally from search engines because they produce plausible text without checking facts or understanding meaning. The central claim is that when students copy AI-generated answers without cognitive effort, they may deliver a good product but acquire no underlying knowledge, and this is the greatest danger of AI in schools. Grounding the argument in cognitive psychology, the paper holds that durable learning requires active processing, effort, and desirable difficulties, so AI must be used to supplement rather than replace genuine mental work. A sympathetic reading is that AI's educational role should be guided by pedagogical rationale across different subjects and age levels, not by the mere availability of the technology.

What carries the argument

The central mechanism is the cognitive processing model, in which stimuli enter working memory and become durable knowledge only through effortful encoding into long-term memory, paired with the principle of desirable difficulties: tasks that make learners struggle productively improve retention. The paper applies this to the 'student + technology' unit, arguing that AI can extend thinking when the student actively processes the material, but becomes a harmful shortcut when it bypasses that processing. This machinery does the work of explaining why the same AI tool can either support or undermine learning depending on how it is used.

What would settle it

A randomized study in which students who receive direct answers from a chatbot on a new topic later perform as well or better on an unaided transfer test than students who solved the same problems without AI would contradict the paper's core claim, provided the experiment controls for prior knowledge, time on task, and tests delayed retention rather than immediate recall.

Watch

Extended reading notes

Core claim

The paper's central claim is that using generative AI as an answer shortcut can derail learning: a student might produce a well-written essay while gaining no lasting knowledge, because learning is the residue of thinking, not of retrieving a product. This claim rests on the cognitive model of attention, working memory, and long-term memory, together with the principle of desirable difficulties, which says that appropriately challenging tasks strengthen encoding and recall. The paper therefore concludes that educators should design AI use to preserve cognitive effort, for example by using chatbots that give hints rather than answers, and that the decision to use AI must be context-dependent, varying by educational stage and subject.

Load-bearing premise

The paper assumes that cognitive-psychology principles established without AI, such as desirable difficulties and working-memory limits, transfer directly to AI-mediated learning; the author acknowledges the AI-specific evidence base is still thin.

Editorial extensions

If this is right

Teachers should assign tasks that are personal, contextual, or tied to recent discussions so that AI-generated answers are less useful and students must engage with the material.
AI tools in schools should default to hint-giving and Socratic questioning rather than direct answers, as illustrated by the GPT Tutor example that shows learning benefits from partial guidance.
Assessment should include process monitoring, oral presentations, or other checks that verify whether students actually understand the work they submit, rather than only grading the final product.
Students need explicit instruction in critical source competence because LLMs can hallucinate, carry political slant, and produce superficially plausible but shallow content.
The role of AI should differ between vocational education, where it can mirror professional practice, and academic preparation programs, where the individual student's unaided competence remains central.

Reading between the lines

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

If the shortcut risk is real, then homework policies may need to treat AI use differently from in-class work, since unsupervised access to answer-generating tools could undermine practice outside school.
AI platform developers could build 'desirable difficulty' defaults that require students to attempt a problem or articulate their reasoning before revealing an answer, turning the technology into a scaffold rather than a substitute.
The argument implies an equity concern: students with strong self-regulation may benefit from AI as a tutor, while students who struggle with persistence may use it as a crutch, widening achievement gaps unless schools actively structure usage.
The 'no pain, no gain' principle, if transferred to AI, suggests that any AI feature that removes necessary struggle from a learning task should be treated as a potential risk rather than a pure efficiency gain.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

If you take one thing from this paper: it is a clear-headed, well-written position piece that tells teachers the sensible thing—AI should supplement, not replace, genuine cognitive effort—and largely avoids the hype and panic that dominate this literature. It is not a research contribution, but it does not pretend to be one.

The best parts are the synthesis and the honesty. The author weaves cognitive load theory, desirable difficulties, classroom ecology, and distributed cognition into a coherent rationale for why unreflective copying from ChatGPT is bad for learning. He repeatedly says the evidence is thin, distinguishes vocational from academic preparation, and gives concrete, usable advice: oral presentations, personalized tasks, discussing academic integrity. That is genuinely useful for practicing teachers.

The soft spots are real but not fatal. The Shein experiment—where all ChatGPT users failed a no-aid follow-up—is presented as a clean demonstration, but it is a secondhand news-style account with no sample size or methods. That is the weakest passage in the paper, and a referee should make the author either verify the underlying study or downgrade the claim. The stress-test worry about transferring desirable difficulties to LLM answer access is legitimate, but the author already hedges it with 'until it is proven otherwise,' so the central argument is more careful than the strongest rhetoric suggests. The citation glitches (Costello 2014 vs 2024, 'Shine' vs 'Shein', 'Bogust') are minor but point to sloppy proofreading.

Is anything actually new? Not really—Mollick, Hodges and Kirschner, and Bastani have all said similar things. But the paper organizes these ideas into a framework that could influence teacher practice, which is a real contribution of its kind.

Who is this for? Teachers, teacher educators, and school leaders who want a balanced, research-informed orientation to AI in classrooms. Researchers will not find new evidence or theory, but they might find the framework useful for framing studies.

My recommendation: this deserves peer review. A serious referee should push for verification of the Shein anecdote, a more explicit statement of the transfer assumption, and cleanup of the references. With those changes, it would be a respectable practitioner-oriented publication.

Referee Report

3 major / 4 minor

Summary. This paper is a conceptual position essay on the integration of generative AI in schools. The author argues that LLMs such as ChatGPT should be used in education only when they support genuine cognitive effort, not as shortcut tools that supply answers without engagement. The argument draws on cognitive psychology (working-memory limits, desirable difficulties, retrieval practice), distributed cognition, and classroom ecological theory. The paper distinguishes LLMs from search engines, emphasizes the need for critical source competence and AI-specific pedagogical judgment, and offers practical advice to teachers. The central claim is that copying AI-generated answers without cognitive engagement can block learning and that AI's role must be context-dependent, guided by the pedagogical rationale of each educational stage and subject. The paper explicitly acknowledges that the empirical evidence on AI-mediated learning is still thin and labels its main position as an argument 'until proven otherwise.'

Significance. The paper achieves a timely and coherent synthesis of cognitive-learning principles and current concerns about generative AI in schools. Its useful distinctions—between machine performance and student learning, between LLMs and search engines, and between vocational and academic-track purposes—provide a reasonable framework for teacher judgment. The paper honestly concedes the evidence base is limited, and it makes a falsifiable prediction: answer-copying without cognitive engagement impairs long-term learning. If that prediction survives empirical testing, the practical recommendations are valuable. Strengths include the explicit acknowledgment of uncertainty, the use of established cognitive psychology, and the recognition of equity and self-discipline concerns. The main weakness is that the direct AI evidence cited is partly secondhand and methodologically underreported, and the strength of the wording in several places exceeds what that evidence supports.

major comments (3)

[Motivation] The report of the Shein (2024) experiment—"all students who participated in this ChatGPT group in the first phase failed the test"—is presented as a decisive reversal, but it is a secondhand account in an ACM news-style piece with no sample size, effect size, or methodological detail. Because this result is used to support the paper's central warning about shortcut use, the author should trace the primary study and either report it with proper statistical and design details or explicitly label it as an anecdotal illustration and temper the claim accordingly. As written, the passage overstates the empirical support for a load-bearing conclusion.
[Desired difficulties in teaching to promote perseverance] The argument that easy access to LLM answers undermines learning relies on the premise that desirable-difficulty and retrieval-practice effects established in non-AI settings transfer unchanged to LLM-mediated answer access. The paper itself notes "until it is proven otherwise," but earlier statements—"This is the greatest danger of AI in schools" and "If a student uses technology to copy answers directly... derails the learning process"—are more categorical than this acknowledged uncertainty allows. The author should consistently frame the strong warning as a hypothesis grounded in cognitive theory, and explicitly discuss boundary conditions, such as whether AI-generated worked examples might benefit novices, whether high-knowledge students are less affected, and how Socratic or hint-based uses differ from raw answer access.
[Learning with and without AI – Some preliminary conclusions] The practical advice for teachers, drawn largely from Hodges and Kirschner (2024), is not tightly connected to the cognitive framework developed in the earlier sections. For example, the recommendation to shift focus from grades to process and to use oral presentations is plausible but is not derived from the desirable-difficulties or distributed-cognition mechanisms that the paper emphasizes. Making this link explicit would strengthen the paper's coherence and would better support the title's promise of 'instructional implications.'

minor comments (4)

[AI in Schools (reference list)] The citation 'Shine, 2024' in the body of the paper does not appear in the reference list; the reference list uses 'Shein, E. (2024).' Please standardize the spelling.
[Introduction] The in-text citation '(Costello et al., 2014)' is inconsistent with the reference list entry, which is dated 2024. Please correct the year.
[Abstract/Introduction] The phrase 'illuminating ideas or uncover assumptions' and the sentence 'The content of school subjects is hierarchically organized, but students read a text line by line...' contain minor grammatical and stylistic infelicities that should be polished.
[Motivation] The two sentences 'This showed that frequent use of ChatGPT correlated with a tendency to procrastinate...' and 'The authors believe that students should be encouraged...' appear to be about the Abbas et al. study, but the preceding paragraph has shifted to the Shein experiment; the transition is confusing and should be clarified.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the paper is an argumentative synthesis whose central warning rests on an acknowledged transfer assumption, not on self-referential fitting or a self-citation chain.

full rationale

This paper contains no equations, fitted parameters, or derivation chain whose output is equivalent to its input by construction. The central claim — that copying answers from generative AI can undermine learning — is presented as an inference from external cognitive-psychology literature (Bjork & Bjork, Giebl et al., Kirschner et al., Willingham) and from AI-specific studies (Bastani et al., Shein, Abbas et al.). The author explicitly acknowledges the evidentiary limits, writing that 'we are currently on uncertain ground regarding the interaction between students and AI tools' and that 'until it is proven otherwise, I argue that a student's cognitive engagement, perseverance and propensity to delve more deeply are of great significance for learning progress when using AI tools.' That is a stated assumption about transfer of desirable-difficulty effects, not a circular step: the claim is not defined in terms of itself, and no fitted quantity is renamed as a prediction. The author cites his own earlier work (Elstad 2008; Elstad 2016) for learning management and libertarian paternalism, but those concepts are secondary to the central argument about AI and cognitive effort, and the central claim does not depend on the authority of those self-citations. Because the paper is self-contained as an argumentative review and does not reduce its conclusion to its inputs, the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper's argument rests on several domain assumptions from cognitive science and AI theory. It also uses the author's own earlier framework (Elstad 2008, 2016) for learning management, but not as the central load-bearing result. No free parameters or invented entities appear.

assumptions (3)

domain assumption LLMs are stochastic parrots: they generate text by probability and do not understand meaning.
Invoked in the sections on AI in Schools and Generative AI as an information retrieval tool. This underpins the claim that AI output needs critical evaluation and cannot replace human understanding, which is load-bearing for the paper's cautionary stance.
domain assumption Durable learning requires effort, retrieval, and desirable difficulties.
Invoked in Learning mechanisms and limitations and Desired difficulties in teaching. The paper imports Bjork's desirable-difficulties framework and cognitive load theory into AI-mediated learning contexts without direct empirical confirmation for AI.
domain assumption Distributed cognition and ecological classroom theory are apt lenses for analyzing AI in classrooms.
Invoked in Generative AI in the classroom ecosystem. The paper assumes these frameworks describe how teacher, students, and AI interact, which is a modeling choice rather than an empirically established fact.

how reviews work

0 comments

Cite this review

Pith. "Pith review of AI in Education: Rationale, Principles, and Instructional Implications." pith.science (2026). https://pith.science/paper/NP3DHQP6

@misc{pith2026241212116,
  author       = {Pith},
  title        = {Pith review of: AI in Education: Rationale, Principles, and Instructional Implications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NP3DHQP6}},
  note         = {Machine review of arXiv:2412.12116}
}

read the original abstract

This study examines the integration of generative AI in schools, assessing its benefits and risks. As AI use by students grows, it's crucial to understand its impact on learning and teaching practices. Generative AI, like ChatGPT, can create human-like content, prompting questions about its educational role. The article differentiates large language models from traditional search engines and stresses the need for students to develop critical source evaluation skills. Although empirical evidence on AI's classroom effects is limited, AI offers personalized learning support and problem-solving tools, alongside challenges like undermining deep learning if misused. The study emphasizes deliberate strategies to ensure AI complements, not replaces, genuine cognitive effort. AI's educational role should be context-dependent, guided by pedagogical goals. The study concludes with practical advice for teachers on effectively utilizing AI to promote understanding and critical engagement, advocating for a balanced approach to enhance students' knowledge and skills development.

Figures

Figures reproduced from arXiv: 2412.12116 by the authors.

**Figure 1.** A model of cognitive processing (Willingham, 2021). When stimuli processed in working memory are transferred to long-term memory, the new information is incorporated into existing schemata or forms new schemata. The knowledge is stored in a hierarchical Environment Working memory Long-term memory Learning Remembering Attention [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Inclusive Education with AI: Supporting Special Needs and Tackling Language Barriers
cs.CY 2025-04 unverdicted novelty 1.0 of 10

A narrative review concludes that AI tools can support inclusive early education when implemented with ethical safeguards, but the evidence base it relies on is not rigorously appraised.

Reference graph

Works this paper leans on

9 extracted references · 6 canonical work pages · cited by 1 Pith paper

[1]

A., & Khan, T

Abbas, M., Jam, F. A., & Khan, T. I. (2024). Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students. International Journal of Educational Technology in Higher Education, 21(1),

work page 2024
[5]

https://doi.org/10.3389/frai.2024.1415782 Klingberg, T. (2023). Framtidens digitala lärande. Natur & Kultur. Klopfer, E., Reich, J., Abelson, H., & Breazeal, C. (2024). Generative AI and K-12 Education: An MIT Perspective. https://mit-genai.pubpub.org/pub/4k9msp17/release/1 Koh, J., Cowling, M., Jha, M., & Sim, K. N. (2023). The human teacher, the AI teac...

work page arXiv 2023
[7]

The Atlantic

ChatGPT is dumber than you think: Treat it like a toy, not a tool. The Atlantic. https://www.theatlantic.com/technology/archive/2022/12/chatgpt-openai-artificial- intelligence-writing-ethics/672386/. 19 Bruner, J. S. (1961). The act of discovery. Harvard Educational Review, 31, 21–32. Bruner, J. S. (1966). Toward a theory of instruction. Harvard Universit...

1961
[9]

https://proceedings.neurips.cc/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf Willingham, D. T. (2021). Why don't students like school?: A cognitive scientist answers questions about how the mind works and what it means for the classroom. John Wiley & Sons. Willingham, D.T. (2023). Outsmart Your Brain. Gallery Books. Wolfram, S. (2023). What I...

work page 2021
[10]

J., Aldrich, N

Alfieri, L., Brooks, P . J., Aldrich, N. J. & Tenenbaum, H. R. (2011). Does discovery-based instruction enhance learning? Journal of Educational Psychology, 103(1), 1–18. Anderson, J. R. (2020). Cognitive psychology and its implications. Macmillan. Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, O., & Mariman, R. (2024). Generative AI Can Harm Learn...

work page 2011
[12]

OpenAI releases o1, its first model with ‘reasoning’ abilities

September). OpenAI releases o1, its first model with ‘reasoning’ abilities. The Verge https://www.theverge.com/2024/9/12/24242439/openai-o1-model-reasoning-strawberry-chatgpt Russel, S. & Norvig, P . (2020). Artificial Intelligence: A Modern Approach (4th Edition). Pearson. Salomon, G. (1993). No distribution without individuals' cognition: a dynamic inte...

work page 2020
[30]

Let's talk evidence–The case for combining inquiry-based and direct instruction

juli). The Impact of AI on Computer Science Education. https://dl.acm.org/doi/full/10.1145/3673428 Shetye, S. (2024). An Evaluation of Khanmigo, a Generative AI Tool, as a Computer-Assisted Language Learning App. Studies in Applied Linguistics and TESOL, 24(1). Sonko, S., Adewusi, A. O., Obi, O. C., Onwusinkwue, S., & Atadoga, A. (2024). A critical review...

doi:10.1145/3673428 2024
[1087]

guide on the side

Duckworth, A. L. (2016). Grit: The Power of Passion and Perseverance. Simon & Schuster. Dweck, C. (2015). Carol Dweck revisits the growth mindset. Education week, 35(5), 20-24. Elstad, E. (2008). Building self-discipline to promote learning: students’ volitional strategies to navigate the demands of schooling. Learning Inquiry, 2, 53-71. 20 Epstein, Z., H...

arXiv 2016

Show all 9 references

[2002]

Mitchell, M

Journal of Research in Science Teaching, 47(4), 474–496. Mitchell, M. (2024). Debates on the nature of artificial general intelligence. Science, 383(6689), eado7069. Mollick, E., & Euchner, J. (2023). The Transformative Potential of Generative AI: A Conversation with Ethan Mol...

2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

A., & Khan, T

Abbas, M., Jam, F. A., & Khan, T. I. (2024). Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students. International Journal of Educational Technology in Higher Education, 21(1),

work page 2024

[2] [5]

https://doi.org/10.3389/frai.2024.1415782 Klingberg, T. (2023). Framtidens digitala lärande. Natur & Kultur. Klopfer, E., Reich, J., Abelson, H., & Breazeal, C. (2024). Generative AI and K-12 Education: An MIT Perspective. https://mit-genai.pubpub.org/pub/4k9msp17/release/1 Koh, J., Cowling, M., Jha, M., & Sim, K. N. (2023). The human teacher, the AI teac...

work page arXiv 2023

[3] [7]

The Atlantic

ChatGPT is dumber than you think: Treat it like a toy, not a tool. The Atlantic. https://www.theatlantic.com/technology/archive/2022/12/chatgpt-openai-artificial- intelligence-writing-ethics/672386/. 19 Bruner, J. S. (1961). The act of discovery. Harvard Educational Review, 31, 21–32. Bruner, J. S. (1966). Toward a theory of instruction. Harvard Universit...

1961

[4] [9]

https://proceedings.neurips.cc/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf Willingham, D. T. (2021). Why don't students like school?: A cognitive scientist answers questions about how the mind works and what it means for the classroom. John Wiley & Sons. Willingham, D.T. (2023). Outsmart Your Brain. Gallery Books. Wolfram, S. (2023). What I...

work page 2021

[5] [10]

J., Aldrich, N

Alfieri, L., Brooks, P . J., Aldrich, N. J. & Tenenbaum, H. R. (2011). Does discovery-based instruction enhance learning? Journal of Educational Psychology, 103(1), 1–18. Anderson, J. R. (2020). Cognitive psychology and its implications. Macmillan. Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, O., & Mariman, R. (2024). Generative AI Can Harm Learn...

work page 2011

[6] [12]

OpenAI releases o1, its first model with ‘reasoning’ abilities

September). OpenAI releases o1, its first model with ‘reasoning’ abilities. The Verge https://www.theverge.com/2024/9/12/24242439/openai-o1-model-reasoning-strawberry-chatgpt Russel, S. & Norvig, P . (2020). Artificial Intelligence: A Modern Approach (4th Edition). Pearson. Salomon, G. (1993). No distribution without individuals' cognition: a dynamic inte...

work page 2020

[7] [30]

Let's talk evidence–The case for combining inquiry-based and direct instruction

juli). The Impact of AI on Computer Science Education. https://dl.acm.org/doi/full/10.1145/3673428 Shetye, S. (2024). An Evaluation of Khanmigo, a Generative AI Tool, as a Computer-Assisted Language Learning App. Studies in Applied Linguistics and TESOL, 24(1). Sonko, S., Adewusi, A. O., Obi, O. C., Onwusinkwue, S., & Atadoga, A. (2024). A critical review...

doi:10.1145/3673428 2024

[8] [1087]

guide on the side

Duckworth, A. L. (2016). Grit: The Power of Passion and Perseverance. Simon & Schuster. Dweck, C. (2015). Carol Dweck revisits the growth mindset. Education week, 35(5), 20-24. Elstad, E. (2008). Building self-discipline to promote learning: students’ volitional strategies to navigate the demands of schooling. Learning Inquiry, 2, 53-71. 20 Epstein, Z., H...

arXiv 2016