Pith. sign in

REVIEW 2 major objections 6 minor 73 references

Why (not) use AI? Analyzing People's Reasoning and Conditions for AI Acceptability

T0 review · 2 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read The paper claims that public acceptability of AI use cases is systematically tied to whether people reason in cost-benefit or rule-based terms.

desk verdict A useful, transparent survey of how people justify AI acceptability, but the central rule-based-to-rejection link reads as post-hoc justification rather than demonstrated decision process. read the letter →

arxiv 2502.07287 v2 pith:4TMAG4FC submitted 2025-02-11 cs.CY

Jimin Mun , Wei Bin Au Yeong , Wesley Hanwen Deng , Jana Schaich Borg , Maarten Sap This is my paper · ORCID
classification cs.CY
keywords AIacceptabilitypublicattitudesmoralreasoningcost-benefitrule-basedusecasesgovernancesurveystudy
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 asks why laypeople accept or reject specific AI use cases, and argues that the style of reasoning in people's justifications is a key part of the answer. Based on a survey of 197 demographically diverse participants judging ten AI use cases, it finds that outcome-focused (cost-benefit) rationales accompany approval while principle-focused (rule-based) rationales accompany disapproval, especially for contested uses like an AI elementary-school teacher. It also reports lower acceptance for labor-replacement than personal-health applications, lower acceptance among non-male participants and people familiar with AI ethics, and higher agreement among participants who used the same reasoning style. If the paper is right, public acceptance of AI is not settled by better impact reports alone; consensus-building must engage the moral frameworks people bring to each use case.

What carries the argument

The load-bearing instrument is a structured survey that combines binary acceptability judgments (should this be developed, would you use it), 5-point confidence ratings, and open-text rationales completed by sentence stems ('should be developed because...' and 'should not be developed if...'). The rationales are classified along three dimensions — reasoning type (cost-benefit vs rule-based vs both vs unclear), moral foundations (Care, Fairness, Purity, Authority, Loyalty), and switching conditions (Functionality, Usage, Societal Impact) — with a large-language-model classifier validated against a human-annotated gold standard using an inter-rater agreement statistic (Gwet's AC1). Linear mixed-effects models then relate these coded rationales to judgments and confidence, with random effects for participants and use cases. The reasoning-type classification carries the argument: it is the coded variable that links the open-text justifications to the acceptability outcomes and to the disagreement pattern across use cases.

What would settle it

Randomly assign participants to write either cost-benefit or rule-based justifications for the same set of AI use cases before stating acceptability, and compare acceptability between the two groups. If the induced reasoning style does not shift acceptability while the vignettes are identical, then the paper's observed correlation between reasoning type and judgment is not a causal relationship. A lighter check consistent with the paper's own second study is that explicitly listing and weighing harms and benefits barely moved judgments, which a strongly causal reading of reasoning-to-acceptance would need to explain.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that acceptability judgments of AI use cases covary with the moral reasoning style in people's written justifications. In linear mixed-effects models on development judgments, rule-based reasoning is negatively associated with acceptability ($\beta_{\mathrm{DEV}} = -0.46$, $p < .001$) while cost-benefit reasoning is positively associated ($\beta_{\mathrm{DEV}} = 0.32$, $p < .01$); for usage judgments, cost-benefit reasoning is negatively associated with raw acceptance but positively with confidence-weighted acceptance. The use cases with the least disagreement (IT Support Specialist, Nutrition Optimizer) had the highest shares of cost-benefit rationales, and the most contested uses (Elementary School Teacher, Lawyer) had the most rule-based rationales. The paper reads this as evidence that public disagreements about AI development are often disagreements between utilitarian, outcome-weighing valuations and deontological, rule-based commitments about what AI should not do.

Load-bearing premise

The load-bearing premise is that the open-text rationales participants wrote are a faithful trace of the thinking that produced their judgments; if those rationales are post-hoc justifications generated after the decision, then the associations between reasoning style and acceptability would not describe how people actually decide.

Editorial extensions

If this is right

  • If rule-based reasoning is what pushes acceptability down, then impact assessments and cost-benefit reports alone will not resolve public opposition; deliberation must engage deontological concerns directly.
  • Because non-male participants, participants with high discrimination chronicity, and participants familiar with AI ethics gave lower acceptability ratings, workplace AI governance should include these stakeholders rather than relying on majority or expert opinion.
  • The sharply lower acceptance of Elementary School Teacher AI relative to IT Support Specialist AI shows that labor-replacement acceptability is not uniform, with care work a particular flashpoint.
  • Use cases with more unified reasoning styles showed less disagreement, implying that consensus interventions that align reasoning frames — not just share facts — may reduce polarization about AI.

Reading between the lines

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

  • A testable extension the paper does not run: randomly prompt participants to justify one vignette in cost-benefit terms and another in rule-based terms; if acceptability does not move with the induced framing, the observed correlation is not a causal path.
  • The paper's second study found that explicitly weighing harms and benefits barely changed judgments; that hints acceptability may be formed before deliberation, so future work could measure decision time or pre-rationale predictions to separate decision from justification.
  • The domain pattern suggests a moral-foundation map of occupations: fairness-centric roles (lawyer, eligibility interviewer) and care-centric roles (teacher) will attract the most rule-based opposition, a prediction that could be tested on new occupational vignettes.
  • Because moral foundations vary across cultures, the US-sample finding that rule-based reasoning marks rejection may not transfer; a cross-cultural replication would clarify whether the reasoning-style signature is universal or local.
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

2 major / 6 minor

Summary. The paper reports a survey study (N=197) in which participants judged ten AI use cases (five labor-replacement, five personal health) on development and usage acceptability, provided open-text rationales, and answered demographic and AI-literacy questions. The authors annotate rationales for cost-benefit versus rule-based reasoning, moral foundations, and switching conditions, then fit mixed-effects models to relate judgments to use-case characteristics, demographics, and rationale features. They find higher acceptability for personal health than for labor-replacement use cases, significant demographic associations, and a negative association between rule-based reasoning and development acceptability (βDEV = −0.46, p < .001, Table 5), interpreting this as evidence that reasoning style drives acceptability and that unified reasoning styles lead to higher consensus.

Significance. This is a relevant and well-motivated contribution to the AI acceptability literature. Its strengths include the open-ended elicitation of lay rationales, the demographically stratified sample, the transparent validation of LLM-based annotation against human annotations using Gwet's AC1 (Appendix C), and the inclusion of a second study in Appendix F that attempts to manipulate deliberation. If the reasoning-type associations were causal, the paper would offer actionable guidance for consensus-building. However, the causal interpretation is not currently supported by the design, and the agreement/reasoning-type link is not formally tested. The paper is therefore a promising empirical resource whose central claims need strengthening before publication.

major comments (2)
  1. [§3.2, §5.3, Table 5] The central association between reasoning type and acceptability (Table 5: βDEV = −0.46, p < .001) is vulnerable to a temporal-ordering problem. In the survey, Q1 (development judgment) is asked immediately before Q3, which asks participants to complete '[Use Case] should [not] be developed because...' (Section 3.2). The open-text rationales are therefore produced after the judgment and are explicitly framed as justifications for an already-made decision. Annotating these texts as 'rule-based reasoning' and then using that code as a predictor of the earlier judgment (Section 5.3, Table 5) conflates the language people use to rationalize a decision with the reasoning that produced it. The authors even report in Appendix F that explicitly weighing harms and benefits (which forces cost-benefit deliberation) had at most one marginal effect (F(1,201.05)=3.371, p=0.0678), which is consistent with the rationalization account. To make a causal claim about reasoning style, the manuscript needs either a manipulation that varies reasoning style independently of judgment, a process measure (e.g., response time or deliberation tasks), or at minimum a careful revision that reframes Table 5 as describing properties of justifications rather than as drivers of acceptability.
  2. [§5.3, §6, Figure 4] The conclusion that 'use cases with less disagreement tended to elicit more cost-benefit reasoning' (Section 6) is based on an informal visual comparison across ten use cases (Figure 4) and on the observation that IT Support Specialist AI (91.0% cost-benefit) and Nutrition Optimizer (92.8%) have low disagreement, while Elementary School Teacher AI (30% rule-based) has high disagreement. No formal test is reported for the relationship between agreement (e.g., the standard deviation of judgment×confidence) and the proportion of reasoning types. A simple correlation or a use-case-level regression across the ten use cases would directly test this claim; as written, the claim is an eyeballed pattern and should either be formally supported or explicitly labeled as observational.
minor comments (6)
  1. [§5.1, Figure 7] The t-test comparing categories treats the 985 judgments as independent, although each participant provides five judgments. Because category assignment is between participants but judgments are repeated within participants, the test's degrees of freedom are inflated. The mixed-model ANOVA in Table 24 supports the same conclusion, so the error does not change the substantive result, but the figure and text should use a valid repeated-measures test or participant-level aggregation.
  2. [Appendix A.4] The description of switching-condition categories contains a copy-paste error: 'usage (e.g., errors, bias in systems, limited capabilities)' repeats the examples given for functionality. The usage dimension is described correctly in Section 4.2 as context of system integration (e.g., supervision, misuse, or unintended use).
  3. [Table 5, §5.3] The coding of 'both' cost-benefit and rule-based reasoning is not defined in the table or text. Because the two indicators are binary and mutually non-exclusive, the intercept and the coefficients are interpretable only if the reference category ('unclear' in the annotation scheme) is explicitly stated; please clarify.
  4. [Table 3, §5.2] The table reports many interaction coefficients (e.g., AsianLabor, Non-maleLabor, Str. LiberalLabor) without any multiple-comparison correction. Given the large number of tests, a footnote acknowledging this and distinguishing hypothesis-driven main effects from exploratory interactions would help.
  5. [Abstract and §6] The phrasing 'unified reasoning type (e.g., cost-benefit reasoning) leading to higher agreement' overstates the evidence; as noted in the major comments, the relationship is not formally tested. Suggest revising to 'associated with' or adding the missing statistical test.
  6. [Figure 4] The percentages for the 'Both' and 'Unclear' segments are not labeled in the figure, making the full distribution difficult to read. Adding exact percentage labels or a detailed legend would improve transparency.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the central associations are empirical, though reasoning measures are self-reported post-decision rationales.

full rationale

The paper's central claims are empirical survey associations rather than derivations from assumed inputs. The key coefficients in Table 5 regress acceptability judgments on reasoning types annotated from Q3/Q4 open-text rationales; the reasoning labels are defined by the content of the justification (cost-benefit vs. rule-based), not by the polarity of the judgment, so the negative rule-based association is not forced by construction. Prior author work (Mun et al. 2024) is used mainly to source use cases and annotation dimensions, while the acceptability judgments and rationale texts are new data, so the conclusions do not reduce to that prior work. The concern that Q3 rationales are elicited after Q1 and may be post-hoc justifications is a construct-validity and causal-inference limitation, not a definitional circularity; Appendix F's near-null effect of explicit harm/benefit weighing is consistent with that alternative but does not make the observed correlations tautological. No parameter is fitted to a subset and then renamed as a prediction, no uniqueness theorem is imported from the authors' prior work, and no ansatz is smuggled in via self-citation. Accordingly, the paper is substantively self-contained against its empirical inputs, with only a minor reliance on the authors' prior taxonomies that is not load-bearing for the main findings.

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

No invented entities or ad hoc parameters. The central claims rest on the validity of self-reported reasoning, the chosen moral-psychology frameworks, the vignette design, and standard statistical assumptions.

assumptions (4)
  • domain assumption Participants' self-reported rationales accurately reflect their underlying reasoning processes.
    The central RQ3 analysis depends on open-text rationales (Q3/Q4) as valid measures of reasoning type, despite potential post-hoc rationalization. Invoked in Section 3.2.
  • domain assumption The dual-process distinction (cost-benefit vs rule-based reasoning) is a valid and applicable framework for AI acceptability judgments.
    The paper adopts utilitarian/deontological categories from moral psychology (Section 4.2) without validating their fit to AI-specific decisions.
  • domain assumption The ten vignettes are representative samples of their categories and vary primarily on the intended factors (education and risk).
    RQ1 interprets acceptability differences across use cases as reflecting category and risk/education factors; confounds with specific job or content attributes (e.g., teaching children) are possible.
  • standard math Statistical models' assumptions (linearity, normality of residuals, random effects structure) hold.
    Repeated-measures ANOVA and lmer models are used throughout; no diagnostics are reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Why (not) use AI? Analyzing People's Reasoning and Conditions for AI Acceptability." pith.science (2026). https://pith.science/paper/4TMAG4FC

@misc{pith2026250207287,
  author       = {Pith},
  title        = {Pith review of: Why (not) use AI? Analyzing People's Reasoning and Conditions for AI Acceptability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4TMAG4FC}},
  note         = {Machine review of arXiv:2502.07287}
}
read the original abstract

In recent years, there has been a growing recognition of the need to incorporate lay-people's input into the governance and acceptability assessment of AI usage. However, how and why people judge acceptability of different AI use cases remains under-explored, despite it being crucial towards understanding and addressing potential sources of disagreement. In this work, we investigate the demographic and reasoning factors that influence people's judgments about AI's development via a survey administered to demographically diverse participants (N=197). As a way to probe into these decision factors as well as inherent variations of perceptions across use cases, we consider ten distinct labor-replacement (e.g., Lawyer AI) and personal health (e.g., Digital Medical Advice AI) AI use cases. We explore the relationships between participants' judgments and their rationales such as reasoning approaches (cost-benefit reasoning vs. rule-based). Our empirical findings reveal a number of factors that influence acceptance. We find lower acceptance of labor-replacement usage over personal health, significant influence of demographics factors such as gender, employment, education, and AI literacy level, and prevalence of rule-based reasoning for unacceptable use cases. Moreover, we observe unified reasoning type (e.g., cost-benefit reasoning) leading to higher agreement. Based on these findings, we discuss the key implications towards understanding and mitigating disagreements on the acceptability of AI use cases to collaboratively build consensus.

Figures

Figures reproduced from arXiv: 2502.07287 by the authors.

Figure 1
Figure 1. Five professional or personal use cases are presented in a random order. For each use case, we ask multiple-choice [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Labor-replacement use case means and distributions of numerically converted Judgment, Confidence, and [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Personal health use case means and distributions of numerically converted Judgment, Confidence, and [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Percentage of reasoning types (cost-benefit and rule-based) by use cases in participant provided rationales (Q3, Q4). [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Proportion (%) of presence of moral foundations in participant’s rationale responses (Q3, Q4) aggregated by use case. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Proportion (%) of presence of switching conditions (Functionality, Usage, Societal Impact) mentioned in participant’s [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Use case category means and distributions of numerically converted Judgment, Confidence, and [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Numerically converted Judgment x Confidence (- [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Mapping of decisions and reasoning types. + and - denote positive and negative acceptance. “C” denotes Cost-benefit [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

73 extracted references · 63 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence

    2023. Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/. Accessed: 2024-01-02

    work page 2023

  4. [4]

    How Do People Feel About AI? A Nationally Representative Survey of Public Attitudes to Artificial Intelligence in Britain

    2023. How Do People Feel About AI? A Nationally Representative Survey of Public Attitudes to Artificial Intelligence in Britain

    work page 2023

  5. [5]

    Aging-and-Tech Job Vulnerability

    Alcover, C.-M.; Guglielmi, D.; Depolo, M.; and Mazzetti, G. 2021. “Aging-and-Tech Job Vulnerability”: A proposed framework on the dual impact of aging and AI, robotics, and automation among older workers. Organizational Psychology Review, 11(2): 175--201

    work page 2021

  6. [6]

    T.; and Dehghani, M

    Atari, M.; Haidt, J.; Graham, J.; Koleva, S.; Stevens, S. T.; and Dehghani, M. 2023. Morality beyond the WEIRD: How the nomological network of morality varies across cultures. Journal of Personality and Social Psychology

    work page 2023

  7. [7]

    Awad, E.; Dsouza, S.; Kim, R.; Schulz, J.; Henrich, J.; Shariff, A.; Bonnefon, J.-F.; and Rahwan, I. 2018. The moral machine experiment. Nature, 563(7729): 59--64

    work page 2018

  8. [8]

    R.; Vaughan, J

    Barocas, S.; Guo, A.; Kamar, E.; Krones, J.; Morris, M. R.; Vaughan, J. W.; Wadsworth, W. D.; and Wallach, H. 2021. Designing Disaggregated Evaluations of AI Systems: Choices, Considerations, and Tradeoffs. In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, AIES '21, 368–378. New York, NY, USA: Association for Computing Machinery. ...

    work page 2021

Show all 73 references
  1. [9]

    Barocas, S.; and Selbst, A. D. 2016. Big data's disparate impact. Calif. L. Rev., 104: 671

    2016

  2. [10]

    S.; Levi, M.; Magnus, D.; Rajala, B

    Bernstein, M. S.; Levi, M.; Magnus, D.; Rajala, B. A.; Satz, D.; and Waeiss, Q. 2021. Ethics and society review: Ethics reflection as a precondition to research funding. Proceedings of the National Academy of Sciences, 118(52): e2117261118

    2021

  3. [11]

    Bhardwaj, E.; Gujral, H.; Wu, S.; Zogheib, C.; Maharaj, T.; and Becker, C. 2024. Machine learning data practices through a data curation lens: An evaluation framework. In The 2024 ACM Conference on Fairness, Accountability, and Transparency, 1055--1067

    2024

  4. [12]

    Birhane, A.; Kalluri, P.; Card, D.; Agnew, W.; Dotan, R.; and Bao, M. 2022. The values encoded in machine learning research. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 173--184

    2022

  5. [13]

    S.; and Read, H

    Borg, J. S.; and Read, H. 2024. What Is Required for Empathic AI? It Depends, and Why That Matters for AI Developers and Users. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, volume 7, 1306--1318

    2024

  6. [14]

    M.; Jakesch, M.; Ribeiro, M

    Bu c inca, Z.; Pham, C. M.; Jakesch, M.; Ribeiro, M. T.; Olteanu, A.; and Amershi, S. 2023. AHA!: Facilitating AI Impact Assessment by Generating Examples of Harms. arXiv preprint arXiv:2306.03280

    2023 arXiv

  7. [15]

    J.; Wang, J

    Chen, R. J.; Wang, J. J.; Williamson, D. F.; Chen, T. Y.; Lipkova, J.; Lu, M. Y.; Sahai, S.; and Mahmood, F. 2023. Algorithmic fairness in artificial intelligence for medicine and healthcare. Nature biomedical engineering, 7(6): 719--742

    2023

  8. [16]

    M.; and Zhu, H

    Cheng, H.-F.; Wang, R.; Zhang, Z.; O'connell, F.; Gray, T.; Harper, F. M.; and Zhu, H. 2019. Explaining decision-making algorithms through UI: Strategies to help non-expert stakeholders. In Proceedings of the 2019 chi conference on human factors in computing systems, 1--12

    2019

  9. [17]

    Cheon, E. 2023. Powerful Futures: How a Big Tech Company Envisions Humans and Technologies in the Workplace of the Future. Proc. ACM Hum.-Comput. Interact., 7(CSCW2)

    2023

  10. [18]

    Cheung, V.; Maier, M.; and Lieder, F. 2024. Measuring the decision process in (moral) dilemmas: Self-report measures of reliance on rules, cost-benefit reasoning, intuition, & deliberation

    2024

  11. [19]

    Cushman, F. 2013. Action, outcome, and value: A dual-system framework for morality. Personality and social psychology review, 17(3): 273--292

    2013

  12. [20]

    H.; Barocas, S.; and Vaughan, J

    Deng, W. H.; Barocas, S.; and Vaughan, J. W. 2024. Supporting Industry Computing Researchers in Assessing, Articulating, and Addressing the Potential Negative Societal Impact of Their Work. arXiv preprint arXiv:2408.01057

    2024 arXiv

  13. [21]

    H.; Guo, B.; Devrio, A.; Shen, H.; Eslami, M.; and Holstein, K

    Deng, W. H.; Guo, B.; Devrio, A.; Shen, H.; Eslami, M.; and Holstein, K. 2023. Understanding Practices, Challenges, and Opportunities for User-Engaged Algorithm Auditing in Industry Practice. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, 1--18

    2023

  14. [22]

    H.; Nagireddy, M.; Lee, M

    Deng, W. H.; Nagireddy, M.; Lee, M. S. A.; Singh, J.; Wu, Z. S.; Holstein, K.; and Zhu, H. 2022. Exploring how machine learning practitioners (try to) use fairness toolkits. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 473--484

    2022

  15. [23]

    H.; Wang, C.; Han, H

    Deng, W. H.; Wang, C.; Han, H. Z.; Hong, J. I.; Holstein, K.; and Eslami, M. 2025. WeAudit: Scaffolding User Auditors and AI Practitioners in Auditing Generative AI. arXiv preprint arXiv:2501.01397

    2025 arXiv

  16. [24]

    DeVos, A.; Dhabalia, A.; Shen, H.; Holstein, K.; and Eslami, M. 2022. Toward User-Driven Algorithm Auditing: Investigating users’ strategies for uncovering harmful algorithmic behavior. In Proceedings of the 2022 CHI conference on human factors in computing systems, 1--19

    2022

  17. [25]

    Evetts, J. 2006. Introduction: Trust and professionalism: Challenges and occupational changes

    2006

  18. [26]

    E.; Khovanskaya, V.; Crivellaro, C.; Salehi, N.; Dombrowski, L.; Kulkarni, C.; Irani, L.; and Forlizzi, J

    Fox, S. E.; Khovanskaya, V.; Crivellaro, C.; Salehi, N.; Dombrowski, L.; Kulkarni, C.; Irani, L.; and Forlizzi, J. 2020. Worker-Centered Design: Expanding HCI Methods for Supporting Labor. In Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems, ...

    2020

  19. [27]

    J.; and Lewis, D

    Golpayegani, D.; Pandit, H. J.; and Lewis, D. 2023. To Be High-Risk, or Not To Be—Semantic Specifications and Implications of the AI Act’s High-Risk AI Applications and Harmonised Standards. In Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparenc...

    2023

  20. [28]

    L.; Lam, M

    Gordon, M. L.; Lam, M. S.; Park, J. S.; Patel, K.; Hancock, J.; Hashimoto, T.; and Bernstein, M. S. 2022. Jury learning: Integrating dissenting voices into machine learning models. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 1--19

    2022

  21. [29]

    A.; Haidt, J.; Iyer, R.; Koleva, S.; and Ditto, P

    Graham, J.; Nosek, B. A.; Haidt, J.; Iyer, R.; Koleva, S.; and Ditto, P. H. 2011. Mapping the moral domain. Journal of personality and social psychology, 101(2): 366

    2011

  22. [30]

    A.; Haidt, J.; Iyer, R.; Spassena, K.; and Ditto, P

    Graham, J.; Nosek, B. A.; Haidt, J.; Iyer, R.; Spassena, K.; and Ditto, P. H. 2008. Moral foundations questionnaire. Journal of Personality and Social Psychology

    2008

  23. [31]

    Herdel, V.; S \'c epanovi \'c , S.; Bogucka, E.; and Quercia, D. 2024. ExploreGen: Large language models for envisioning the uses and risks of AI technologies. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, volume 7, 584--596

    2024

  24. [32]

    Jakesch, M.; Bu c inca, Z.; Amershi, S.; and Olteanu, A. 2022. How different groups prioritize ethical values for responsible AI. In proceedings of the 2022 ACM conference on fairness, accountability, and transparency, 310--323

    2022

  25. [33]

    M.; and Sambasivan, N

    Kapania, S.; Siy, O.; Clapper, G.; Sp, A. M.; and Sambasivan, N. 2022. ``Because AI is 100\ In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 1--18

    2022

  26. [34]

    Kawakami, A.; Taylor, J.; Fox, S.; Zhu, H.; and Holstein, K. 2024. AI Failure Loops in Feminized Labor: Understanding the Interplay of Workplace AI and Occupational Devaluation. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, volume 7, 683--683

    2024

  27. [35]

    Kelly, J. 2025. Jobs AI will replace first in the workplace shift

    2025

  28. [36]

    Kieslich, K.; Diakopoulos, N.; and Helberger, N. 2023. Anticipating Impacts: Using Large-Scale Scenario Writing to Explore Diverse Implications of Generative AI in the News Environment. arXiv preprint arXiv:2310.06361

    2023 arXiv

  29. [37]

    Kieslich, K.; Helberger, N.; and Diakopoulos, N. 2024. My Future with My Chatbot: A Scenario-Driven, User-Centric Approach to Anticipating AI Impacts. In Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency, FAccT '24, 2071–2085. New York, NY, U...

    2024

  30. [38]

    H.; Lee, J.; Zhang, S.; Eslami, M.; Holstein, K.; Hong, J

    Kingsley, S.; Zhi, J.; Deng, W. H.; Lee, J.; Zhang, S.; Eslami, M.; Holstein, K.; Hong, J. I.; Li, T.; and Shen, H. 2024. Investigating What Factors Influence Users’ Rating of Harmful Algorithmic Bias and Discrimination. In Proceedings of the AAAI Conference on Human Computati...

    2024

  31. [39]

    Kolata, G. 2024. Chatgpt defeated doctors at diagnosing illness - The New York Times

    2024

  32. [40]

    F.; Schaich Borg, J.; Conitzer, V.; and Sinnott-Armstrong, W

    Kramer, M. F.; Schaich Borg, J.; Conitzer, V.; and Sinnott-Armstrong, W. 2018. When Do People Want AI to Make Decisions? In Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society, AIES '18, 204–209. New York, NY, USA: Association for Computing Machinery. ISBN 9...

    2018

  33. [41]

    Z.; Zhang, A

    Kuo, T.-S.; Chen, Q. Z.; Zhang, A. X.; Hsieh, J.; Zhu, H.; and Holstein, K. 2024. PolicyCraft: Supporting Collaborative and Participatory Policy Design through Case-Grounded Deliberation. arXiv preprint arXiv:2409.15644

    2024 arXiv

  34. [42]

    Lee, H. R. 2024. Contrasting Perspectives of Workers: Exploring Labor Relations in Workplace Automation and Potential Interventions. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, 1--17

    2024

  35. [43]

    K.; Kusbit, D.; Kahng, A.; Kim, J

    Lee, M. K.; Kusbit, D.; Kahng, A.; Kim, J. T.; Yuan, X.; Chan, A.; See, D.; Noothigattu, R.; Lee, S.; Psomas, A.; and Procaccia, A. D. 2019. WeBuildAI: Participatory Framework for Algorithmic Governance. Proc. ACM Hum.-Comput. Interact., 3(CSCW)

    2019

  36. [44]

    Li, J.-J.; Pyatkin, V.; Kleiman-Weiner, M.; Jiang, L.; Dziri, N.; Collins, A. G. E.; Schaich Borg, J.; Sap, M.; Choi, Y.; and Levine, S. 2024. SafetyAnalyst: Interpretable, transparent, and steerable LLM safety moderation. arXiv

    2024

  37. [45]

    Lin, H.-T.; Balcan, M.-F.; Hadsell, R.; and Ranzato, M. 2020. Getting Started with NeurIPS 2020. NeurIPS blog

    2020

  38. [46]

    McClain, C. 2025. How the U.S. public and AI experts view Artificial Intelligence

    2025

  39. [47]

    Michaels, E.; Thomas, M.; Reeves, A.; Price, M.; Hasson, R.; Chae, D.; and Allen, A. 2019. Coding the Everyday Discrimination Scale: implications for exposure assessment and associations with hypertension and depression among a cross section of mid-life African American women....

    2019

  40. [48]

    A.; Kawakami, A.; Verdezoto, N.; and Cheon, E

    Ming, J.; Pei, L.; Varanasi, R. A.; Kawakami, A.; Verdezoto, N.; and Cheon, E. 2024. Labor, Visibility, and Technology: Weaving Together Academic Insights and On-Ground Realities. In Companion Publication of the 2024 Conference on Computer-Supported Cooperative Work and Social...

    2024

  41. [49]

    Mun, J.; Jiang, L.; Liang, J.; Cheong, I.; DeCario, N.; Choi, Y.; Kohno, T.; and Sap, M. 2024. Particip-AI: A Democratic Surveying Framework for Anticipating Future AI Use Cases, Harms and Benefits. arXiv:2403.14791

    2024 arXiv

  42. [50]

    I.; and Technology

    of Standards, N. I.; and Technology. 2023. Artificial Intelligence Risk Management Framework (AI RMF 1.0)

    2023

  43. [51]

    OpenAI. 2022. OpenAI: Our approach to alignment research

    2022

  44. [52]

    Parliament, E. 2023. Artificial Intelligence Act: deal on comprehensive rules for trustworthy AI

    2023

  45. [53]

    W.; Jurafsky, D.; Koh, P

    Pierson, E.; Shanmugam, D.; Movva, R.; Kleinberg, J.; Agrawal, M.; Dredze, M.; Ferryman, K.; Gichoya, J. W.; Jurafsky, D.; Koh, P. W.; et al. 2025. Using Large Language Models to Promote Health Equity

    2025

  46. [54]

    Pistilli, G.; Mu \ n oz Ferrandis, C.; Jernite, Y.; and Mitchell, M. 2023. Stronger together: on the articulation of ethical charters, legal tools, and technical documentation in ML. In Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency, 343--354

    2023

  47. [55]

    D.; Scheuerman, M

    Raji, I. D.; Scheuerman, M. K.; and Amironesei, R. 2021. You Can't Sit With Us: Exclusionary Pedagogy in AI Ethics Education. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT '21, 515–525. New York, NY, USA: Association for Computi...

    2021

  48. [56]

    Rajpurkar, P.; Chen, E.; Banerjee, O.; and Topol, E. J. 2022. AI in health and medicine. Nature medicine, 28(1): 31--38

    2022

  49. [57]

    H.; Chattopadhyay, A.; Wu, Z

    Shen, H.; Deng, W. H.; Chattopadhyay, A.; Wu, Z. S.; Wang, X.; and Zhu, H. 2021. Value cards: An educational toolkit for teaching social impacts of machine learning through deliberation. In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency, 850--861

    2021

  50. [58]

    J.; Beattie, L.; and Cramer, H

    Smith, J. J.; Beattie, L.; and Cramer, H. 2023. Scoping fairness objectives and identifying fairness metrics for recommender systems: The practitioners’ perspective. In Proceedings of the ACM Web Conference 2023, 3648--3659

    2023

  51. [59]

    Solaiman, I. 2023. The Gradient of Generative AI Release: Methods and Considerations. In Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency, FAccT '23, 111–122. New York, NY, USA: Association for Computing Machinery. ISBN 9798400701924

    2023

  52. [60]

    L.; Chen, C.; Daum \'e III, H.; Dodge, J.; Duan, I.; et al

    Solaiman, I.; Talat, Z.; Agnew, W.; Ahmad, L.; Baker, D.; Blodgett, S. L.; Chen, C.; Daum \'e III, H.; Dodge, J.; Duan, I.; et al. 2023. Evaluating the social impact of generative ai systems in systems and society. arXiv preprint arXiv:2306.05949

    2023 arXiv

  53. [61]

    J.; Cabral Berenfus, F.; Burke, R.; and Fiesler, C

    Sonboli, N.; Smith, J. J.; Cabral Berenfus, F.; Burke, R.; and Fiesler, C. 2021. Fairness and transparency in recommendation: The users’ perspective. In Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization, 274--279

    2021

  54. [62]

    Sorensen, T.; Jiang, L.; Hwang, J.; Levine, S.; Pyatkin, V.; West, P.; Dziri, N.; Lu, X.; Rao, K.; Bhagavatula, C.; Sap, M.; Tasioulas, J.; and Choi, Y. 2024. Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties. In AAAI

    2024

  55. [63]

    Svensson, L. G. 2006. Professional occupations and status: A sociological study of professional occupations, status and trust

    2006

  56. [64]

    The White House . 2023. National Artificial Intelligence Research Resource Task Force Releases Final Report

    2023

  57. [65]

    Times, T. N. Y. 2024. Will Chatbots Teach Your Children?

    2024

  58. [66]

    P.; and Yuan, T

    Wang, B.; Rau, P.-L. P.; and Yuan, T. 2023. Measuring user competence in using artificial intelligence: validity and reliability of artificial intelligence literacy scale. Behaviour & information technology, 42(9): 1324--1337

    2023

  59. [67]

    J.; Kulkarni, C.; Wilcox, L.; Terry, M.; and Madaio, M

    Wang, Z. J.; Kulkarni, C.; Wilcox, L.; Terry, M.; and Madaio, M. 2024. Farsight: Fostering Responsible AI Awareness During AI Application Prototyping. In Proceedings of the CHI Conference on Human Factors in Computing Systems, 1--40

    2024

  60. [68]

    Y.; and Nguyen, T

    Wong, R. Y.; and Nguyen, T. 2021. Timelines: A world-building activity for values advocacy. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems, 1--15

    2021

  61. [69]

    Wongpakaran, N.; Wongpakaran, T.; Wedding, D.; and Gwet, K. L. 2013. A comparison of Cohen’s Kappa and Gwet’s AC1 when calculating inter-rater reliability coefficients: a study conducted with personality disorder samples. BMC medical research methodology, 13: 1--7

    2013

  62. [70]

    G.; Cho, J.; Ahmadpour, N.; Roto, V.; Sachathep, T.; Liu, J.; Sawan, M.; Song, D.; C ai\' c , M.; Cheng, L.; Liu, R.; Kettley, S.; Soares, L.; Grace, K.; and Astell-Burt, T

    Wu, Y.; Lee, J.-J.; Pillai, A. G.; Cho, J.; Ahmadpour, N.; Roto, V.; Sachathep, T.; Liu, J.; Sawan, M.; Song, D.; C ai\' c , M.; Cheng, L.; Liu, R.; Kettley, S.; Soares, L.; Grace, K.; and Astell-Burt, T. 2024. Collective Imaginaries for the Futures of Care Work. In Companion ...

    2024

  63. [71]

    Zhai, C.; Wibowo, S.; and Li, L. D. 2024. The effects of over-reliance on AI dialogue systems on students' cognitive abilities: a systematic review. Smart Learning Environments, 11(1): 28

    2024

  64. [72]

    Zhang, A.; Walker, O.; Nguyen, K.; Dai, J.; Chen, A.; and Lee, M. K. 2023. Deliberating with AI: Improving Decision-Making for the Future through Participatory AI Design and Stakeholder Deliberation. Proc. ACM Hum.-Comput. Interact., 7(CSCW1)

    2023

  65. [73]

    Zhang, B.; and Dafoe, A. 2020. US public opinion on the governance of artificial intelligence. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 187--193

    2020

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.