arxiv: 2605.03210 · v1 · submitted 2026-05-04 · 💻 cs.CY · cs.AI· econ.GN· q-fin.EC

Recognition: unknown

Human-Provenance Verification should be Treated as Labor Infrastructure in AI-Saturated Markets

David Khachaturov, Erin McGurk

Pith reviewed 2026-05-08 17:01 UTC · model grok-4.3

classification 💻 cs.CY cs.AIecon.GNq-fin.EC

keywords human provenanceAI governancelabor infrastructureperformative humanityknowledge workverification systemsbarbell economyVeblen goods

0 comments

The pith

In AI-saturated markets, verified human presence will command premiums that require treating provenance verification as core labor infrastructure.

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

The paper argues that generative AI will make many standardized cognitive and creative tasks abundant and cheap at low marginal cost, weakening scarcity premiums for middle-tier knowledge work and producing an asymmetric barbell economy. At one pole lies high-volume synthetic production controlled by AI infrastructure owners; at the other lies scarce human labor valued specifically for its verified human character. Because these human-provenance premiums only materialize when buyers can credibly distinguish human involvement, the authors conclude that AI governance must treat verification systems as labor infrastructure rather than optional authenticity labels. They define three forms of such valued labor—relational presence, aesthetic provenance, and accountability—and propose that human elements retain premium value in hybrid work only when they are constitutive of the output rather than incidental.

Core claim

AI compresses the value of standardized middle-tier labor by making good-enough synthetic substitutes scalable at low marginal cost, hollowing out the middle of the skill distribution. This compression reallocates demand toward work valued for its visible human character, which the authors term performative humanity and divide into relational presence, aesthetic provenance, and accountability. As these premiums depend on credible verification, AI governance should treat human-provenance systems as labor infrastructure rather than luxury authenticity labels. Hybrid human-AI work retains human value when human judgment, attention, accountability, authorship, or relational participation is not

What carries the argument

Human-provenance premiums, which arise when verified human presence adds value to outputs in markets where AI supplies scalable substitutes for standardized tasks.

If this is right

Standardized middle-tier cognitive and creative work will lose scarcity value as AI substitutes become widely available.
Labor demand will shift toward roles where visible human character is central, including relational presence, aesthetic provenance, and accountability.
Hybrid human-AI outputs will be valued for human elements only when those elements are constitutive of what is purchased.
Governance should prioritize scalable verification systems as essential labor infrastructure to enable capture of human-provenance premiums.

Where Pith is reading between the lines

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

Private control of verification tools could allow AI owners to gate access to human-premium markets and concentrate economic power.
Training and education systems might need to emphasize skills in relational and accountability work over routine expertise.
The same logic could apply to domains such as healthcare or legal services where buyer willingness to pay for verified human accountability emerges.
Without public investment in verification, the barbell structure may reduce the overall share of income going to labor.

Load-bearing premise

AI will reliably produce scalable, good-enough substitutes for middle-tier standardized cognitive and creative tasks, driving an asymmetric reallocation of demand toward performative human labor.

What would settle it

Empirical observation that middle-tier knowledge-work wages and employment remain stable or rise as AI adoption scales, or that markets show no sustained price differential between verified human and synthetic outputs in creative and professional services.

Figures

Figures reproduced from arXiv: 2605.03210 by David Khachaturov, Erin McGurk.

**Figure 1.** Figure 1: The barbell order as a stylized value-capture curve. The horizontal axis is a latent ordering view at source ↗

**Figure 2.** Figure 2: The verification gate separating premium human-provenance markets from synthetic view at source ↗

**Figure 3.** Figure 3: Historical shapes of labor regimes. The panels schematically compare how different eco view at source ↗

**Figure 4.** Figure 4: Schematic transition from a broad middle-class economy to an AI-era barbell structure. The view at source ↗

read the original abstract

We argue that AI-saturated markets are likely to create Veblen-good premiums, which we term human-provenance premiums, for verified human presence, and hence AI governance should treat human-provenance verification as labor infrastructure. Generative and agentic AI systems lower the cost of many standardized cognitive, creative, and coordination tasks, weakening the scarcity premiums that have supported much middle-tier knowledge work. We argue that this pressure may produce an asymmetric barbell-shaped structure of value capture in advanced economies: high-volume synthetic production controlled by owners of AI infrastructure at one pole, and scarce, high-status human labor valued for verified human presence at the other. We advance three claims. First, AI compresses the value of standardized middle-tier labor by making good-enough synthetic substitutes scalable at low marginal cost, hollowing out the middle of the skill distribution currently categorized by knowledge work. Second, this compression reallocates demand for human labor toward work valued for its visible human character. We term this performative humanity and distinguish three forms of labor: relational presence, aesthetic provenance, and accountability. Third, as these premiums depend on credible verification, AI governance should treat human-provenance systems as labor infrastructure rather than as luxury authenticity labels. To evaluate hybrid human-AI work, we propose constitutive human presence as the relevant standard: human labor retains premium value when human judgment, attention, accountability, authorship, or relational participation is not incidental to the output but constitutive of what is being purchased.

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

The paper frames AI-driven task compression as likely to create verifiable-human premiums in a barbell labor market and calls for treating provenance systems as infrastructure, but the case rests on untested premises without data or modeling.

read the letter

The core claim is that generative AI will hollow out middle-tier knowledge work through cheap substitutes, leaving demand concentrated on high-volume synthetic output and on scarce human labor valued specifically for its verified presence. The paper breaks this into three forms—relational presence, aesthetic provenance, and accountability—and proposes “constitutive human presence” as the test for when human input actually adds the premium rather than being incidental. It then concludes that governance should build verification infrastructure around these premiums instead of treating them as optional authenticity labels. That framing is the main thing worth noting; it applies Veblen-good logic to current AI capabilities in a direct way and gives a usable distinction for hybrid work evaluation. The logic from cost compression to asymmetric reallocation follows cleanly from the stated premises. The three forms of performative humanity are spelled out plainly enough to be usable in discussion. The soft spots are straightforward and central. No empirical patterns, elasticity estimates, or equilibrium model are offered to show the barbell will actually form or that the human-premium tasks will remain hard to mimic at scale. The definitions of the new terms are largely circular: the premiums are defined by the value they are meant to explain. The stress-test point lands—the argument needs AI substitutes to stay inadequate for relational and accountability work, yet the paper gives no independent reason to expect that limit to hold once persistent identity or simulation improves. This is a policy-oriented conceptual piece rather than a result with new evidence. Readers working on AI labor policy or governance will find the terminology and the infrastructure recommendation useful to chew on, even if they disagree with the direction. It is coherent on its own terms and engages the relevant economic ideas, so it deserves a serious referee to test the assumptions and see whether the framing can be tightened or falsified.

Referee Report

3 major / 2 minor

Summary. The paper claims that generative and agentic AI will compress the value of standardized middle-tier cognitive and creative labor through scalable low-cost substitutes, producing an asymmetric barbell labor market: high-volume synthetic output controlled by AI infrastructure owners at one pole, and scarce high-status human labor valued for verified presence (termed performative humanity in relational presence, aesthetic provenance, and accountability forms) at the other. It introduces human-provenance premiums and constitutive human presence as the relevant standard for hybrid work, and concludes that AI governance must treat credible human-provenance verification systems as labor infrastructure rather than luxury authenticity labels.

Significance. If the barbell structure and associated premiums materialize, the reframing would shift AI governance from regulating synthetic content to building verification infrastructure that sustains new forms of human labor value capture. The distinction between incidental and constitutive human elements provides a policy-relevant lens for evaluating hybrid systems, though the manuscript offers no empirical data, equilibrium model, or falsifiable predictions to ground the claims.

major comments (3)

[Abstract, paragraph advancing the three claims] The second claim (reallocation toward performative humanity) and the resulting recommendation rest on the premise that AI cannot credibly scale substitutes for relational presence, aesthetic provenance, or accountability tasks; no equilibrium model, elasticity estimates, or counter-scenario analysis is supplied to derive the barbell outcome or rule out future model advances that close this gap.
[Abstract, definition of constitutive human presence] Human-provenance premiums and constitutive human presence are defined in terms of the premium value they are intended to explain (e.g., human elements are constitutive when they retain premium value), rendering the central argument self-referential rather than independently grounded.
[Abstract, first claim] The first claim (compression of middle-tier labor) is logically consistent with cost-reduction premises but provides no quantitative bounds, historical analogies, or falsifiable predictions to support the specific asymmetric reallocation to the three forms of performative humanity.

minor comments (2)

[Terminology throughout] The manuscript introduces several novel terms without explicit mapping to or differentiation from existing concepts in labor economics (e.g., Veblen goods) or AI ethics (e.g., authenticity and provenance frameworks).
[References section] No references are provided to empirical studies on AI substitution effects in knowledge work or to governance proposals for verification systems.

Simulated Author's Rebuttal

3 responses · 0 unresolved

Thank you for the referee's thoughtful review. We respond to each major comment below, clarifying the conceptual nature of the paper and indicating revisions where appropriate to strengthen the argument.

read point-by-point responses

Referee: The second claim (reallocation toward performative humanity) and the resulting recommendation rest on the premise that AI cannot credibly scale substitutes for relational presence, aesthetic provenance, or accountability tasks; no equilibrium model, elasticity estimates, or counter-scenario analysis is supplied to derive the barbell outcome or rule out future model advances that close this gap.

Authors: We acknowledge that the manuscript does not include a formal equilibrium model or quantitative elasticity estimates, as it is a conceptual paper proposing a reframing of AI governance rather than an empirical or modeling study. The premise is based on current limitations in AI for tasks requiring genuine relational presence and accountability, supported by references in the full text to AI capability assessments. To address the concern about future advances, we will add a dedicated subsection in the revision discussing potential counter-scenarios where AI might close these gaps and the implications for the barbell structure. revision: partial
Referee: Human-provenance premiums and constitutive human presence are defined in terms of the premium value they are intended to explain (e.g., human elements are constitutive when they retain premium value), rendering the central argument self-referential rather than independently grounded.

Authors: This is a valid point regarding the definitional structure. We will revise the abstract and the relevant sections to first ground the concepts in observable market phenomena—such as existing premiums for human-verified services in fields like law, medicine, and creative arts—before linking them to the proposed premiums. This will make the definitions more independent and less circular. revision: yes
Referee: The first claim (compression of middle-tier labor) is logically consistent with cost-reduction premises but provides no quantitative bounds, historical analogies, or falsifiable predictions to support the specific asymmetric reallocation to the three forms of performative humanity.

Authors: The full manuscript does reference historical analogies from the automation of routine tasks and cites literature on labor market polarization. However, we agree that more explicit quantitative bounds and falsifiable predictions are not provided, as the paper focuses on qualitative implications for governance. In the revision, we will expand the discussion of historical analogies and clarify that the asymmetric reallocation is presented as a plausible outcome under current trends rather than a definitive prediction. revision: partial

Circularity Check

0 steps flagged

No significant circularity; conceptual argument is self-contained

full rationale

The paper advances three sequential claims about AI-driven labor compression, reallocation toward performative human labor, and the resulting policy implication for treating verification as infrastructure. These rest on posited economic premises about AI scalability and the non-mimickable character of certain human attributes rather than on any internal definitions, fitted parameters, or self-citations that reduce the conclusions to the inputs by construction. New terms such as 'human-provenance premiums' and 'constitutive human presence' are introduced to label the posited phenomena and the proposed evaluative standard, but the text does not define the premiums in terms of the infrastructure recommendation or vice versa; the 'hence' linkage follows from the stated dependence of premiums on credible verification. No equations, self-referential loops, or load-bearing citations to prior author work appear in the derivation chain. The argument is therefore a forward-looking policy analysis whose validity hinges on external empirical assumptions about AI capabilities, not on circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 3 invented entities

The central claims rest on several domain assumptions about AI capabilities and market responses plus three newly introduced conceptual entities that lack independent empirical grounding.

axioms (2)

domain assumption Generative and agentic AI systems lower the cost of many standardized cognitive, creative, and coordination tasks at low marginal cost.
Invoked in the opening paragraph to establish the compression of middle-tier labor value.
domain assumption Demand for human labor will reallocate toward work valued specifically for its visible human character when synthetic substitutes are available.
Stated as the second claim without supporting data or model.

invented entities (3)

human-provenance premiums no independent evidence
purpose: To name the market premium for verified human presence in AI-saturated markets.
New term introduced to frame the value reallocation; no independent evidence provided.
performative humanity no independent evidence
purpose: To categorize labor valued for visible human character into relational presence, aesthetic provenance, and accountability.
Invented framing device to organize the second claim; no external validation.
constitutive human presence no independent evidence
purpose: To define the standard for when human involvement retains premium value in hybrid work.
Proposed evaluative criterion that the third claim depends on; defined within the paper.

pith-pipeline@v0.9.0 · 5576 in / 1580 out tokens · 93805 ms · 2026-05-08T17:01:17.478977+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Reference graph

Works this paper leans on

62 extracted references · 30 canonical work pages

[1]

Macmillan, 1899

Thorstein Veblen.The Theory of the Leisure Class. Macmillan, 1899
[2]

Leibenstein, Bandwagon, snob, and veblen effects in the theory of consumers’ demand, The Quarterly Journal of Economics 64 (2) (1950) 183–207.doi:10.2307/1882692

Harvey Leibenstein. Bandwagon, snob, and veblen effects in the theory of consumers’ demand. The Quarterly Journal of Economics, 64(2):183–207, 1950. doi: 10.2307/1882692

work page doi:10.2307/1882692 1950
[4]

Felten, Manav Raj, and Robert Seamans

Edward W. Felten, Manav Raj, and Robert Seamans. Occupational heterogeneity in exposure to automation: Evidence from task-based measures.Strategic Management Journal, 42(10): 1855–1870, 2021. doi: 10.1002/smj.3287

work page doi:10.1002/smj.3287 2021
[5]

Generative AI and jobs: A global analysis of potential effects on job quantity and quality

Pawel Gmyrek, Janine Berg, and David Bescond. Generative AI and jobs: A global analysis of potential effects on job quantity and quality. International Labour Organization Working Paper No. 96, 2023

2023
[6]

AI will transform the global economy

Kristalina Georgieva. AI will transform the global economy. let’s make sure it benefits humanity. IMF Blog, 2024

2024
[7]

URLhttps://www.aeaweb.org/articles?id=10.1257/aer

David H. Autor and David Dorn. The growth of low-skill service jobs and the polarization of the US labor market.American Economic Review, 103(5):1553–1597, 2013. doi: 10.1257/aer. 103.5.1553

work page doi:10.1257/aer 2013
[8]

Explaining job polarization: Routine-biased technological change and offshoring.American Economic Review, 104(8):2509–2526, 2014

Maarten Goos, Alan Manning, and Anna Salomons. Explaining job polarization: Routine-biased technological change and offshoring.American Economic Review, 104(8):2509–2526, 2014. doi: 10.1257/aer.104.8.2509

work page doi:10.1257/aer.104.8.2509 2014
[9]

Varian.Information Rules: A Strategic Guide to the Network Economy

Carl Shapiro and Hal R. Varian.Information Rules: A Strategic Guide to the Network Economy. Harvard Business School Press, 1999

1999
[10]

Polity, 2017

Nick Srnicek.Platform Capitalism. Polity, 2017

2017
[11]

PublicAffairs, 2019

Shoshana Zuboff.The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs, 2019

2019
[12]

Melville House, 2024

Yanis Varoufakis.Technofeudalism: What Killed Capitalism. Melville House, 2024

2024
[13]

Techno-feudalism or platform capitalism? conceptualising the digital society

Jeremy Gilbert. Techno-feudalism or platform capitalism? conceptualising the digital society. European Journal of Social Theory, 27(4):561–578, 2024. doi: 10.1177/13684310241276474

work page doi:10.1177/13684310241276474 2024
[14]

Douglas Bernheim

Laurie Simon Bagwell and B. Douglas Bernheim. Veblen effects in a theory of conspicuous consumption.American Economic Review, 86(3):349–373, 1996

1996
[15]

Christoph Fuchs, Martin Schreier, and Stijn M. J. Van Osselaer. The handmade effect: What’s love got to do with it?Journal of Marketing, 79(2):98–110, 2015. doi: 10.1509/jm.14.0018

work page doi:10.1509/jm.14.0018 2015
[17]

Darby and Edi Karni

Michael R. Darby and Edi Karni. Free competition and the optimal amount of fraud.Journal of Law and Economics, 16(1):67–88, 1973. doi: 10.1086/466756

work page doi:10.1086/466756 1973
[18]

Technical change, globalization, and the labour market: British and american experience since 1620.Oxford Open Economics, 3(Supplement 1):i178–i211, 07 2024

Robert C Allen. Technical change, globalization, and the labour market: British and american experience since 1620.Oxford Open Economics, 3(Supplement 1):i178–i211, 07 2024. ISSN 2752-5074. doi: 10.1093/ooec/odad033. URL https://doi.org/10.1093/ooec/odad033

work page doi:10.1093/ooec/odad033 2024
[19]

The trademark function of authorship.Boston University Law Review, 85(4):1171–1242, 2005

Greg Lastowka. The trademark function of authorship.Boston University Law Review, 85(4):1171–1242, 2005. URL https://www.bu.edu/law/journals-archive/bulr/ volume85n4/LASTOWKA.pdf. 10

2005
[20]

Economic arguments in favour of reducing copyright protection for generative ai inputs and outputs

Bertin Martens. Economic arguments in favour of reducing copyright protection for generative ai inputs and outputs. Working Paper 09/2024, Bruegel, April 2024. URL https://www. econstor.eu/bitstream/10419/297853/1/188870473X.pdf

2024
[21]

M.et al.Identification of short-range ordering motifs in semiconductors.Science389, 1342–1346 (2025)

Shakked Noy and Whitney Zhang. Experimental evidence on the productivity effects of generative artificial intelligence.Science, 381(6654):187–192, 2023. doi: 10.1126/science. adh2586

work page doi:10.1126/science 2023
[22]

The Quarterly Journal of Economics , author =

Erik Brynjolfsson, Danielle Li, and Lindsey Raymond. Generative AI at work.The Quarterly Journal of Economics, 140(2):889–942, 2025. doi: 10.1093/qje/qjae044

work page doi:10.1093/qje/qjae044 2025
[23]

Generative AI and jobs: A refined global index of occupational exposure

Pawel Gmyrek, Janine Berg, Karol Kami´nski, Filip Konopczy´nski, Agnieszka Ładna, Balint Nafradi, Konrad Rosłaniec, and Marek Troszy´nski. Generative AI and jobs: A refined global index of occupational exposure. ILO Working Paper 140, International Labour Organization, 2025

2025
[24]

Kim A. Weeden. Why do some occupations pay more than others? social closure and earnings inequality in the united states.American Journal of Sociology, 108(1):55–101, 2002. doi: 10.1086/344121

work page doi:10.1086/344121 2002
[25]

Rosenthal

Stephen Lippmann and Jeffrey E. Rosenthal. Do displaced workers lose occupational prestige? Social Science Research, 37(2):642–656, 2008. doi: 10.1016/j.ssresearch.2007.08.006

work page doi:10.1016/j.ssresearch.2007.08.006 2008
[26]

What can machine learning do? workforce implications

Erik Brynjolfsson and Tom Mitchell. What can machine learning do? workforce implications. Science, 358(6370):1530–1534, 2017. doi: 10.1126/science.aap8062. URL https://www. science.org/doi/10.1126/science.aap8062

work page doi:10.1126/science.aap8062 2017
[27]

Autor, Frank Levy, and Richard J

David H. Autor, Frank Levy, and Richard J. Murnane. The skill content of recent techno- logical change: An empirical exploration.The Quarterly Journal of Economics, 118(4): 1279–1333, 2003. doi: 10.1162/003355303322552801. URL https://doi.org/10.1162/ 003355303322552801

work page doi:10.1162/003355303322552801 2003
[28]

Humans versus ai: Whether and why we prefer human-created compared to ai-created artwork.Cognitive Research: Principles and Implications, 8(1):42, 2023

Lucas Bellaiche, Rohin Shahi, Martin Harry Turpin, Anya Ragnhildstveit, Shawn Sprockett, Nathaniel Barr, Alexander Christensen, and Paul Seli. Humans versus ai: Whether and why we prefer human-created compared to ai-created artwork.Cognitive Research: Principles and Implications, 8(1):42, 2023. doi: 10.1186/s41235-023-00499-6. URL https://doi.org/10. 1186...

work page doi:10.1186/s41235-023-00499-6 2023
[29]

Univer- sity of California Press, 1983

Arlie Russell Hochschild.The Managed Heart: Commercialization of Human Feeling. Univer- sity of California Press, 1983

1983
[30]

Zelizer.The Social Meaning of Money

Viviana A. Zelizer.The Social Meaning of Money. Basic Books, New York, 1994

1994
[31]

Too human and not human enough: A grounded theory analysis of mental health harms from emotional dependence on the social chatbot Replika.New Media & Society, 26: 5923–5941, 2024

Linnea Laestadius, Andrea Bishop, Michael Gonzalez, Diana Illen ˇcík, and Celeste Campos- Castillo. Too human and not human enough: A grounded theory analysis of mental health harms from emotional dependence on the social chatbot Replika.New Media & Society, 26: 5923–5941, 2024. doi: 10.1177/14614448221142007

work page doi:10.1177/14614448221142007 2024
[32]

Kathleen Kara Fitzpatrick, Alison Darcy, and Molly Vierhile. Delivering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (Woebot): A randomized controlled trial.JMIR Mental Health, 4(2):e19,
[33]

doi: 10.2196/mental.7785

work page doi:10.2196/mental.7785
[34]

The New Press, New York, 2001

Nancy Folbre.The Invisible Heart: Economics and Family Values. The New Press, New York, 2001

2001
[35]

Contradictions of capital and care.New Left Review, 100:99–117, 2016

Nancy Fraser. Contradictions of capital and care.New Left Review, 100:99–117, 2016

2016
[36]

Art and the machine: Why people devalue AI-generated creative work

Graelin Mandel and Alex Imas. Art and the machine: Why people devalue AI-generated creative work. SSRN Working Paper No. 6302659, March 2026. URL https://papers.ssrn.com/ sol3/papers.cfm?abstract_id=6302659

2026
[37]

Morris M. Kleiner. Occupational licensing.Journal of Economic Perspectives, 14(4):189–202,
[38]

URLhttps://doi.org/10.1257/jep.14.4.189

doi: 10.1257/jep.14.4.189. URLhttps://doi.org/10.1257/jep.14.4.189. 11

work page doi:10.1257/jep.14.4.189
[39]

The Review of Economic Studies , author =

Carl Shapiro. Investment, moral hazard, and occupational licensing.The Review of Economic Studies, 53(5):843–862, 1986. doi: 10.2307/2297722. URL https://doi.org/10.2307/ 2297722

work page doi:10.2307/2297722 1986
[40]

Regulation (EU) 2024/1689 of the european parliament and of the council of 13 june 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)

European Union. Regulation (EU) 2024/1689 of the european parliament and of the council of 13 june 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). Official Journal of the European Union, 2024

2024
[41]

David H. Autor. The task approach to labor markets: An overview.Journal for Labour Market Research, 46(3):185–199, 2013. doi: 10.1007/s12651-013-0128-z. URL https: //doi.org/10.1007/s12651-013-0128-z

work page doi:10.1007/s12651-013-0128-z 2013
[42]

Newman and Paul Bloom

George E. Newman and Paul Bloom. Art and authenticity: The importance of originals in judgments of value.Journal of Experimental Psychology: General, 141(3):558–569, 2012. doi: 10.1037/a0026035. URLhttps://doi.org/10.1037/a0026035

work page doi:10.1037/a0026035 2012
[43]

University of Chicago Press, Chicago, 2001

Eliot Freidson.Professionalism: The Third Logic: On the Practice of Knowledge. University of Chicago Press, Chicago, 2001. ISBN 9780226262031

2001
[44]

Brotheridge and Alicia A

Celeste M. Brotheridge and Alicia A. Grandey. Emotional labor and burnout: Comparing two perspectives of people work.Journal of Vocational Behavior, 60(1):17–39, 2002. doi: 10.1006/jvbe.2001.1815. URLhttps://doi.org/10.1006/jvbe.2001.1815

work page doi:10.1006/jvbe.2001.1815 2002
[45]

Hülsheger and Anna F

Ute R. Hülsheger and Anna F. Schewe. On the costs and benefits of emotional labor: A meta-analysis of three decades of research.Journal of Occupational Health Psychology, 16(3): 361–389, 2011. doi: 10.1037/a0022876. URLhttps://doi.org/10.1037/a0022876

work page doi:10.1037/a0022876 2011
[46]

Zelizer.The Purchase of Intimacy

Viviana A. Zelizer.The Purchase of Intimacy. Princeton University Press, Princeton, NJ, 2005. ISBN 9780691124085

2005
[47]

Stanford University Press, Stanford, CA, 2010

Eileen Boris and Rhacel Salazar Parreñas, editors.Intimate Labors: Cultures, Technologies, and the Politics of Care. Stanford University Press, Stanford, CA, 2010. ISBN 9780804761932

2010
[48]

About cara, 2022

Cara Project. About cara, 2022. URLhttps://blog.cara.app/about

2022
[49]

Artificial intelligence, 2025

Writers Guild of America West. Artificial intelligence, 2025. URL https://www.wga.org/ contracts/know-your-rights/artificial-intelligence

2025
[50]

Harvard University Press, Cambridge, MA, 1976

Fred Hirsch.Social Limits to Growth. Harvard University Press, Cambridge, MA, 1976

1976
[51]

Journal of Political Economy , author =

Daron Acemoglu and Pascual Restrepo. Robots and jobs: Evidence from US labor markets. Journal of Political Economy, 128(6):2188–2244, 2020. doi: 10.1086/705716

work page doi:10.1086/705716 2020
[52]

Recommendation on the ethics of artificial intelligence

UNESCO. Recommendation on the ethics of artificial intelligence. Technical report, United Nations Educational, Scientific and Cultural Organization, Paris, November 2021. URL https: //unesdoc.unesco.org/ark:/48223/pf0000381137. Adopted by UNESCO’s General Conference at its 41st session on 23 November 2021

2021
[53]

Automation and new tasks: How technology displaces and reinstates labor.Journal of Economic Perspectives, 33(2):3–30, 2019

Daron Acemoglu and Pascual Restrepo. Automation and new tasks: How technology displaces and reinstates labor.Journal of Economic Perspectives, 33(2):3–30, 2019. doi: 10.1257/jep.33. 2.3. URLhttps://doi.org/10.1257/jep.33.2.3

work page doi:10.1257/jep.33 2019
[54]

Automation, bargaining power, and labor market fluctuations

Sylvain Leduc and Zheng Liu. Automation, bargaining power, and labor market fluctuations. American Economic Journal: Macroeconomics, 16(4):311–349, 2024. doi: 10.1257/mac. 20220181. URLhttps://doi.org/10.1257/mac.20220181

work page doi:10.1257/mac 2024
[55]

Kirk and Julian Givi

Colleen P. Kirk and Julian Givi. The ai-authorship effect: Understanding authenticity, moral disgust, and consumer responses to ai-generated marketing communications.Journal of Business Research, 186:114984, 2025. doi: 10.1016/j.jbusres.2024.114984. URL https://doi.org/ 10.1016/j.jbusres.2024.114984

work page doi:10.1016/j.jbusres.2024.114984 2025
[56]

Artificial intelligence risk management framework (AI RMF 1.0)

Elham Tabassi. Artificial intelligence risk management framework (AI RMF 1.0). National Institute of Standards and Technology, 2023. 12

2023
[57]

C2PA and content credentials ex- plainer

Coalition for Content Provenance and Authenticity. C2PA and content credentials ex- plainer. C2PA Technical Specification Explainer, 2024. URL https://spec.c2pa.org/ specifications/specifications/2.4/explainer/Explainer.html

2024
[58]

S., Kumar, A., Balasubramanian, S., Wang, W., and Feizi, S

Vinu Sankar Sadasivan, Aounon Kumar, Sriram Balasubramanian, Wenxiao Wang, and Soheil Feizi. Can ai-generated text be reliably detected?, 2023. URL https://arxiv.org/abs/ 2303.11156

work page arXiv 2023
[59]

Watermarking needs input repetition masking, 2025

David Khachaturov, Robert Mullins, Ilia Shumailov, and Sumanth Dathathri. Watermarking needs input repetition masking, 2025. URLhttps://arxiv.org/abs/2504.12229

work page arXiv 2025
[60]

Chandra et al

B. Chandra et al. Reducing risks posed by synthetic content: An overview of technical approaches to digital content transparency. Technical Report NIST AI 100-4, National Institute of Standards and Technology, 2024. URLhttps://www.nist.gov/publications/ reducing-risks-posed-synthetic-content-overview-technical-approaches-digital-content

2024
[61]

Can companies arbitrarily reduce pay and terminate contracts amid the ai wave? a hangzhou intermediate people’s court press conference gives the final word, 2026

Jiawei Xu, Lan Wei, and Jialing Song. Can companies arbitrarily reduce pay and terminate contracts amid the ai wave? a hangzhou intermediate people’s court press conference gives the final word, 2026. URL https://ori.hangzhou.com.cn/ornews/content/2026-04/29/ content_9214417.htm

2026
[62]

Generative AI and the SME workforce: New survey evidence

OECD. Generative AI and the SME workforce: New survey evidence. OECD Publishing, 2025

2025
[63]

Michael Spence

A. Michael Spence. Job market signaling.The Quarterly Journal of Economics, 87(3):355–374,
[64]

13 A Summary of paper’s core claims Table 2: Evidentiary status of the paper’s core claims

doi: 10.2307/1882010. 13 A Summary of paper’s core claims Table 2: Evidentiary status of the paper’s core claims. Claim Evidence currently available Status AI reduces the cost of standard- ized cognitive work Task-level productivity studies and information-goods cost theory [ 21, 22, 9] Supported as mechanism Middle-tier labor loses scarcity premium Occup...

work page doi:10.2307/1882010