Why this paper is being considered
The paper maps possible AI capability paths into scenarios for GDP, labour share, wages, reallocation and unemployment through 2030 (modest, substantial, extreme). It matters for how policymakers and funders plan for a range of AI economic outcomes. An evaluation would focus on the model structure, how the scenarios are anchored to evidence, and the interactive explorer's sensitivity to assumptions.
What an expert evaluation could add: An independent reproduction covers the published calculations; public debate leaves the economic assumptions open. A focused expert evaluation of calibration, labor adjustment, demand and policy interpretation could still add substantial value. Evidence and commissioning case. AI-assisted judgment, 2026-10-07; separate from ratings and completed evaluations.
AI-generated criterion ratings and reasoning
Scores come from AI prioritization; the accompanying explanations include AI-assisted source checks. These are provisional judgments, separate from human ratings and commissioned evaluations.
Decision relevance
7.2/10
The research has a strong but not exceptional global-welfare/VoI case: it addresses potentially large human welfare stakes from AI-driven productivity growth, labor displacement, unemployment, inequality, and institutional adaptation, but its direct scope is mainly US macro-labor outcomes through 2030. Its incremental contribution is a structured model linking AI capability/adoption assumptions to economic outcomes, which could improve planning for redistribution, labor-market policy, insurance, and governance triggers; however, it does not assign probabilities, directly estimate causal effects, or model LMIC welfare, animal welfare, or catastrophic-risk channels. The welfare value depends heavily on uptake and interpretation: useful as a scenario framework, weaker as decisive evidence for pacing, export controls, or compute governance.
- Paper claim to check A task-based automation framework can translate assumptions about AI capabilities, adoption, productivity gains, and automation versus augmentation into paths for GDP, labor share, wages, reallocation, and unemployment through 2030. Source abstract
Value of added scrutiny
8.5/10
The paper received early public coverage. An independent Econ-ARK reimplementation now reports reproducing all 226 checkable published numbers after recovering twelve omitted or rounded implementation details. This checks implementation rather than establishing the plausibility of the scenarios or economic assumptions; a new evaluation should concentrate on those substantive questions.
Timing
9.4/10
Timing value is very high: this is a September 2026 working paper from The Anthropic Institute and appears to be at a stage where independent critique could still affect revisions and downstream policy use. The supplied targeted public-scrutiny evidence is empty, so I cannot verify whether substantial independent expert review already exists; I therefore treat prior scrutiny as unclear rather than absent. Given the paper's prominence and likely policy salience, even limited prior scrutiny would leave a large marginal role for an Unjournal-style evaluation unless there is already a substantial public expert debate not captured here.
- Source record Publication-stage and date evidence should be checked in the linked paper record. TARGETED_CURATED
Methodological potential
7.2/10
Evaluation would need economists comfortable with task-based automation, macro growth modeling, search-and-matching labor models, survey design, and AI capability forecasting. The hardest issue is not checking a single empirical estimate but assessing whether the scenario assumptions, calibration choices, omitted mechanisms, and interpretation are appropriate for policy use; reviewers should separate model-internal validity from whether the scenarios deserve decision weight.
- Paper claim to check A task-based automation framework can translate assumptions about AI capabilities, adoption, productivity gains, and automation versus augmentation into paths for GDP, labor share, wages, reallocation, and unemployment through 2030. Source abstract
Prominence
8.2/10
The scoring model estimated prominence from the paper's venue, authors, institutional setting, and visibility. The model did not supply a criterion-specific explanation. Current public-attention status: Independent computational reproduction; substantive model evaluation still valuable.
- Publisher page Anthropic published the paper with an interactive scenario explorer. The publisher is also the authors' employer. Anthropic
Likely influence
7.4/10
The scoring model estimated how far the findings could shape later research or decisions, without supplying a criterion-specific explanation. Current public-attention status: Independent computational reproduction; substantive model evaluation still valuable. See public-attention evidence below.
- Publisher page Anthropic published the paper with an interactive scenario explorer. The publisher is also the authors' employer. Anthropic
What the paper says
Source abstract · Abstract text copied from the working-paper PDF; line breaks and whitespace normalized.
This paper presents a framework for assessing the economic consequences of AI between 2026 and 2030. In the model, AI automates a growing share of cognitive work, raising productivity and displacing workers who must search for jobs in other occupations. The model maps future paths of AI capabilities into implied paths for GDP, the labor share, wages, labor reallocation, and unemployment. We illustrate the framework by considering three scenarios: modest, substantial, and extreme. Under modest change, AI adds less than half a point to GDP growth by 2030 and raises unemployment by a tenth of a point. In the extreme change scenario, AI has transformative effects, with AI performing almost half of today’s cognitive work by 2030. GDP growth then rises to 15 percent per year, the labor share of income falls from 60 to 45 percent, and nearly one in five cognitive workers is unemployed. We also surveyed US adults about their expectations for AI. Views vary widely, but the median respondent’s answers are consistent with our substantial change scenario in which, by 2030, GDP rises by 8 percent and cognitive employment declines by 4 percent. The model offers a structured way to compare possibilities for our economic future under different expectations about AI.
Claims to check
- A task-based automation framework can translate assumptions about AI capabilities, adoption, productivity gains, and automation versus augmentation into paths for GDP, labor share, wages, reallocation, and unemployment through 2030.
- In an extreme-change scenario, AI performs almost half of today's cognitive work by 2030, GDP growth rises to about 15 percent per year, labor's income share falls from 60 percent to 45 percent, and nearly one in five cognitive workers is unemployed.
- A representative survey of 10,980 US adults implies median expectations close to the paper's substantial-change scenario, with GDP about 8 percent higher by 2030 and cognitive employment about 4 percent lower.
Methodological or theoretical issues flagged for evaluation
Evaluation would need economists comfortable with task-based automation, macro growth modeling, search-and-matching labor models, survey design, and AI capability forecasting. The hardest issue is not checking a single empirical estimate but assessing whether the scenario assumptions, calibration choices, omitted mechanisms, and interpretation are appropriate for policy use; reviewers should separate model-internal validity from whether the scenarios deserve decision weight.
Opus 5.5 re-evaluation (experimental)
Read with care. This is an experimental re-evaluation by a different model (Claude Opus 5.5, 2026-10-01). Across a calibration sample Opus scored about 15 points lower than GPT-5.5, so 15 points are added to compare it with GPT-5.5 scores. The calibration is a small sample (n=44) and is not human-validated, and the two models disagree about the order of papers within the top group. The AI score at the top of this page is the usual evaluation-priority score shifted by the difference between the adjusted Opus score and the earlier holistic score; the synthesis uses that shifted value, and the unshifted score is shown beside it. On the dashboard you can switch the Opus scores off. Read the methods note.
Opus rationale (AI-generated)
This is a working paper from Anthropic's Economic Research team, co-authored with two of the most prominent economists working on AI and growth (Chad Jones and Anton Korinek). It sets out a tractable macro-labor framework that maps AI capability paths into GDP, the labor share, wages, reallocation and unemployment over 2026–2030, and anchors it with a survey of US adults. It is squarely within The Unjournal's 'Emerging technologies (focus: AI)' area and is quantitative social science with direct policy relevance. Likely users are fiscal and labor-policy bodies (CBO, Treasury, DOL, Federal Reserve staff), international organisations (IMF, OECD, World Bank, ILO), AI-policy think tanks (GovAI, Brookings, CSET), and funders weighing AI-economy preparedness (Open Philanthropy and others). The headline numbers (15 percent growth, labor share falling to 45 percent, nearly one in five cognitive workers unemployed in the extreme case) will be widely quoted and easily misread as forecasts. It is an unrefereed report from an AI developer, so independent scrutiny of the calibration, the reallocation/search frictions, and how the scenarios are framed would add clear value. Concerns: the outputs depend heavily on the assumed capability paths, the model is US/closed-economy, and the survey component is descriptive. Evaluators should be asked whether the framework adds decision information beyond existing AI-macro models (Acemoglu, Epoch GATE, Korinek-Suh) or mostly repackages assumptions. One more note: this assessment was produced by an Anthropic model about an Anthropic co-authored paper; human assessors should check the framing.
Dashboard details and provenance
Full AI dashboard scoring rationale
This paper from The Anthropic Institute is squarely in Unjournal's wheelhouse: quantitative economic modeling of AI impacts with direct relevance to labor-market preparation, fiscal redistribution, public communication, and AI-governance planning. Organizations such as OECD, ILO, US Treasury, CEA, UK DSIT, EU AI Office, CSET, GovAI, Open Philanthropy, and AI-safety policy teams could use or cite this framework, and it is a September 2026 working paper rather than a peer-reviewed publication. The main concern is that it is scenario modeling rather than a validated forecast or causal estimate, and it omits catastrophic risk and much political economy, but that is also why independent evaluation would add clear value before the paper becomes a policy reference point.
AI decision-relevance rationale
The research has a strong but not exceptional global-welfare/VoI case: it addresses potentially large human welfare stakes from AI-driven productivity growth, labor displacement, unemployment, inequality, and institutional adaptation, but its direct scope is mainly US macro-labor outcomes through 2030. Its incremental contribution is a structured model linking AI capability/adoption assumptions to economic outcomes, which could improve planning for redistribution, labor-market policy, insurance, and governance triggers; however, it does not assign probabilities, directly estimate causal effects, or model LMIC welfare, animal welfare, or catastrophic-risk channels. The welfare value depends heavily on uptake and interpretation: useful as a scenario framework, weaker as decisive evidence for pacing, export controls, or compute governance.
AI timing assessment
Timing value is very high: this is a September 2026 working paper from The Anthropic Institute and appears to be at a stage where independent critique could still affect revisions and downstream policy use. The supplied targeted public-scrutiny evidence is empty, so I cannot verify whether substantial independent expert review already exists; I therefore treat prior scrutiny as unclear rather than absent. Given the paper's prominence and likely policy salience, even limited prior scrutiny would leave a large marginal role for an Unjournal-style evaluation unless there is already a substantial public expert debate not captured here.
Intake, review, and crux connections
AI governance and the economics of AI policy: priority papers
· 2026-09-30
Papers identified in The Unjournal's September 2026 AI-governance scoping (David Reinstein's internal planning). On integration (2026-09-30), titles, authors, dates and abstracts were re-resolved from canonical sources (arXiv, Crossref, NBER, or the publisher's own page), never from the scoping notes; the papers were deduplicated against the dashboard and scored by the standard Codex GPT-5.5 (medium reasoning) subscription path. Five related papers already on the dashboard but still awaiting a genuine model score were re-scored by the same path and are labeled as surfaced existing records
The papers were identified in The Unjournal's September 2026 AI-governance scoping (David Reinstein's internal planning). Several are central to live policy debates but were missing from the dashboard or not yet scored. Inclusion is not an endorsement or a completed Unjournal team decision.
Public attention and use
The paper received early public coverage. An independent Econ-ARK reimplementation now reports reproducing all 226 checkable published numbers after recovering twelve omitted or rounded implementation details. This checks implementation rather than establishing the plausibility of the scenarios or economic assumptions; a new evaluation should concentrate on those substantive questions.
Publisher page listing / discoverability
Anthropic published the paper with an interactive scenario explorer. The publisher is also the authors' employer.
Source: Anthropic · Relationship: author institution
Independent media visible circulation
A technology outlet summarised the three scenarios. It reports rather than evaluates the model.
Source: The Next Web · Relationship: independent media
Commentary public mention
A university business-school site discusses the scenarios and their range for jobs and GDP by 2030. We did not check authorship or relationship to the paper.
Source: Darden School of Business · Relationship: relationship not checked
Independent computational reproduction substantive scrutiny
The reimplementation reports matching every checkable published number and the explorer paths after recovering omitted or rounded details. It validates the implementation, while leaving the scenario assumptions, empirical calibration, and external validity for substantive evaluation.
Source: Econ-ARK REMARK · Relationship: independent reimplementation
What the search did not establish
- No exact-title EA Forum or LessWrong discussion surfaced in the targeted search.
- No government or central-bank document using the scenarios was found.
- Substantive checks of scenario assumptions, labor reallocation, and external validity remain useful; the independent reproduction already addresses implementation correctness.
Targeted search checked 2026-10-05. Search scope: Targeted exact-title searches across the open web, institutional and author pages, news and public social-media results, EA Forum/LessWrong, and policy/white-paper contexts. Evidence records distinguish commissioning or report use from independent discussion, media attention, indexing, and post-publication policy use. A search miss is reported as uncertainty, not proof of absence. Entries with check_depth "quick" rest on roughly one to two searches and should be read as especially uncertain; entries checked on 2026-10-01 were added for the shortlist of papers without human ratings.