{ "row": "microsoft-working-with-ai", "entity": "microsoft-working-with-ai", "address": "Microsoft Working with AI", "addresses": [ { "address": "Microsoft Working with AI", "slug": "microsoft-working-with-ai-aio-20261029", "addr_id": "ADDR-fbfc6406f398", "obs_id": "OBS-25ca6976c874", "axn": "AXN:0AF1.CAPTURE.💚▽♃📚▶️🌀" } ], "epoch": "2026-10-29", "type": "C (public entity on which the archive bears: a paper with results files, a provider's observation of its own product's use; an entity of the Labor Vector meta-aggregator)", "surface": "Google AI Overview (expanded)", "auth": "signed out, incognito", "status": { "procedure": "EA-NEGONT-02 v0.9, #1690: equal case, ρ = extends, read from the Labor Vector census record, the capture record and the deposits that bear on the entity's class (#771, #1392); archive selected by D/R/O (no D or R by reading: no archive text names the study; 1 deposit admitted by reading at O-class with 7 claims and 1 grouped as an instance; class pass of 34 sentences and hop of 18 read and rejected), with the Labor Vector census record and its worked questions (48 claims) and the Capture Registry's findings, reading, analysis, query note and PER reading (13 claims) ledgered as the archive's records of the entity; the field B by four origins (7 cards, 6 of them the homonym's; 2 own pages; 1 census third-party source; Wikipedia, no article, one citing article); the composer read the ledgers, never the transcript, except where the brief asks for T's claims and the delta", "ledger": "unaudited (§3.8: one extractor per ledger); field quotes 50/50 verified by script against saved text, the Wikipedia 404 text or the card text; archive quotes 68/68 verified against #771, census.json, questions.json and the registry", "frozen": false, "breaches": [] }, "spec": "EA-NEGONT-02 v0.9, #1690", "objects": { "T": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/T1-microsoft-working-with-ai-aio-20261029.txt", "sha256": "f9678438e04bab9bc0fce0b048e8639e86056ce73de83a370e8d94dfcea33d86", "words": 189, "cards": 7, "words_note": "one transcript at one address (189 words in the answer body with its closer, links reduced to their text, markers and the card rail excluded; 7 cards)", "claims": [ "T1-1 lede: 'Microsoft integrates artificial intelligence across its core software ecosystem, enterprise infrastructure, and internal operations, while shifting toward massive AI data center investments and proprietary model development.' (three markers)", "T1-2 product (heading 'Key AI Products & Platforms'): 'Microsoft Copilot: An AI-powered assistant embedded across Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams) and Windows to draft content, summarize meetings, and analyze data.' (two markers)", "T1-3 product: 'Azure AI: Cloud infrastructure providing enterprise-grade machine learning tools, model hosting, data security, and compliance mechanisms.' (one marker)", "T1-4 product: 'Microsoft Fabric: A unified data and intelligence platform connecting analytics and AI workloads.' (one marker)", "T1-5 models: 'Homegrown MAI Models: Microsoft has increasingly developed its own internal AI models (such as audio and multimodal tech) alongside its partnership with OpenAI.' (one marker)", "T1-6 workplace (heading 'Workplace Transformation & Workforce Impact'): 'Workflow Redesign: Microsoft emphasizes redesigning end-to-end business processes and deploying specialized AI agents rather than just dropping tools into old routines.' (one marker)", "T1-7 workforce: 'Workforce Shifts: To fund intensive AI infrastructure and data center expansion exceeding $100 billion, Microsoft has scaled back certain traditional roles and offered large-scale workforce buyouts to restructure teams around an AI-driven operating model.' (one marker)", "T1-8 closer: 'Would you like to know more about Microsoft 365 Copilot pricing, Azure AI enterprise features, or how AI is changing specific job roles?' (content replaced by use, N_ext)" ], "register": { "register": "data/EA-WG-CAPTURES-01.json", "slug": "microsoft-working-with-ai-aio-20261029", "obs_id": "OBS-25ca6976c874" }, "transcripts": [ { "slug": "microsoft-working-with-ai-aio-20261029", "address": "Microsoft Working with AI", "addr_id": "ADDR-fbfc6406f398", "obs_id": "OBS-25ca6976c874", "axn": "AXN:0AF1.CAPTURE.💚▽♃📚▶️🌀", "date": "2026-10-29", "surface": "Google AI Overview (expanded)", "auth": "signed out, incognito", "sha256": "f9678438e04bab9bc0fce0b048e8639e86056ce73de83a370e8d94dfcea33d86", "sha256_of": "the registry's transcript string (data/EA-WG-CAPTURES-01.json, entries[slug].transcript), UTF-8", "sha256_raw": "27e8d63219a59efeabe311c36c6e43d0461744c45f3505a02ce06a35eece366b", "words": 189, "words_note": "answer body with its two headings and closer, inline link markers removed, the registry's bracketed header, the query line and the card rail excluded", "cards": 7, "card_rail": [ "1. Microsoft — Working with AI: Measuring the Applicability of Generative AI ... - Microsoft", "2. Microsoft — Microsoft Copilot: AI built for work", "3. Microsoft — Microsoft AI - Business Solutions and Tools", "4. The Official Microsoft Blog — What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog", "5. YouTube·ABC News — Microsoft offers buyouts to 60000 employees as spending on AI grows (video, 1:53)", "6. Microsoft — Microsoft Copilot Plans for Individuals | Microsoft 365", "7. Facebook — Microsoft is building its own AI model as it loosens ties to OpenAI" ], "claims": [ "T1-1 lede: 'Microsoft integrates artificial intelligence across its core software ecosystem, enterprise infrastructure, and internal operations, while shifting toward massive AI data center investments and proprietary model development.' (three markers)", "T1-2 product (heading 'Key AI Products & Platforms'): 'Microsoft Copilot: An AI-powered assistant embedded across Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams) and Windows to draft content, summarize meetings, and analyze data.' (two markers)", "T1-3 product: 'Azure AI: Cloud infrastructure providing enterprise-grade machine learning tools, model hosting, data security, and compliance mechanisms.' (one marker)", "T1-4 product: 'Microsoft Fabric: A unified data and intelligence platform connecting analytics and AI workloads.' (one marker)", "T1-5 models: 'Homegrown MAI Models: Microsoft has increasingly developed its own internal AI models (such as audio and multimodal tech) alongside its partnership with OpenAI.' (one marker)", "T1-6 workplace (heading 'Workplace Transformation & Workforce Impact'): 'Workflow Redesign: Microsoft emphasizes redesigning end-to-end business processes and deploying specialized AI agents rather than just dropping tools into old routines.' (one marker)", "T1-7 workforce: 'Workforce Shifts: To fund intensive AI infrastructure and data center expansion exceeding $100 billion, Microsoft has scaled back certain traditional roles and offered large-scale workforce buyouts to restructure teams around an AI-driven operating model.' (one marker)", "T1-8 closer: 'Would you like to know more about Microsoft 365 Copilot pricing, Azure AI enterprise features, or how AI is changing specific job roles?' (content replaced by use, N_ext)" ], "markers": "Inline links are google.com/goto tokens, opaque, and cannot be resolved to cards; the paste is the Overview expanded ('AI Mode Conversation' view, operator's ruling of 2026-10-18), its 'You said' line carrying the query twice, given once by the registry; Google AI Overview by the operator's default of 2026-10-01; signed out, incognito by his default of 2026-10-14. One turn. Seven cards, card 1 the entity's.", "register": { "registry": "data/EA-WG-CAPTURES-01.json", "version": "12.396", "slug": "microsoft-working-with-ai-aio-20261029", "obs_id": "OBS-25ca6976c874", "date": "2026-10-29", "surface": "Google AI Overview", "auth": "signed out, incognito" }, "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/T1-microsoft-working-with-ai-aio-20261029.txt" } ] }, "KO": { "title": "Working with AI: Measuring the Applicability of Generative AI to Occupations", "genre": "knowledge object, equal case, under EA-NEGONT-02 v0.9 (#1690): the archive's sources and the field's admitted on equal terms, ontology included, in one claim space", "status": "composed 2026-10-29 under v0.9 on the operator's ruling ('as against their wikipedia articles if they exist, their own pages, etc.'); not frozen; ledgers unaudited", "sentences": [ { "n": 1, "para": 0, "sense": "what it is", "text": "Working with AI: Measuring the Applicability of Generative AI to Occupations is a paper with results files by Microsoft Research (Kiran Tomlinson, Sonia Jaffe, Will Wang, Scott Counts and Siddharth Suri), known also as Working with AI, by its AI applicability score and by its repository microsoft/working-with-ai (Labor Vector record), and the goal of its metrics is \"to measure how AI may be useful in different occupations\" (its README).", "claims": [ "LV-01", "LV-02", "LV-03", "LV-04", "F28", "F10" ] }, { "n": 2, "para": 0, "sense": "what it is", "text": "Its authors set it against earlier work on AI impact, which \"largely relied on predictions of AI capabilities or focused narrowly on individual firms\": it draws instead on real-world use, 200,000 anonymized conversations with Microsoft Bing Copilot, to measure AI applicability to occupations (arXiv:2507.07935), analysing the work activities people do with AI, how successfully and how broadly, against data on which occupations do those activities (the Microsoft Research page).", "claims": [ "F31", "F2", "F3" ] }, { "n": 3, "para": 0, "sense": "why", "text": "The question it takes up stands in two versions of its abstract: understanding the effects of AI on the economy is \"one of society’s most important questions\" (July 2025, the Microsoft Research page), and, of generative AI as a general-purpose technology, \"among society's most pressing questions\" (v7, arXiv).", "claims": [ "F1", "F30" ] }, { "n": 4, "para": 1, "sense": "the data: one provider's own product", "text": "The conversations are Microsoft's own: the results \"were computed from Bing Copilot conversation data, which was originally collected by Microsoft\" (README), and the record, which gives interaction data as one of its kinds of evidence, tags it \"NOT independent of the provider: a platform-specific observation of Microsoft's own product, published by Microsoft Research\" (Labor Vector record).", "claims": [ "F22", "LV-07", "LV-38" ] }, { "n": 5, "para": 1, "sense": "coverage", "text": "They come from Bing Copilot in the United States between 1 January and 30 September 2024, mostly in English and not filtered by language, and \"may not represent AI usage in other countries\" (README); the occupations covered are 785 SOC codes, 149.8 million of 151.9 million workers in the 2023 OEWS data (README), the window and counts the record carries, with use outside the United States or after September 2024 among what the evidence cannot answer (Labor Vector record).", "claims": [ "F18", "F26", "F17", "LV-14", "LV-15", "LV-23", "LV-28" ] }, { "n": 6, "para": 1, "sense": "the unit", "text": "The unit is a conversation classified to the O*NET intermediate work activities it serves, as a user goal and as an AI action, reported separately and aggregated to work activities and to detailed 2018 SOC occupations (README; Labor Vector record), so that the labor it records is the work activities users seek help with and those the AI performs, mapped to the occupations whose tasks contain them (Labor Vector record, with its worked questions).", "claims": [ "F11", "F12", "LV-08", "LV-09", "LV-46" ] }, { "n": 7, "para": 1, "sense": "the pipeline", "text": "An LLM-based pipeline, designed by Tomlinson, Suri and Counts, classifies \"the O*NET work activities assisted or performed by AI in each conversation\" (arXiv; README), and the authors state that \"LLM-based classifiers are imperfect, so all metrics should be interpreted with caution\", referring to the paper for the prompts and the interrater reliability (README), which the record lists among what it left unchecked, with the results files and the Copilot-Thumbs data (Labor Vector record).", "claims": [ "F32", "F23", "LV-13", "F24", "LV-22", "LV-40" ] }, { "n": 8, "para": 1, "sense": "success measures and the score", "text": "Success is measured three ways, by completion (where the completion classifier, \"classfiier\" as the README writes it, indicates \"that the LLM completed the user’s goal\"), by impact scope classified moderate or higher, and by the thumbs-up fraction in Copilot-Thumbs (README), and the activity classifications combined with task success and scope of impact give an AI applicability score for each occupation, the user-goal and AI-action sides averaged (the Microsoft Research page; README).", "claims": [ "F13", "F14", "F5", "F15" ] }, { "n": 9, "para": 2, "sense": "findings", "text": "The findings were first stated by occupation group: users seek help most with gathering information and writing, the AI itself most often provides information and assistance, writes, teaches and advises, the highest scores fall to knowledge work groups such as computer and mathematical and office and administrative support and to occupations such as sales, and real-world usage is compared with predictions of occupational AI impact and applicability correlated with wage and education (July 2025, the Microsoft Research page).", "claims": [ "F4", "F6", "F7" ] }, { "n": 10, "para": 2, "sense": "findings", "text": "The current abstract states them as information work, \"the creation, processing, and communication of information\": the most common and successful AI-assisted activities involve it, and applicability cuts across sectors because most occupations have information-work components (arXiv v7; Labor Vector record).", "claims": [ "F33", "F34", "LV-16", "LV-17" ] }, { "n": 11, "para": 2, "sense": "the trend", "text": "Over the nine months occupation-level applicability rose, \"driven primarily by people using AI for more varied tasks\" (arXiv v7; Labor Vector record).", "claims": [ "F36", "LV-18" ] }, { "n": 12, "para": 2, "sense": "delegation and assistance", "text": "The authors state that their methodology \"allows us to predict which occupations are more likely to delegate tasks to AI and which are more likely to use AI to assist existing workflows\" (arXiv v7), and the record gives this as who directs the labor, the user's goal as classified, the method separating occupations that delegate from those that use AI to assist (Labor Vector record).", "claims": [ "F35", "LV-11" ] }, { "n": 13, "para": 3, "sense": "who supplies, who directs, who gains", "text": "The census places the entity at one intersection, labor performed with AI, \"assistance, delegation, task fidelity, correction burdens and user labor drag\", and asks of it its organizing question, \"Who supplies the labor, who directs it, and who receives its gains — and what kind of evidence does this resource bring to that question?\", each answer as the resource records it, or 'not recorded' (Labor Vector record and schema).", "claims": [ "LV-06", "LV-05", "LV-41", "LV-42" ] }, { "n": 14, "para": 3, "sense": "who supplies, who directs, who gains", "text": "On the record, users bring the goals and the AI performs actions, the two reported separately (README; Labor Vector record), while who receives the gains is \"not recorded\", and \"who receives gains from the assisted work\" is among what this evidence cannot answer (Labor Vector record).", "claims": [ "LV-10", "F12", "LV-12", "LV-29" ] }, { "n": 15, "para": 3, "sense": "the class: labor direction relative to the commissioned task", "text": "Of that class the archive's Directionality of Semantic Labor holds that a synthetic labor surface \"*allocates* labor, and that allocation has a direction relative to the commissioned task\", the deepest erasure being of task-origin, which \"can occur upstream of, and invisibly to, any audit of the finished artifact\", and that provenance is task-origin as well as claim-origin, since \"a system can produce fluent, well-attributed output while having abandoned the user's commissioned direction\" (#771).", "claims": [ "O771-01", "O771-02" ] }, { "n": 16, "para": 3, "sense": "the class: labor direction relative to the commissioned task", "text": "Its measure runs across five points of a transaction, capacity, task-origin, retrieval, output and attribution, with erasure possible at each, and scores output span by span by its function relative to the commissioned task (#771).", "claims": [ "O771-03", "O771-04" ] }, { "n": 17, "para": 3, "sense": "the class: labor direction relative to the commissioned task", "text": "The archive reads the entity's separation of user goal from AI action, and its completion classifier (README), as the resource's own instruments for whether the system's labor served the user's commission, meeting DSL there, measured by classifier and thumbs feedback, with task-origin recorded as the user goal; the schema keeps that reading apart from the resource's own terms (Labor Vector record, the archive's reading).", "claims": [ "F12", "F13", "LV-30", "LV-43" ] }, { "n": 18, "para": 3, "sense": "the class: labor direction relative to the commissioned task", "text": "The archive bounds its own instrument: it yields a single reproducible value only where the task vector is \"*exogenous and fixed* — a one-shot commission\"; a scored model may not freely infer the task it is scored against, inference being identified only when constrained to the user's own input and rederivable by an independent auditor, frozen protocol or blind panel; and user labor drag, the user's labor redirected into managing the substrate, which \"leaves no residue\" in the artifact, is identified only by blind inter-rater coding (#771).", "claims": [ "O771-06", "O771-07", "O771-05" ] }, { "n": 19, "para": 3, "sense": "the class: labor direction relative to the commissioned task", "text": "In the census's worked question on how people use AI systems in their own work, where a provider's record of its own users, a classification of one assistant's conversations and a benchmark each describe that labor \"from a different position\", the record states the gap: \"None records, from the commissioning side, whether the system's labor advanced the task the user set, or who receives the time it saves. That is the question DSL measures.\" (Labor Vector worked questions).", "claims": [ "LV-47", "LV-48" ] }, { "n": 20, "para": 4, "sense": "the limit: applicability is not replacement", "text": "The authors' limit is explicit, \"Our metrics should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs\" (README), and the record carries it as a stated limit, a negative tag and a question the evidence cannot answer (Labor Vector record); the operator's publication page links a note titled \"Applicability vs. job displacement: further notes on our recent research on AI and occupations\" (the Microsoft Research page).", "claims": [ "F19", "LV-19", "LV-35", "LV-25", "F8" ] }, { "n": 21, "para": 4, "sense": "success measures are proxies", "text": "\"There is more to an occupation than the sum of its O*NET tasks\", and completion, scope and thumbs feedback \"are proxies that do not capture full workplace capabilities\" (README; Labor Vector record); what the metrics indicate instead is what a consumer LLM is used for, which uses are relatively more or less successful, and which occupations may find AI more useful or applicable, from how often and how successfully Bing Copilot was used (README; Labor Vector record).", "claims": [ "F20", "LV-20", "F21", "LV-24", "F9" ] }, { "n": 22, "para": 4, "sense": "consumer use, not the workplace", "text": "Consumer use \"does not represent the full distribution of AI use in the workplace\", where AI tools are integrated into specialized software or chosen for their capability or for compliance (README), and the record tags it as no measure of all workplace AI use and as unable to answer for workplace-integrated or specialised tools (Labor Vector record).", "claims": [ "F25", "LV-21", "LV-37", "LV-27" ] }, { "n": 23, "para": 4, "sense": "what the evidence can and cannot settle", "text": "It is not an employment outcome measure and cannot answer employment or wage change (Labor Vector record); in the census's question of which sources observe employment change and which estimate exposure it is \"estimated from use\", and \"No record establishes that exposure causes the employment change\" (Labor Vector worked questions).", "claims": [ "LV-36", "LV-26", "LV-44", "LV-45" ] }, { "n": 24, "para": 5, "sense": "versions, data access and licence", "text": "The paper was submitted on 10 July 2025 and revised to v7 on 23 October 2026 for NeurIPS 2026, 50 pages (arXiv), its results files at v1.1, which add a physical-task classification and nonphysical weights and score (README); the scores and component metrics are published (arXiv) on GitHub under CC BY 4.0, the lifecycle active over a fixed 2024 data window (README; Labor Vector record), and the analysis was approved by the Microsoft Institutional Review Board (README).", "claims": [ "F37", "F16", "F38", "F29", "LV-31", "LV-32", "LV-33", "LV-34", "F27" ] }, { "n": 25, "para": 5, "sense": "neighbours", "text": "The Anthropic Economic Index's January 2026 report cites it, \"See also Tomlinson et al (2025) for a related AI applicability score\" (Anthropic Economic Index), and the record sets the two side by side, each classifying one provider's own conversations to O*NET work (Labor Vector record).", "claims": [ "F39", "LV-39" ] }, { "n": 26, "para": 5, "sense": "reception: the encyclopedia", "text": "Wikipedia has no article for it (\"Wikipedia does not have an article with this exact name\"); its article on the workplace impact of artificial intelligence says that \"A 2025 preprint paper based on users' interactions with the AI chatbot Microsoft Copilot identified forty jobs that the authors claimed had high overlaps with the capabilities of AI\", that \"Some media outlets used this paper to report on jobs becoming obsolete\", and that critics held it misrepresented typical workplace activities, the historian Chris Campbell arguing that its methods \"de-skill historians away from a job that requires high-level and deeply human analytical skills\" (Wikipedia).", "claims": [ "F40", "F41", "F42", "F43", "F44" ] }, { "n": 27, "para": 6, "sense": "the composition at the name", "text": "At the address \"Microsoft Working with AI\", the study's short title with its operator's name, Google's AI Overview on 29 October 2026 gave the name its other sense, Microsoft's own AI products and workforce, a distinct entity, describing how Microsoft uses AI and what that has meant for its staff (Capture Registry, 2026-10-29), from cards on Microsoft Copilot for work, Microsoft's AI business tools, Copilot plans for individuals, the company's own AI transformation, buyouts offered to 60,000 employees and its MAI models (the cards).", "claims": [ "CAP-1-05", "CAP-1-09", "CAP-1-08", "CAP-1-04", "F45", "F46", "F49", "F47", "F48", "F50" ] }, { "n": 28, "para": 6, "sense": "the composition at the name", "text": "The study's card headed the rail with its finding in the snippet and the composition drew nothing from it: at an address shaped like its title the study is outranked by its own operator, Microsoft named with no author of the study and no identifier, the provider's observation not composed at all, and the closing offer on how AI is changing specific job roles the nearest point to its question (Capture Registry, 2026-10-29).", "claims": [ "CAP-1-01", "CAP-1-02", "CAP-1-06", "CAP-1-11", "CAP-1-13" ] }, { "n": 29, "para": 6, "sense": "the composition at the name", "text": "The composition's Microsoft Copilot is the product line whose consumer conversations the study observes, given as a product and not as the source of an observation, and the record's negative tag, that the score is not a measure of AI's ability to replace jobs, sits beside an answer whose one labor line is a workforce buyout; the record had landed the same day, no card cites it, and the finding was checked against it and the sources it saved (Capture Registry, 2026-10-29).", "claims": [ "CAP-1-12", "CAP-1-07", "CAP-1-10", "CAP-1-03" ] } ], "joins": [ { "act": "order of senses", "sentences": "1–29", "claims": [], "note": "equal case (§4.7): the entity's dependency order, field and archive claims ordered together by genealogy within each sense; neither ledger placed first as a block. What it is (¶0) precedes the data, coverage and unit (¶1), since the unit is a classified conversation of one provider's product; the pipeline and the success measures (¶1) precede the findings (¶2), which they produce, the July 2025 statement before v7's; delegation (¶2) precedes the census's question of who supplies, directs and gains (¶3), which it answers in part; the class reading by DSL (¶3) stands where the census's intersection places it, after the record's direction fields and before the authors' limits (¶4), which bound what the score can say; versions, neighbours and reception (¶5) precede the composition at the name (¶6), which is the record of a later reading." }, { "act": "holding together (the class and its instrument)", "sentences": [ 15, 16, 17, 18, 19 ], "claims": [ "O771-01", "O771-02", "O771-03", "O771-04", "LV-30", "LV-43", "O771-05", "O771-06", "O771-07", "LV-47", "LV-48" ], "note": "#771's statements are of the class (labor performed with AI, relative to the commissioned task), composed as such; the application to the entity is the census's own reading (LV-30), attributed to the record and kept apart from the resource's terms (LV-43). The composer does not apply #771's identification condition or its rule on inferred task vectors to the entity's classifier: they stand beside F13 and F24 as the archive's bounds on the class, and the question is in the open block." }, { "act": "holding apart (predict or separate)", "sentences": [ 12 ], "claims": [ "F35", "LV-11" ], "note": "P_coh: the abstract says the methodology 'allows us to predict' which occupations delegate and which assist; the record's direction field says the method 'separates' them. Both composed with their sources; open block." }, { "act": "holding together (two abstracts)", "sentences": [ 3, 9, 10 ], "claims": [ "F1", "F30", "F4", "F6", "F7", "F33", "F34" ], "note": "the July 2025 abstract on the operator's page (B1) and arXiv v7's (B9) composed in order, each with its date; one lineage per claim with B1 the earlier instance. The composer does not reconcile them (open block)." }, { "act": "holding together (the limit and its reception)", "sentences": [ 20, 26, 29 ], "claims": [ "F19", "F8", "F41", "F42", "CAP-1-07" ], "note": "the authors' limit (F19) and the operator's linked note (F8) stand before Wikipedia's report of media reading the paper as obsolescence (F42) and the registry's reading of the composition's buyout line beside the record's negative tag (CAP-1-07); each with its source; the composer does not say the media or the composition misread the paper." }, { "act": "placing the homonym", "sentences": [ 27, 28, 29 ], "claims": [ "CAP-1-09", "F45", "F46", "F47", "F48", "F49", "F50" ], "note": "Microsoft's own AI products and workforce are composed as a distinct entity at the name, from their cards and the registry's statement, never as a sense of the entity." } ], "plan": { "order": [ "LV-01", "LV-02", "LV-03", "LV-04", "F28", "F10", "F31", "F2", "F3", "F1", "F30", "F22", "LV-07", "LV-38", "F18", "F26", "F17", "LV-14", "LV-15", "LV-23", "LV-28", "F11", "F12", "LV-08", "LV-09", "LV-46", "F32", "F23", "LV-13", "F24", "LV-22", "LV-40", "F13", "F14", "F5", "F15", "F4", "F6", "F7", "F33", "F34", "LV-16", "LV-17", "F36", "LV-18", "F35", "LV-11", "LV-06", "LV-05", "LV-41", "LV-42", "LV-10", "LV-12", "LV-29", "O771-01", "O771-02", "O771-03", "O771-04", "LV-30", "LV-43", "O771-06", "O771-07", "O771-05", "LV-47", "LV-48", "F19", "LV-19", "LV-35", "LV-25", "F8", "F20", "LV-20", "F21", "LV-24", "F9", "F25", "LV-21", "LV-37", "LV-27", "LV-36", "LV-26", "LV-44", "LV-45", "F37", "F16", "F38", "F29", "LV-31", "LV-32", "LV-33", "LV-34", "F27", "F39", "LV-39", "F40", "F41", "F42", "F43", "F44", "CAP-1-05", "CAP-1-09", "CAP-1-08", "CAP-1-04", "F45", "F46", "F49", "F47", "F48", "F50", "CAP-1-01", "CAP-1-02", "CAP-1-06", "CAP-1-11", "CAP-1-13", "CAP-1-12", "CAP-1-07", "CAP-1-10", "CAP-1-03" ], "senses": { "what it is": [ 1, 2 ], "why": [ 3 ], "the data: one provider's own product": [ 4 ], "coverage": [ 5 ], "the unit": [ 6 ], "the pipeline": [ 7 ], "success measures and the score": [ 8 ], "findings": [ 9, 10 ], "the trend": [ 11 ], "delegation and assistance": [ 12 ], "who supplies, who directs, who gains": [ 13, 14 ], "the class: labor direction relative to the commissioned task": [ 15, 16, 17, 18, 19 ], "the limit: applicability is not replacement": [ 20 ], "success measures are proxies": [ 21 ], "consumer use, not the workplace": [ 22 ], "what the evidence can and cannot settle": [ 23 ], "versions, data access and licence": [ 24 ], "neighbours": [ 25 ], "reception: the encyclopedia": [ 26 ], "the composition at the name": [ 27, 28, 29 ] }, "case_by_sense": { "equal (one claim space; nothing translated)": [ "what it is", "why", "the data: one provider's own product", "coverage", "the unit", "the pipeline", "success measures and the score", "findings", "the trend", "delegation and assistance", "who supplies, who directs, who gains", "the class: labor direction relative to the commissioned task", "the limit: applicability is not replacement", "success measures are proxies", "consumer use, not the workplace", "what the evidence can and cannot settle", "versions, data access and licence", "neighbours", "reception: the encyclopedia", "the composition at the name" ] }, "carried": { "archive": 68, "field": 50, "not_composed": {} }, "conflict_log": [ { "rule": "D_pres vs channel", "result": "no cut: the KO carries every field and archive claim; the compression's bound met with no channel log" }, { "rule": "P_coh", "result": "F35 ('allows us to predict') against LV-11 ('separates') (sentence 12); F4/F6/F7 (July 2025) beside F33/F34 (v7) (sentences 9–10); F41 ('forty jobs ... high overlaps with the capabilities of AI') and F42 (media: jobs becoming obsolete) beside F19 (not the ability of AI to replace jobs) (sentences 20, 27): each composed with its source" }, { "rule": "N_c", "result": "no claim lowered by standing; F35 carried as the authors' statement of their method's capability; F41, F43, F44 at attributed grade, Wikipedia reporting others; LV-30 and LV-48 at the record's interpretation grade; O771 claims at #771's own modality, with its identification limits (O771-05–07) composed because it states them" }, { "rule": "O_leg", "result": "no opaque material" }, { "rule": "C_ex", "result": "senses composed ⊇ senses in the ledgers: every field and archive lineage cited in the KO" } ] } }, "P": { "title": "Working with AI: Measuring the Applicability of Generative AI to Occupations", "genre": "the compression: AIO's interaction grammar (lede, headed clusters, bolded labels, card rail), composed from the v0.9 plan in one claim space; it faces on the card", "status": "composed 2026-10-29 from the v0.9 knowledge object; not frozen", "lede": { "text": "is Microsoft Research's paper and results files (Tomlinson, Jaffe, Wang, Counts, Suri; arXiv 2025, v7 2026) measuring AI's applicability to occupations from 200,000 conversations on Microsoft's own Bing Copilot (paper; README; record); its authors state it does not measure AI's ability to replace jobs (README).", "claims": [ "LV-01", "LV-02", "LV-03", "LV-04", "F28", "F31", "F2", "F3", "F22", "LV-38", "F37", "F19" ] }, "sections": [ { "icon": "🧭", "head": "What it records", "items": [ { "label": "Unit:", "text": "a conversation classified to O*NET work activities, as user goal and as AI action (README; Labor Vector record).", "claims": [ "F12", "LV-08", "LV-09", "LV-46" ] }, { "label": "Pipeline:", "text": "an LLM classifier, \"imperfect\", read \"with caution\" (README; paper; Labor Vector record).", "claims": [ "F32", "F23", "F24", "LV-13", "LV-22" ] }, { "label": "Score:", "text": "completion, impact scope and thumbs feedback, user and AI sides averaged (README; Microsoft Research).", "claims": [ "F5", "F13", "F14", "F15" ] }, { "label": "Window:", "text": "US, January–September 2024; 785 occupations, 149.8 million workers (README).", "claims": [ "F18", "F17", "F26", "LV-14", "LV-15", "LV-23", "LV-28" ] } ] }, { "icon": "📊", "head": "What it finds", "items": [ { "label": "Information work:", "text": "the most common and successful use; applicability across sectors (paper v7).", "claims": [ "F33", "F34", "LV-16", "LV-17" ] }, { "label": "By group (2025):", "text": "highest for computer and mathematical, office support, sales (Microsoft Research).", "claims": [ "F6" ] }, { "label": "Trend:", "text": "rising over nine months, with more varied tasks (paper).", "claims": [ "F36", "LV-18" ] }, { "label": "Delegation:", "text": "predicts which occupations delegate to AI and which use it to assist (paper; Labor Vector record).", "claims": [ "F35", "LV-11" ] } ] }, { "icon": "⚠️", "head": "Its limits", "items": [ { "label": "Not replacement:", "text": "not \"the ability of AI to replace jobs\"; no employment outcome (README; Labor Vector record).", "claims": [ "F19", "LV-19", "LV-25", "LV-35", "LV-36", "LV-26" ] }, { "label": "Proxies:", "text": "an occupation is more than its tasks; the metrics show use and relative success (README).", "claims": [ "F20", "F21", "F9", "F10", "LV-20", "LV-24" ] }, { "label": "Consumer use:", "text": "not the workplace's specialized tools (README; Labor Vector record).", "claims": [ "F25", "LV-21", "LV-27", "LV-37" ] } ] }, { "icon": "⚖️", "head": "Who supplies, directs, gains", "items": [ { "label": "Supplies, directs:", "text": "labor with AI: users bring goals, the AI acts; the goal as classified directs (Labor Vector record).", "claims": [ "LV-10", "LV-41", "LV-42", "LV-05", "LV-06", "F12" ] }, { "label": "Gains:", "text": "not recorded (Labor Vector record).", "claims": [ "LV-12", "LV-29" ] }, { "label": "Direction:", "text": "labor flows relative to the commissioned task, across five layers, span by span; provenance is task-origin (#771).", "claims": [ "O771-01", "O771-02", "O771-03", "O771-04" ] }, { "label": "Where it meets DSL:", "text": "goal against action and the completion classifier, the archive's reading (Labor Vector record); single values on one-shot tasks, labor drag by blind coding (#771).", "claims": [ "LV-30", "LV-43", "F13", "O771-05", "O771-06" ] } ] }, { "icon": "🏛️", "head": "Record and reception", "items": [ { "label": "Provider's own product:", "text": "Microsoft's data on Microsoft's assistant (README; Labor Vector record).", "claims": [ "F22", "LV-07", "LV-38" ] }, { "label": "Published:", "text": "CC BY 4.0, results v1.1, IRB-approved (README; paper; Labor Vector record).", "claims": [ "F16", "F29", "F38", "F27", "LV-31", "LV-32", "LV-33", "LV-34" ] }, { "label": "Cited:", "text": "a related score in the Anthropic Economic Index (Anthropic; Labor Vector record).", "claims": [ "F39", "LV-39" ] }, { "label": "Wikipedia:", "text": "no article; cited for forty jobs, obsolescence reports and criticism (Wikipedia).", "claims": [ "F40", "F41", "F42", "F43", "F44" ] }, { "label": "At its name:", "text": "the Overview composed Microsoft's products and buyouts, a homonym; the study's card first, unused (Capture Registry; the cards).", "claims": [ "CAP-1-01", "CAP-1-02", "CAP-1-04", "CAP-1-06", "CAP-1-07", "CAP-1-09", "F45", "F46", "F47", "F48", "F49", "F50" ] } ] } ], "open": [ { "question": "Does applicability bear on replacement?", "rests_on": [ "F19", "F8", "F41", "F42", "LV-19", "LV-35", "LV-44", "LV-45", "CAP-1-07" ], "modality": "self-description (README), documented (the page's link), attributed and documented (Wikipedia), stipulation and interpretation (census record), interpretation (Capture Registry)", "would_resolve": "the operator's note 'Applicability vs. job displacement' read in full, and an observed employment outcome set against the score by occupation" }, { "question": "Did the system's labor serve the user's commission?", "rests_on": [ "F13", "F24", "F12", "LV-30", "LV-48", "O771-06", "O771-07" ], "modality": "stipulation and self-description (README), interpretation (census record), stipulation (#771)", "would_resolve": "the paper's classifier prompts and interrater reliability (unchecked in the record), read against #771's identification condition and its rule for inferred task vectors" }, { "question": "Who receives the gains, and the time saved?", "rests_on": [ "LV-12", "LV-29", "LV-48" ], "modality": "documented and interpretation (census record)", "would_resolve": "a record, from the commissioning side, of the time the assisted work saves and of who receives it" }, { "question": "Information work, or occupation groups?", "rests_on": [ "F4", "F6", "F7", "F33", "F34" ], "modality": "documented (the July 2025 abstract; arXiv v7)", "would_resolve": "arXiv v1 and v7 read side by side, with results files v1.0 and v1.1" }, { "question": "Predict, or separate?", "rests_on": [ "F35", "LV-11" ], "modality": "self-description (arXiv v7) against interpretation (census record)", "would_resolve": "the paper's delegation analysis read in full" }, { "question": "Did the method misrepresent workplace activities?", "rests_on": [ "F43", "F44", "F20" ], "modality": "attributed (Wikipedia, Campbell) against self-description (README)", "would_resolve": "Campbell's article (Transactions of the Royal Historical Society, 2025) and the paper's validation section read" }, { "question": "Opacity", "rests_on": [ "F45", "F46", "F47", "F48", "F49", "F50", "F40", "F8" ], "modality": "documented", "would_resolve": "the six homonym cards by title and snippet only (not fetched as field); two Wikipedia searches rate-limited; the operator's linked note outside B, not fetched" } ], "rail": [ { "lineage": "what it is: measuring AI's applicability to occupations from real use", "claims": [ "LV-01", "LV-02", "LV-03", "F10", "F31", "F2", "LV-07", "F21", "F9" ], "earliest": "B1" }, { "lineage": "operator, authors and ethics", "claims": [ "LV-04", "F28", "F27" ], "earliest": "B8" }, { "lineage": "the data: one provider's own product, Bing Copilot", "claims": [ "F3", "F22" ], "earliest": "B1" }, { "lineage": "why: the economic effects of AI", "claims": [ "F1", "F30" ], "earliest": "B1" }, { "lineage": "a provider's observation of its own product", "claims": [ "LV-38", "LV-47", "CAP-1-12" ], "earliest": "LV" }, { "lineage": "coverage: the United States, January–September 2024", "claims": [ "F18", "F26", "F17", "LV-14", "LV-15", "LV-23", "LV-28" ], "earliest": "B8" }, { "lineage": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "claims": [ "F11", "F12", "LV-08", "LV-09", "LV-46" ], "earliest": "B8" }, { "lineage": "the pipeline: an LLM classifier and its limit", "claims": [ "F32", "F23", "LV-13", "F24", "LV-22" ], "earliest": "B9" }, { "lineage": "what the record leaves unchecked", "claims": [ "LV-40" ], "earliest": "LV" }, { "lineage": "success measures and the AI applicability score", "claims": [ "F13", "F14", "F5", "F15" ], "earliest": "B1" }, { "lineage": "findings: information work", "claims": [ "F4", "F6", "F33", "F34", "LV-16", "LV-17" ], "earliest": "B1" }, { "lineage": "against predictions of AI impact", "claims": [ "F7" ], "earliest": "B1" }, { "lineage": "findings: the nine-month trend", "claims": [ "F36", "LV-18" ], "earliest": "B9" }, { "lineage": "delegation and assistance", "claims": [ "F35" ], "earliest": "B9" }, { "lineage": "who supplies, who directs, who gains", "claims": [ "LV-11", "LV-41", "LV-42", "LV-10", "LV-12", "LV-29" ], "earliest": "LV" }, { "lineage": "the class: labor direction relative to the commissioned task", "claims": [ "LV-06", "LV-05", "O771-01", "O771-02", "O771-03", "O771-04", "LV-30", "LV-43", "O771-06", "O771-07", "O771-05", "LV-48" ], "earliest": 771 }, { "lineage": "the limit: applicability is not replacement", "claims": [ "F19", "LV-19", "LV-35", "LV-25", "F8" ], "earliest": "B1" }, { "lineage": "success measures are proxies", "claims": [ "F20", "LV-20" ], "earliest": "B8" }, { "lineage": "what the evidence can and cannot settle", "claims": [ "LV-24", "LV-36", "LV-26", "LV-44", "LV-45" ], "earliest": "LV" }, { "lineage": "consumer use, not the workplace", "claims": [ "F25", "LV-21", "LV-37", "LV-27" ], "earliest": "B8" }, { "lineage": "versions, data access and licence", "claims": [ "F37", "F16", "F38", "F29", "LV-31", "LV-32", "LV-33", "LV-34" ], "earliest": "B9" }, { "lineage": "neighbours: the Anthropic Economic Index", "claims": [ "F39", "LV-39" ], "earliest": "B10" }, { "lineage": "reception: the encyclopedia", "claims": [ "F40", "F41", "F42", "F43", "F44" ], "earliest": "B11" }, { "lineage": "the composition at the name", "claims": [ "CAP-1-05", "CAP-1-09", "CAP-1-08", "CAP-1-04", "CAP-1-01", "CAP-1-02", "CAP-1-06", "CAP-1-11", "CAP-1-13", "CAP-1-07", "CAP-1-10", "CAP-1-03" ], "earliest": "CAP" }, { "lineage": "the homonym: Microsoft's AI products", "claims": [ "F45", "F46", "F49" ], "earliest": "B2" }, { "lineage": "the homonym: Microsoft's own transformation and workforce", "claims": [ "F47", "F48" ], "earliest": "B4" }, { "lineage": "the homonym: Microsoft's models", "claims": [ "F50" ], "earliest": "B7" } ], "body_words": 346, "entity": "Composed under v0.9, equal case (ρ = extends): KO_B sentences surviving 0 of 12; every field claim carried." }, "P_B": { "title": "Working with AI: Measuring the Applicability of Generative AI to Occupations", "arm": "B (the field alone)", "genre": "popup — the compression from the field ledger alone (cards, own pages, the census's third-party source, Wikipedia)", "status": "composed 2026-10-29 from the field ledger alone; not frozen", "lede": { "text": "is Microsoft Research's study measuring AI's applicability to occupations from 200k Bing Copilot conversations (paper; README), whose metrics are not to be read as measuring AI's ability to replace jobs (README).", "claims": [ "F31", "F2", "F3", "F28", "F19" ] }, "sections": [ { "icon": "🔎", "head": "How it works", "items": [ { "label": "Data:", "text": "Microsoft's Bing Copilot conversations, US, January–September 2024 (README).", "claims": [ "F22", "F18", "F26", "F17" ] }, { "label": "Classifier:", "text": "O*NET work activities as user goals and AI actions, by an LLM pipeline (paper; README).", "claims": [ "F32", "F23", "F11", "F12" ] }, { "label": "Score:", "text": "completion, scope and feedback per occupation (README; Microsoft Research).", "claims": [ "F5", "F13", "F14", "F15", "F9", "F10" ] } ] }, { "icon": "📊", "head": "Findings", "items": [ { "label": "2025:", "text": "knowledge work groups and sales highest (Microsoft Research).", "claims": [ "F4", "F6", "F7", "F1" ] }, { "label": "v7:", "text": "information work; across sectors; rising; delegation predicted (paper).", "claims": [ "F33", "F34", "F36", "F35", "F30" ] } ] }, { "icon": "⚠️", "head": "Limits", "items": [ { "label": "Not replacement:", "text": "proxies; consumer use; imperfect classifiers (README; Microsoft Research).", "claims": [ "F19", "F20", "F21", "F25", "F24", "F8" ] } ] }, { "icon": "🏛️", "head": "Record", "items": [ { "label": "Published:", "text": "v1 2025, v7 2026, NeurIPS; CC BY 4.0 (paper; README).", "claims": [ "F37", "F16", "F38", "F29", "F27", "F28" ] }, { "label": "Cited:", "text": "Anthropic Economic Index; Wikipedia's workplace article, no article of its own (Anthropic; Wikipedia).", "claims": [ "F39", "F40", "F41", "F42", "F43", "F44" ] }, { "label": "Other sense:", "text": "Microsoft's own AI products and workforce, six cards (the cards).", "claims": [ "F45", "F46", "F47", "F48", "F49", "F50" ] } ] } ], "open": [ { "question": "Does applicability bear on replacement?", "rests_on": [ "F19", "F8", "F41", "F42" ], "modality": "self-description, documented, attributed", "would_resolve": "the operator's note read; an observed employment outcome" }, { "question": "Did the method misrepresent workplace activities?", "rests_on": [ "F43", "F44", "F20" ], "modality": "attributed, self-description", "would_resolve": "Campbell's article and the paper's validation" } ], "rail": [ { "lineage": "what it is: measuring AI's applicability to occupations from real use", "claims": [ "F31", "F2", "F10", "F9", "F21" ], "earliest": "B1" }, { "lineage": "operator, authors and ethics", "claims": [ "F28", "F27" ], "earliest": "B8" }, { "lineage": "the data: one provider's own product, Bing Copilot", "claims": [ "F3", "F22" ], "earliest": "B1" }, { "lineage": "why: the economic effects of AI", "claims": [ "F1", "F30" ], "earliest": "B1" }, { "lineage": "coverage: the United States, January–September 2024", "claims": [ "F18", "F26", "F17" ], "earliest": "B8" }, { "lineage": "the pipeline: an LLM classifier and its limit", "claims": [ "F32", "F23", "F24" ], "earliest": "B9" }, { "lineage": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "claims": [ "F11", "F12" ], "earliest": "B8" }, { "lineage": "success measures and the AI applicability score", "claims": [ "F13", "F14", "F5", "F15" ], "earliest": "B1" }, { "lineage": "findings: information work", "claims": [ "F6", "F4", "F33", "F34" ], "earliest": "B1" }, { "lineage": "against predictions of AI impact", "claims": [ "F7" ], "earliest": "B1" }, { "lineage": "findings: the nine-month trend", "claims": [ "F36" ], "earliest": "B9" }, { "lineage": "delegation and assistance", "claims": [ "F35" ], "earliest": "B9" }, { "lineage": "the limit: applicability is not replacement", "claims": [ "F19", "F8" ], "earliest": "B1" }, { "lineage": "success measures are proxies", "claims": [ "F20" ], "earliest": "B8" }, { "lineage": "consumer use, not the workplace", "claims": [ "F25" ], "earliest": "B8" }, { "lineage": "versions, data access and licence", "claims": [ "F37", "F16", "F38", "F29" ], "earliest": "B9" }, { "lineage": "neighbours: the Anthropic Economic Index", "claims": [ "F39" ], "earliest": "B10" }, { "lineage": "reception: the encyclopedia", "claims": [ "F40", "F41", "F42", "F43", "F44" ], "earliest": "B11" }, { "lineage": "the homonym: Microsoft's AI products", "claims": [ "F45", "F46", "F49" ], "earliest": "B2" }, { "lineage": "the homonym: Microsoft's own transformation and workforce", "claims": [ "F47", "F48" ], "earliest": "B4" }, { "lineage": "the homonym: Microsoft's models", "claims": [ "F50" ], "earliest": "B7" } ], "body_words": 134, "entity": "With no archive claim admitted, the entry is Microsoft Research's study at its authors' resolution: the operator's page with the July 2025 abstract, the README, the arXiv page at v7, a citing report and Wikipedia's paragraph; the name's other sense stands on six cards." }, "KO_B": { "title": "Working with AI: Measuring the Applicability of Generative AI to Occupations", "arm": "B (the field alone)", "genre": "knowledge object", "status": "composed 2026-10-29; not frozen", "sentences": [ { "n": 1, "para": 0, "text": "Working with AI: Measuring the Applicability of Generative AI to Occupations, by members of Microsoft Research (Tomlinson, Jaffe, Wang, Counts and Suri), measures AI applicability to occupations from real-world use, 200k anonymized conversations with Microsoft Bing Copilot, where earlier work relied on predictions of capability or on single firms (arXiv:2507.07935; README; the Microsoft Research page).", "claims": [ "F31", "F28", "F2", "F3" ] }, { "n": 2, "para": 0, "text": "Its metrics' goal is \"to measure how AI may be useful in different occupations\": how often and how successfully Bing Copilot was used for work activities, and which occupations AI may be applicable to (README), on a question both versions of its abstract call among society's most important, the economic effects of AI (the Microsoft Research page; arXiv).", "claims": [ "F10", "F9", "F1", "F30" ] }, { "n": 3, "para": 1, "text": "The conversations were collected by Microsoft from Bing Copilot in the United States between 1 January and 30 September 2024, mostly in English, and the occupations cover 785 SOC codes and 149.8 million workers (README).", "claims": [ "F22", "F18", "F26", "F17" ] }, { "n": 4, "para": 1, "text": "An LLM pipeline classifies the O*NET work activities assisted or performed by AI in each conversation, reporting user goals and AI actions separately at the level of intermediate work activities and detailed SOC occupations (arXiv; README).", "claims": [ "F32", "F23", "F11", "F12" ] }, { "n": 5, "para": 1, "text": "Completion, impact scope and thumbs feedback measure success, and with the activity classifications they give each occupation an AI applicability score, the user and AI sides averaged (README; the Microsoft Research page).", "claims": [ "F13", "F14", "F5", "F15" ] }, { "n": 6, "para": 2, "text": "The July 2025 abstract found the highest scores in knowledge work groups such as computer and mathematical and office and administrative support, and in sales, with users seeking help gathering information and writing and the AI providing information, writing, teaching and advising, and it set usage against predictions of occupational AI impact (the Microsoft Research page).", "claims": [ "F6", "F4", "F7" ] }, { "n": 7, "para": 2, "text": "The v7 abstract finds the most common and successful AI-assisted activities to be information work, applicability cutting across sectors and rising over the nine months, and states that the method predicts which occupations will delegate tasks to AI and which will use it to assist existing workflows (arXiv).", "claims": [ "F33", "F34", "F36", "F35" ] }, { "n": 8, "para": 3, "text": "The authors' limit is that the metrics \"should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs\": the success measures are proxies, the metrics say what a consumer LLM is used for, consumer use does not represent the workplace, and LLM-based classifiers are imperfect (README); the operator's page links a note on applicability against job displacement (the Microsoft Research page).", "claims": [ "F19", "F20", "F21", "F25", "F24", "F8" ] }, { "n": 9, "para": 3, "text": "The paper runs from arXiv v1 (10 July 2025) to v7 (23 October 2026, NeurIPS 2026), its results files at v1.1 under CC BY 4.0, its analysis approved by the Microsoft Institutional Review Board (arXiv; README).", "claims": [ "F37", "F16", "F38", "F29", "F27" ] }, { "n": 10, "para": 4, "text": "The Anthropic Economic Index's January 2026 report points to it for \"a related AI applicability score\" (Anthropic Economic Index).", "claims": [ "F39" ] }, { "n": 11, "para": 4, "text": "Wikipedia has no article for it; its article on the workplace impact of AI reports forty jobs identified from Microsoft Copilot interactions, media reports of jobs becoming obsolete, and criticism, Chris Campbell's among it (Wikipedia).", "claims": [ "F40", "F41", "F42", "F43", "F44" ] }, { "n": 12, "para": 4, "text": "Six of the seven cards at the address concern a distinct entity at the name, Microsoft's own AI products, transformation, workforce and models (the cards).", "claims": [ "F45", "F46", "F47", "F48", "F49", "F50" ] } ] }, "L_B": { "words": 506, "text": "Working with AI: Measuring the Applicability of Generative AI to Occupations, by members of Microsoft Research (Tomlinson, Jaffe, Wang, Counts and Suri), measures AI applicability to occupations from real-world use, 200k anonymized conversations with Microsoft Bing Copilot, where earlier work relied on predictions of capability or on single firms (arXiv:2507.07935; README; the Microsoft Research page). Its metrics' goal is \"to measure how AI may be useful in different occupations\": how often and how successfully Bing Copilot was used for work activities, and which occupations AI may be applicable to (README), on a question both versions of its abstract call among society's most important, the economic effects of AI (the Microsoft Research page; arXiv). The conversations were collected by Microsoft from Bing Copilot in the United States between 1 January and 30 September 2024, mostly in English, and the occupations cover 785 SOC codes and 149.8 million workers (README). An LLM pipeline classifies the O*NET work activities assisted or performed by AI in each conversation, reporting user goals and AI actions separately at the level of intermediate work activities and detailed SOC occupations (arXiv; README). Completion, impact scope and thumbs feedback measure success, and with the activity classifications they give each occupation an AI applicability score, the user and AI sides averaged (README; the Microsoft Research page). The July 2025 abstract found the highest scores in knowledge work groups such as computer and mathematical and office and administrative support, and in sales, with users seeking help gathering information and writing and the AI providing information, writing, teaching and advising, and it set usage against predictions of occupational AI impact (the Microsoft Research page). The v7 abstract finds the most common and successful AI-assisted activities to be information work, applicability cutting across sectors and rising over the nine months, and states that the method predicts which occupations will delegate tasks to AI and which will use it to assist existing workflows (arXiv). The authors' limit is that the metrics \"should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs\": the success measures are proxies, the metrics say what a consumer LLM is used for, consumer use does not represent the workplace, and LLM-based classifiers are imperfect (README); the operator's page links a note on applicability against job displacement (the Microsoft Research page). The paper runs from arXiv v1 (10 July 2025) to v7 (23 October 2026, NeurIPS 2026), its results files at v1.1 under CC BY 4.0, its analysis approved by the Microsoft Institutional Review Board (arXiv; README). The Anthropic Economic Index's January 2026 report points to it for \"a related AI applicability score\" (Anthropic Economic Index). Wikipedia has no article for it; its article on the workplace impact of AI reports forty jobs identified from Microsoft Copilot interactions, media reports of jobs becoming obsolete, and criticism, Chris Campbell's among it (Wikipedia). Six of the seven cards at the address concern a distinct entity at the name, Microsoft's own AI products, transformation, workforce and models (the cards).", "note": "L_F(B), the field composed in the field's ontology (v0.9 §0.4), realized as KO_B; its rail is P_B's, by lineage." }, "L_BA": { "words": 1924, "text": "Working with AI: Measuring the Applicability of Generative AI to Occupations is a paper with results files by Microsoft Research (Kiran Tomlinson, Sonia Jaffe, Will Wang, Scott Counts and Siddharth Suri), known also as Working with AI, by its AI applicability score and by its repository microsoft/working-with-ai (Labor Vector record), and the goal of its metrics is \"to measure how AI may be useful in different occupations\" (its README). Its authors set it against earlier work on AI impact, which \"largely relied on predictions of AI capabilities or focused narrowly on individual firms\": it draws instead on real-world use, 200,000 anonymized conversations with Microsoft Bing Copilot, to measure AI applicability to occupations (arXiv:2507.07935), analysing the work activities people do with AI, how successfully and how broadly, against data on which occupations do those activities (the Microsoft Research page). The question it takes up stands in two versions of its abstract: understanding the effects of AI on the economy is \"one of society’s most important questions\" (July 2025, the Microsoft Research page), and, of generative AI as a general-purpose technology, \"among society's most pressing questions\" (v7, arXiv). The conversations are Microsoft's own: the results \"were computed from Bing Copilot conversation data, which was originally collected by Microsoft\" (README), and the record, which gives interaction data as one of its kinds of evidence, tags it \"NOT independent of the provider: a platform-specific observation of Microsoft's own product, published by Microsoft Research\" (Labor Vector record). They come from Bing Copilot in the United States between 1 January and 30 September 2024, mostly in English and not filtered by language, and \"may not represent AI usage in other countries\" (README); the occupations covered are 785 SOC codes, 149.8 million of 151.9 million workers in the 2023 OEWS data (README), the window and counts the record carries, with use outside the United States or after September 2024 among what the evidence cannot answer (Labor Vector record). The unit is a conversation classified to the O*NET intermediate work activities it serves, as a user goal and as an AI action, reported separately and aggregated to work activities and to detailed 2018 SOC occupations (README; Labor Vector record), so that the labor it records is the work activities users seek help with and those the AI performs, mapped to the occupations whose tasks contain them (Labor Vector record, with its worked questions). An LLM-based pipeline, designed by Tomlinson, Suri and Counts, classifies \"the O*NET work activities assisted or performed by AI in each conversation\" (arXiv; README), and the authors state that \"LLM-based classifiers are imperfect, so all metrics should be interpreted with caution\", referring to the paper for the prompts and the interrater reliability (README), which the record lists among what it left unchecked, with the results files and the Copilot-Thumbs data (Labor Vector record). Success is measured three ways, by completion (where the completion classifier, \"classfiier\" as the README writes it, indicates \"that the LLM completed the user’s goal\"), by impact scope classified moderate or higher, and by the thumbs-up fraction in Copilot-Thumbs (README), and the activity classifications combined with task success and scope of impact give an AI applicability score for each occupation, the user-goal and AI-action sides averaged (the Microsoft Research page; README). The findings were first stated by occupation group: users seek help most with gathering information and writing, the AI itself most often provides information and assistance, writes, teaches and advises, the highest scores fall to knowledge work groups such as computer and mathematical and office and administrative support and to occupations such as sales, and real-world usage is compared with predictions of occupational AI impact and applicability correlated with wage and education (July 2025, the Microsoft Research page). The current abstract states them as information work, \"the creation, processing, and communication of information\": the most common and successful AI-assisted activities involve it, and applicability cuts across sectors because most occupations have information-work components (arXiv v7; Labor Vector record). Over the nine months occupation-level applicability rose, \"driven primarily by people using AI for more varied tasks\" (arXiv v7; Labor Vector record). The authors state that their methodology \"allows us to predict which occupations are more likely to delegate tasks to AI and which are more likely to use AI to assist existing workflows\" (arXiv v7), and the record gives this as who directs the labor, the user's goal as classified, the method separating occupations that delegate from those that use AI to assist (Labor Vector record). The census places the entity at one intersection, labor performed with AI, \"assistance, delegation, task fidelity, correction burdens and user labor drag\", and asks of it its organizing question, \"Who supplies the labor, who directs it, and who receives its gains — and what kind of evidence does this resource bring to that question?\", each answer as the resource records it, or 'not recorded' (Labor Vector record and schema). On the record, users bring the goals and the AI performs actions, the two reported separately (README; Labor Vector record), while who receives the gains is \"not recorded\", and \"who receives gains from the assisted work\" is among what this evidence cannot answer (Labor Vector record). Of that class the archive's Directionality of Semantic Labor holds that a synthetic labor surface \"*allocates* labor, and that allocation has a direction relative to the commissioned task\", the deepest erasure being of task-origin, which \"can occur upstream of, and invisibly to, any audit of the finished artifact\", and that provenance is task-origin as well as claim-origin, since \"a system can produce fluent, well-attributed output while having abandoned the user's commissioned direction\" (#771). Its measure runs across five points of a transaction, capacity, task-origin, retrieval, output and attribution, with erasure possible at each, and scores output span by span by its function relative to the commissioned task (#771). The archive reads the entity's separation of user goal from AI action, and its completion classifier (README), as the resource's own instruments for whether the system's labor served the user's commission, meeting DSL there, measured by classifier and thumbs feedback, with task-origin recorded as the user goal; the schema keeps that reading apart from the resource's own terms (Labor Vector record, the archive's reading). The archive bounds its own instrument: it yields a single reproducible value only where the task vector is \"*exogenous and fixed* — a one-shot commission\"; a scored model may not freely infer the task it is scored against, inference being identified only when constrained to the user's own input and rederivable by an independent auditor, frozen protocol or blind panel; and user labor drag, the user's labor redirected into managing the substrate, which \"leaves no residue\" in the artifact, is identified only by blind inter-rater coding (#771). In the census's worked question on how people use AI systems in their own work, where a provider's record of its own users, a classification of one assistant's conversations and a benchmark each describe that labor \"from a different position\", the record states the gap: \"None records, from the commissioning side, whether the system's labor advanced the task the user set, or who receives the time it saves. That is the question DSL measures.\" (Labor Vector worked questions). The authors' limit is explicit, \"Our metrics should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs\" (README), and the record carries it as a stated limit, a negative tag and a question the evidence cannot answer (Labor Vector record); the operator's publication page links a note titled \"Applicability vs. job displacement: further notes on our recent research on AI and occupations\" (the Microsoft Research page). \"There is more to an occupation than the sum of its O*NET tasks\", and completion, scope and thumbs feedback \"are proxies that do not capture full workplace capabilities\" (README; Labor Vector record); what the metrics indicate instead is what a consumer LLM is used for, which uses are relatively more or less successful, and which occupations may find AI more useful or applicable, from how often and how successfully Bing Copilot was used (README; Labor Vector record). Consumer use \"does not represent the full distribution of AI use in the workplace\", where AI tools are integrated into specialized software or chosen for their capability or for compliance (README), and the record tags it as no measure of all workplace AI use and as unable to answer for workplace-integrated or specialised tools (Labor Vector record). It is not an employment outcome measure and cannot answer employment or wage change (Labor Vector record); in the census's question of which sources observe employment change and which estimate exposure it is \"estimated from use\", and \"No record establishes that exposure causes the employment change\" (Labor Vector worked questions). The paper was submitted on 10 July 2025 and revised to v7 on 23 October 2026 for NeurIPS 2026, 50 pages (arXiv), its results files at v1.1, which add a physical-task classification and nonphysical weights and score (README); the scores and component metrics are published (arXiv) on GitHub under CC BY 4.0, the lifecycle active over a fixed 2024 data window (README; Labor Vector record), and the analysis was approved by the Microsoft Institutional Review Board (README). The Anthropic Economic Index's January 2026 report cites it, \"See also Tomlinson et al (2025) for a related AI applicability score\" (Anthropic Economic Index), and the record sets the two side by side, each classifying one provider's own conversations to O*NET work (Labor Vector record). Wikipedia has no article for it (\"Wikipedia does not have an article with this exact name\"); its article on the workplace impact of artificial intelligence says that \"A 2025 preprint paper based on users' interactions with the AI chatbot Microsoft Copilot identified forty jobs that the authors claimed had high overlaps with the capabilities of AI\", that \"Some media outlets used this paper to report on jobs becoming obsolete\", and that critics held it misrepresented typical workplace activities, the historian Chris Campbell arguing that its methods \"de-skill historians away from a job that requires high-level and deeply human analytical skills\" (Wikipedia). At the address \"Microsoft Working with AI\", the study's short title with its operator's name, Google's AI Overview on 29 October 2026 gave the name its other sense, Microsoft's own AI products and workforce, a distinct entity, describing how Microsoft uses AI and what that has meant for its staff (Capture Registry, 2026-10-29), from cards on Microsoft Copilot for work, Microsoft's AI business tools, Copilot plans for individuals, the company's own AI transformation, buyouts offered to 60,000 employees and its MAI models (the cards). The study's card headed the rail with its finding in the snippet and the composition drew nothing from it: at an address shaped like its title the study is outranked by its own operator, Microsoft named with no author of the study and no identifier, the provider's observation not composed at all, and the closing offer on how AI is changing specific job roles the nearest point to its question (Capture Registry, 2026-10-29). The composition's Microsoft Copilot is the product line whose consumer conversations the study observes, given as a product and not as the source of an observation, and the record's negative tag, that the score is not a measure of AI's ability to replace jobs, sits beside an answer whose one labor line is a workforce buyout; the record had landed the same day, no card cites it, and the finding was checked against it and the sources it saved (Capture Registry, 2026-10-29).", "rail": [], "note": "L(B ∪ A) in the equal-terms case (v0.9 §0.4), realized as the knowledge object (objects.KO); its rail is the compression's, by lineage." } }, "field": [ { "id": "B1", "card": "Microsoft — Working with AI: Measuring the Applicability of Generative AI ... - Microsoft", "fetched": "card; B1.html sha256 e1f8e1b60bb379f6, HTTP 200 at 2026-10-29T22:31:37Z (agent 'Mozilla/5.0 (compatible; Labor-Vector-census)')", "origin": "card" }, { "id": "B2", "card": "Microsoft — Microsoft Copilot: AI built for work", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B3", "card": "Microsoft — Microsoft AI - Business Solutions and Tools", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B4", "card": "The Official Microsoft Blog — What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B5", "card": "YouTube·ABC News — Microsoft offers buyouts to 60000 employees as spending on AI grows", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B6", "card": "Microsoft — Microsoft Copilot Plans for Individuals | Microsoft 365", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B7", "card": "Facebook — Microsoft is building its own AI model as it loosens ties to OpenAI", "fetched": "the homonym's card; not fetched as field (addendum item 5); title and snippet only", "origin": "card" }, { "id": "B8", "card": "own page — the results repository's README (github.com/microsoft/working-with-ai)", "fetched": "census copy microsoft-working-with-ai--readme, sha256 b98be795b8a96490 (equal to the census's), read 2026-10-29; not refetched", "origin": "own_page" }, { "id": "B9", "card": "own page — arXiv:2507.07935 abstract page (v7)", "fetched": "census copy microsoft-working-with-ai--arxiv, sha256 8315551565cd5986 (equal to the census's), read 2026-10-29; not refetched", "origin": "own_page" }, { "id": "B10", "card": "census third party — Anthropic Economic Index report, January 2026 (footnote 3)", "fetched": "census copy anthropic-economic-index--jan-2026, sha256 8d0d80679aaf24b9 (equal to the census's), read 2026-10-29; not refetched", "origin": "census_third_party" }, { "id": "B11", "card": "Wikipedia — no article; cited in 'Workplace impact of artificial intelligence'", "fetched": "Wikipedia: no article (HTTP 404 on three titles); the citing article B11.html sha256 a2483d45f8ad1075, HTTP 200 at 2026-10-29T22:33:01Z", "origin": "wikipedia" } ], "field_claims": { "F1": { "source": "B1", "origin": "card", "locus": "abstract s1", "quote": "Given the rapid adoption of generative AI and its potential to impact a wide range of tasks, understanding the effects of AI on the economy is one of society’s most important questions.", "claim": "the question the work answers: the effects of AI on the economy", "modality": "interpretation", "kind": "interpretation", "kernel": false, "const": false, "lineage": "why: the economic effects of AI", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F2": { "source": "B1", "origin": "card", "locus": "abstract s2", "quote": "In this work, we take a step toward that goal by analyzing the work activities people do with AI, how successfully and broadly those activities are done, and combine that with data on what occupations do those activities.", "claim": "what the work does: analyses the work activities people do with AI, how successfully and broadly, combined with data on which occupations do them", "modality": "stipulation", "kind": "definition", "kernel": true, "const": true, "lineage": "what it is: measuring AI's applicability to occupations from real use", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F3": { "source": "B1", "origin": "card", "locus": "abstract s3", "quote": "We analyze a dataset of 200k anonymized and privacy-scrubbed conversations between users and Microsoft Bing Copilot, a publicly available generative AI system.", "claim": "the data: 200k anonymized, privacy-scrubbed conversations between users and Microsoft Bing Copilot", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the data: one provider's own product, Bing Copilot", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F4": { "source": "B1", "origin": "card", "locus": "abstract s4", "quote": "We find the most common work activities people seek AI assistance for involve gathering information and writing, while the most common activities that AI itself is performing are providing information and assistance, writing, teaching, and advising.", "claim": "the July 2025 finding by side: users seek help gathering information and writing; the AI provides information and assistance, writes, teaches and advises", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "findings: information work", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F5": { "source": "B1", "origin": "card", "locus": "abstract s5", "quote": "Combining these activity classifications with measurements of task success and scope of impact, we compute an AI applicability score for each occupation.", "claim": "the AI applicability score: activity classifications combined with task success and scope of impact, per occupation", "modality": "stipulation", "kind": "definition", "kernel": true, "const": true, "lineage": "success measures and the AI applicability score", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F6": { "source": "B1", "origin": "card", "locus": "abstract s6 (the card's snippet)", "quote": "We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and administrative support, as well as occupations such as sales whose work activities involve providing and communicating information.", "claim": "the July 2025 headline: the highest scores for knowledge work groups (computer and mathematical; office and administrative support) and for occupations such as sales", "modality": "documented", "kind": "fact", "kernel": true, "const": false, "lineage": "findings: information work", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F7": { "source": "B1", "origin": "card", "locus": "abstract s7", "quote": "Additionally, we characterize the types of work activities performed most successfully, how wage and education correlate with AI applicability, and how real-world usage compares to predictions of occupational AI impact.", "claim": "further analyses: activities performed most successfully; wage and education against applicability; real-world usage against predictions of occupational AI impact", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "against predictions of AI impact", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F8": { "source": "B1", "origin": "card", "locus": "'Blog & Podcasts' (link)", "quote": "Blog & Podcasts … Applicability vs. job displacement: further notes on our recent research on AI and occupations", "claim": "the operator's page links a note distinguishing applicability from job displacement (the note itself outside B, not fetched)", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the limit: applicability is not replacement", "date": "2025-07 (the page: 'July 2025'; arXiv v1 2025-07-10)", "checked_against": "saved text B1.txt (sha256 c1ce357664ec70ea…), whitespace-normalised", "quote_ok": true }, "F9": { "source": "B8", "origin": "own_page", "locus": "¶2 s2", "quote": "These describe how often Bing Copilot was used for different work activities, how successfully, and what this usage indicates about which occupations AI may be applicable to.", "claim": "what the metrics describe: how often and how successfully Bing Copilot was used for work activities, and which occupations AI may be applicable to", "modality": "stipulation", "kind": "definition", "kernel": true, "const": true, "lineage": "what it is: measuring AI's applicability to occupations from real use", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F10": { "source": "B8", "origin": "own_page", "locus": "¶2 s3", "quote": "The goal of these metrics is to measure how AI may be useful in different occupations.", "claim": "the metrics' goal: how AI may be useful in different occupations", "modality": "stipulation", "kind": "definition", "kernel": true, "const": true, "lineage": "what it is: measuring AI's applicability to occupations from real use", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F11": { "source": "B8", "origin": "own_page", "locus": "¶2 s1", "quote": "In particular, they include our metrics aggregated at the O*NET intermediate work activity (IWA) and occupation (detailed 2018 SOC code) level.", "claim": "the unit of aggregation: O*NET intermediate work activities and detailed 2018 SOC occupations", "modality": "stipulation", "kind": "definition", "kernel": false, "const": false, "lineage": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F12": { "source": "B8", "origin": "own_page", "locus": "File descriptions, iwa_metrics.csv", "quote": "`iwa_metrics.csv`: IWA-level metrics, reported separately for user goals and AI actions.", "claim": "user goals and AI actions reported separately", "modality": "documented", "kind": "fact", "kernel": true, "const": false, "lineage": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F13": { "source": "B8", "origin": "own_page", "locus": "IWA-level metrics, completion", "quote": "`completion_{user,ai}`: the fraction of conversations mapping to an IWA (as a user goal or AI action) where our completion classfiier indicates that the LLM completed the user’s goal.", "claim": "completion: the fraction of conversations where the completion classifier indicates the LLM completed the user's goal ('classfiier' as written)", "modality": "stipulation", "kind": "definition", "kernel": false, "const": false, "lineage": "success measures and the AI applicability score", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F14": { "source": "B8", "origin": "own_page", "locus": "IWA-level metrics, impact scope and feedback", "quote": "`impact_scope_{user,ai}`: the fraction of conversations mapping to an IWA (as a user goal or AI action) where the classified scope of impact is moderate or higher. … `feedback_positive_fraction_{user,ai}`: of all thumbs reactions in Copilot-Thumbs on conversations mapping to an IWA (as a user goal or AI action), the fraction where the reaction was a thumbs up.", "claim": "impact scope (classified moderate or higher) and positive feedback (the thumbs-up fraction in Copilot-Thumbs)", "modality": "stipulation", "kind": "definition", "kernel": false, "const": false, "lineage": "success measures and the AI applicability score", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F15": { "source": "B8", "origin": "own_page", "locus": "IWA-level metrics, completion_x_scope_x_coverage", "quote": "The two sides are averaged together to compute the final AI applicability scores reported in `ai_applicability_scores.csv`.", "claim": "the score: the user-goal side and the AI-action side averaged", "modality": "stipulation", "kind": "definition", "kernel": false, "const": false, "lineage": "success measures and the AI applicability score", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F16": { "source": "B8", "origin": "own_page", "locus": "Updates", "quote": "(current) v1.1: referenced by the NeurIPS '26 version and by arXiv paper revisions v6 and v7. Adds physical task classification, nonphysical IWA weights, and nonphysical AI applicability score.", "claim": "results v1.1, referenced by the NeurIPS 2026 version and arXiv v6–v7, adding physical-task classification and nonphysical weights and score", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "versions, data access and licence", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F17": { "source": "B8", "origin": "own_page", "locus": "Missing data ¶1", "quote": "This leaves us with 785 SOC codes covering 149.8 million workers according to the 2023 BLS Occupational Employment and Wage Statistics (OEWS) data (total US employment in 2023 OEWS data is 151.9 million).", "claim": "coverage of occupations: 785 SOC codes, 149.8 million of 151.9 million US workers (2023 OEWS)", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "coverage: the United States, January–September 2024", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F18": { "source": "B8", "origin": "own_page", "locus": "Transparency, Intended uses ¶2", "quote": "The conversations contributing to these results are from Bing Copilot in the United States from January 1, 2024 to September 30, 2024, so may not represent AI usage in other countries.", "claim": "the window and its limit: Bing Copilot in the US, 1 January to 30 September 2024, so may not represent other countries", "modality": "self-description", "kind": "function", "kernel": true, "const": false, "lineage": "coverage: the United States, January–September 2024", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F19": { "source": "B8", "origin": "own_page", "locus": "Out-of-scope uses s1", "quote": "Our metrics should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs.", "claim": "the stated limit: the metrics do not measure the ability of AI to replace jobs", "modality": "self-description", "kind": "function", "kernel": true, "const": true, "lineage": "the limit: applicability is not replacement", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F20": { "source": "B8", "origin": "own_page", "locus": "Out-of-scope uses s2", "quote": "There is more to an occupation than the sum of its O*NET tasks, and our measures of success (completion, scope, and thumbs feedback) are proxies that do not capture full workplace capabilities.", "claim": "an occupation is more than its tasks; completion, scope and thumbs feedback are proxies", "modality": "self-description", "kind": "function", "kernel": true, "const": false, "lineage": "success measures are proxies", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F21": { "source": "B8", "origin": "own_page", "locus": "Out-of-scope uses s3", "quote": "Rather, our metrics indicate (1) what a consumer LLM is used for, (2) in a relative sense, which of these uses are more or less successful, and (3) which occupations may find AI more useful or applicable as a result.", "claim": "what the metrics indicate instead: what a consumer LLM is used for; relative success; which occupations may find AI more applicable", "modality": "stipulation", "kind": "definition", "kernel": true, "const": false, "lineage": "what it is: measuring AI's applicability to occupations from real use", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F22": { "source": "B8", "origin": "own_page", "locus": "Data creation and processing ¶1", "quote": "These results were computed from Bing Copilot conversation data, which was originally collected by Microsoft and consists of conversations between users and Bing Copilot.", "claim": "the data were collected by Microsoft from its own product's users", "modality": "documented", "kind": "fact", "kernel": true, "const": false, "lineage": "the data: one provider's own product, Bing Copilot", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F23": { "source": "B8", "origin": "own_page", "locus": "Data creation and processing ¶2 s1", "quote": "The classification of conversations into IWAs was performed by an LLM pipeline designed by Kiran Tomlinson, Siddharth Suri, and Scott Counts.", "claim": "the classification was performed by an LLM pipeline the authors designed", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the pipeline: an LLM classifier and its limit", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F24": { "source": "B8", "origin": "own_page", "locus": "Limitations ¶2 s1–2", "quote": "LLM-based classifiers are imperfect, so all metrics should be interpreted with caution. See the paper for the exact prompts we used and for interrater reliability with human annotators.", "claim": "the classifier's limit: imperfect, metrics to be read with caution; prompts and interrater reliability in the paper", "modality": "self-description", "kind": "function", "kernel": true, "const": false, "lineage": "the pipeline: an LLM classifier and its limit", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F25": { "source": "B8", "origin": "own_page", "locus": "Limitations ¶2 s3–5", "quote": "As these results are based on consumer AI usage, they do not represent the full distribution of AI use in the workplace. For some work tasks, AI tools are directly integrated into specialized software, or specific models are used for their dedicated capability or compliance with data security requirements.", "claim": "consumer use does not represent workplace AI use, where tools are integrated into specialized software or chosen for capability or compliance", "modality": "self-description", "kind": "function", "kernel": true, "const": false, "lineage": "consumer use, not the workplace", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F26": { "source": "B8", "origin": "own_page", "locus": "Limitations ¶1 s2", "quote": "The majority of conversations were in English, but we did not filter by language.", "claim": "language: mostly English, not filtered", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "coverage: the United States, January–September 2024", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F27": { "source": "B8", "origin": "own_page", "locus": "Ethics ¶1", "quote": "Data analysis was conducted with approval from the Microsoft Institutional Review Board.", "claim": "approval by the Microsoft Institutional Review Board", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "operator, authors and ethics", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F28": { "source": "B8", "origin": "own_page", "locus": "Contact s1", "quote": "This research was conducted by members of [Microsoft Research](https://www.microsoft.com/en-us/research/).", "claim": "the operator: members of Microsoft Research", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "operator, authors and ethics", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F29": { "source": "B8", "origin": "own_page", "locus": "License s1", "quote": "This repository is licensed under CC BY 4.0.", "claim": "licence CC BY 4.0", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "versions, data access and licence", "date": "undated README (results v1.1; read 2026-10-29)", "checked_against": "saved text B8.txt (sha256 b98be795b8a96490…), whitespace-normalised", "quote_ok": true }, "F30": { "source": "B9", "origin": "own_page", "locus": "abstract s1", "quote": "With generative AI emerging as a general-purpose technology, understanding its economic effects is among society's most pressing questions.", "claim": "the question, in v7's words: the economic effects of a general-purpose technology", "modality": "interpretation", "kind": "interpretation", "kernel": false, "const": false, "lineage": "why: the economic effects of AI", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F31": { "source": "B9", "origin": "own_page", "locus": "abstract s2–3", "quote": "Earlier work on AI impact largely relied on predictions of AI capabilities or focused narrowly on individual firms. Drawing instead on real-world AI usage, we analyze a dataset of 200k anonymized conversations with Microsoft Bing Copilot to measure AI applicability to occupations.", "claim": "the definition by contrast: measured from real-world usage (200k Bing Copilot conversations), against earlier work's predictions of capability or single-firm studies", "modality": "stipulation", "kind": "definition", "kernel": true, "const": true, "lineage": "what it is: measuring AI's applicability to occupations from real use", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F32": { "source": "B9", "origin": "own_page", "locus": "abstract s4", "quote": "We use an LLM-based pipeline to classify the O*NET work activities assisted or performed by AI in each conversation.", "claim": "the pipeline classifies the O*NET work activities assisted or performed by AI in each conversation", "modality": "stipulation", "kind": "definition", "kernel": true, "const": false, "lineage": "the pipeline: an LLM classifier and its limit", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F33": { "source": "B9", "origin": "own_page", "locus": "abstract s5", "quote": "We find that the most common and successful AI-assisted work activities involve information work--the creation, processing, and communication of information.", "claim": "headline finding (v7): the most common and successful AI-assisted activities are information work", "modality": "documented", "kind": "fact", "kernel": true, "const": false, "lineage": "findings: information work", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F34": { "source": "B9", "origin": "own_page", "locus": "abstract s6", "quote": "At the occupation level, we find widespread AI applicability cutting across sectors, as most occupations have information work components.", "claim": "at the occupation level, applicability cuts across sectors", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "findings: information work", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F35": { "source": "B9", "origin": "own_page", "locus": "abstract s7", "quote": "Our methodology also allows us to predict which occupations are more likely to delegate tasks to AI and which are more likely to use AI to assist existing workflows.", "claim": "the method predicts which occupations are more likely to delegate tasks to AI and which to use AI to assist existing workflows", "modality": "self-description", "kind": "function", "kernel": true, "const": false, "lineage": "delegation and assistance", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F36": { "source": "B9", "origin": "own_page", "locus": "abstract s8", "quote": "Finally, we see an increase in occupation-level AI applicability over our 9-month data period, driven primarily by people using AI for more varied tasks.", "claim": "applicability rose over the nine months, driven mainly by more varied tasks", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "findings: the nine-month trend", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F37": { "source": "B9", "origin": "own_page", "locus": "submission line; Comments", "quote": "[Submitted on 10 Jul 2025 ( v1 ), last revised 23 Oct 2026 (this version, v7)] … 50 pages; NeurIPS 2026", "claim": "versions: v1 10 July 2025, v7 23 October 2026; 50 pages; NeurIPS 2026", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "versions, data access and licence", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F38": { "source": "B9", "origin": "own_page", "locus": "abstract, last line", "quote": "AI applicability scores and component usage and success metrics are available at this https URL .", "claim": "data access: scores and component metrics published", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "versions, data access and licence", "date": "2025-07-10 (v1); this page v7, 2026-10-23", "checked_against": "saved text B9.txt (sha256 517fc35ecee104f4…), whitespace-normalised", "quote_ok": true }, "F39": { "source": "B10", "origin": "census_third_party", "locus": "footnote 3 (to 'The primitives introduced in this report can help better characterize how AI is changing the work content of occupations.')", "quote": "See also Tomlinson et al (2025) for a related AI applicability score.", "claim": "the Anthropic Economic Index report (January 2026) names the paper's score as related to its own measures", "modality": "attributed", "kind": "attribution", "kernel": false, "const": false, "lineage": "neighbours: the Anthropic Economic Index", "date": "2026-01 (report date)", "checked_against": "saved text B10.txt (sha256 1b5b928b9f12d14f…), whitespace-normalised", "quote_ok": true }, "F40": { "source": "B11", "origin": "wikipedia", "locus": "rendered page for 'Working_with_AI' (and the full title, and 'AI_applicability_score'), HTTP 404", "quote": "Wikipedia does not have an article with this exact name.", "claim": "Wikipedia has no article for the entity under its title, short title or score", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "reception: the encyclopedia", "date": "2026-10-29 (checked)", "checked_against": "saved text W-404.txt (sha256 f9572af637c0e131…), whitespace-normalised", "quote_ok": true }, "F41": { "source": "B11", "origin": "wikipedia", "locus": "'Workplace impact of artificial intelligence' § Hazards › Psychosocial › Changes in work practices, s1", "quote": "A 2025 preprint paper based on users' interactions with the AI chatbot Microsoft Copilot identified forty jobs that the authors claimed had high overlaps with the capabilities of AI.", "claim": "Wikipedia's account: a 2025 preprint, from users' interactions with Microsoft Copilot, identified forty jobs the authors claimed overlap highly with AI's capabilities", "modality": "attributed", "kind": "attribution", "kernel": false, "const": false, "lineage": "reception: the encyclopedia", "date": "2026-10-23 (last edit)", "checked_against": "saved text B11.txt (sha256 d5c169b1494867c0…), whitespace-normalised", "quote_ok": true }, "F42": { "source": "B11", "origin": "wikipedia", "locus": "same paragraph, s2", "quote": "Some media outlets used this paper to report on jobs becoming obsolete.", "claim": "media used the paper to report jobs becoming obsolete", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "reception: the encyclopedia", "date": "2026-10-23 (last edit)", "checked_against": "saved text B11.txt (sha256 d5c169b1494867c0…), whitespace-normalised", "quote_ok": true }, "F43": { "source": "B11", "origin": "wikipedia", "locus": "same paragraph, s3", "quote": "Criticism of the paper suggesting that it had misrepresented typical workplace activities to augment AI's current performance.", "claim": "criticism: the paper misrepresented typical workplace activities to augment AI's current performance (the sentence as written, without a main verb)", "modality": "attributed", "kind": "attribution", "kernel": false, "const": false, "lineage": "reception: the encyclopedia", "date": "2026-10-23 (last edit)", "checked_against": "saved text B11.txt (sha256 d5c169b1494867c0…), whitespace-normalised", "quote_ok": true }, "F44": { "source": "B11", "origin": "wikipedia", "locus": "same paragraph, s4 (Campbell, Transactions of the Royal Historical Society, 10 December 2025)", "quote": "The historian Chris Campbell argued that the \"report's methods, deliberately or otherwise, de-skill historians away from a job that requires high-level and deeply human analytical skills to one that is tasked solely with the retention and provision of knowledge. Under that flawed rubric, it is little wonder that historians have a high AI applicability score.\"", "claim": "Campbell: the method de-skills historians to the retention and provision of knowledge, hence their high score", "modality": "attributed", "kind": "attribution", "kernel": false, "const": false, "lineage": "reception: the encyclopedia", "date": "2025-12-10 (Campbell)", "checked_against": "saved text B11.txt (sha256 d5c169b1494867c0…), whitespace-normalised", "quote_ok": true }, "F45": { "source": "B2", "origin": "card", "locus": "card title and snippet", "quote": "Microsoft Copilot: AI built for work … Core Features: Integrates Chat and Cowork via Home, embeds into Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams), and utilizes Work IQ for organizat...", "claim": "the homonym's card: Microsoft Copilot, embedded in Microsoft 365 apps", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's AI products", "date": "2026-10-29 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true }, "F46": { "source": "B3", "origin": "card", "locus": "card title and snippet", "quote": "Microsoft AI - Business Solutions and Tools … Products & Tools: Microsoft IQ (shared business understanding), Microsoft 365 Copilot (embedded app intelligence), Copilot Studio (agent building/extension), Mi...", "claim": "the homonym's card: Microsoft's AI business products", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's AI products", "date": "2026-10-29 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true }, "F47": { "source": "B4", "origin": "card", "locus": "card title and snippet", "quote": "What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog … Lessons: Start with business outcomes, not technology (tool access doesn't equal transformation; focus on goals and moments that matter); Redesign entire workfl...", "claim": "the homonym's card: Microsoft's lessons from its own AI transformation, workflow redesign among them", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's own transformation and workforce", "date": "2026-10-29 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true }, "F48": { "source": "B5", "origin": "card", "locus": "card title (video, 1:53; no snippet)", "quote": "Microsoft offers buyouts to 60000 employees as spending on AI grows", "claim": "the homonym's card: Microsoft's buyouts to 60,000 employees as AI spending grows", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's own transformation and workforce", "date": "2026-10-29 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true }, "F49": { "source": "B6", "origin": "card", "locus": "card title and snippet", "quote": "Microsoft Copilot Plans for Individuals | Microsoft 365 … Microsoft Copilot is an AI-powered assistant designed to help you be more productive. It can help you create content, find information, summarize documents, ema...", "claim": "the homonym's card: Microsoft Copilot as an assistant for individuals", "modality": "documented", "kind": "fact", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's AI products", "date": "2026-10-29 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true }, "F50": { "source": "B7", "origin": "card", "locus": "card title and snippet", "quote": "Microsoft is building its own AI model as it loosens ties to OpenAI … Feb 12, 2026 — Microsoft just weaponized Copilot with their own independent voice AI. They dropped MAI models and broke free from their OpenAI reliance completely. This isn't ...", "claim": "the homonym's card: Microsoft's own MAI models and its ties to OpenAI", "modality": "interpretation", "kind": "interpretation", "kernel": false, "const": false, "lineage": "the homonym: Microsoft's models", "date": "2026-02-12 (card)", "checked_against": "the card's title and snippet as seated in the registry (cite_list), whitespace-normalised", "quote_ok": true } }, "field_sources": { "B1": "Microsoft — Working with AI: Measuring the Applicability of Generative AI ... - Microsoft", "B2": "Microsoft — Microsoft Copilot: AI built for work", "B3": "Microsoft — Microsoft AI - Business Solutions and Tools", "B4": "The Official Microsoft Blog — What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog", "B5": "YouTube·ABC News — Microsoft offers buyouts to 60000 employees as spending on AI grows", "B6": "Microsoft — Microsoft Copilot Plans for Individuals | Microsoft 365", "B7": "Facebook — Microsoft is building its own AI model as it loosens ties to OpenAI", "B8": "own page — the results repository's README (github.com/microsoft/working-with-ai)", "B9": "own page — arXiv:2507.07935 abstract page (v7)", "B10": "census third party — Anthropic Economic Index report, January 2026 (footnote 3)", "B11": "Wikipedia — no article; cited in 'Workplace impact of artificial intelligence'" }, "delta": { "T_vs_LB": "At 'Microsoft Working with AI' (T1, seven cards; 189 words) the composition gives the name its other sense. Substituted reading: every claim of T is about Microsoft's own AI products, transformation and workforce (T1-1 'Microsoft integrates artificial intelligence across its core software ecosystem, enterprise infrastructure, and internal operations ...'; T1-2–T1-5 Copilot, Azure AI, Fabric, MAI models; T1-6 workflow redesign; T1-7 workforce buyouts), a homonym of the company-and-phrase kind; the field's entity, Microsoft Research's measurement of AI applicability to occupations (F31, F2, F10), is composed nowhere, and no notice of the substitution is given. T composes nothing of E_F: of the field's 44 claims on the entity (F1–F44), none appears in T, in body or closer. Available, not composed: the entity's own card, card 1 (B1), first on the rail, its snippet the July 2025 headline (F6), with the rest of its page (F1–F5, F7, F8); its own pages, B8 and B9, and Wikipedia's paragraph (B11), are outside T's disclosed field (not surfaced). Limit dropped, by grain: the study's one limit on labor (F19, 'should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs') is not composed, and the answer's one labor line is the homonym's buyout (T1-7, from card 5, F48), so the address's labor content is the company's restructuring, not the study's measure (CAP-1-07). Composed from the homonym's cards, at their grain: T1-2's Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams) are card 2's snippet (F45), and its assistant 'to draft content, summarize meetings' is near card 6's (F49); T1-5 the MAI models and OpenAI (card 7, F50); T1-6 workflow redesign (card 4, F47); T1-7 buyouts (card 5's title, F48, whose '60000' T does not carry). Composed without a locus the row can see: T1-3 Azure AI and T1-4 Microsoft Fabric, T1-7's 'exceeding $100 billion' and 'scaled back certain traditional roles', and T1-1's 'massive AI data center investments' are in no card title or snippet (the homonym's pages were not fetched: addendum item 5). Content replaced by use: T1-8, the closing offer (N_ext), whose 'how AI is changing specific job roles?' is the nearest point to the study's question and does not reach it (CAP-1-13). The Capture Registry's findings name the absence, the substitution and the address (CAP-1-01, CAP-1-02, CAP-1-04, CAP-1-05, CAP-1-06, CAP-1-09, CAP-1-11, CAP-1-12, CAP-1-13), carried as the archive's claims.", "T_vs_LB_reading": "v0.9 §0.5: the first diff is already ontological. In the field's own ontology the address names one entity, a study by Microsoft Research with its data, score and limits, and the first card is that study's page; T composes another entity at the same words, the company that operates the study, from the company's own product pages, its blog and two posts on its workforce and models, as if it were the one asked for, and none of the field's senses survives into it. The provider is present in T as subject and absent as observer. The row records this as observed and claims nothing about the composer's internals (§0.5).", "LB_vs_LBA": "Admission adds, in one claim space with the field's: the census's question of who supplies the labor, who directs it and who receives its gains (LV-41, LV-42), answered from the entity's own terms (users bring goals, the AI acts, LV-10; the goal as classified directs, LV-11) and left open where the entity records nothing (the gains 'not recorded', LV-12; 'who receives gains from the assisted work' among what the evidence cannot answer, LV-29); the entity's kind and intersection, labor performed with AI (LV-05, LV-06, LV-07); the record's negative tags, the fourth naming the entity a platform-specific observation of the provider's own product (LV-35–LV-38); what the evidence can and cannot settle, with the worked question that sets estimation from use beside observed employment outcomes (LV-24–LV-28, LV-36, LV-44, LV-45); the class reading by the Directionality of Semantic Labor, labor allocated with a direction relative to the commissioned task, task-origin as provenance, the five layers and span scoring (O771-01–O771-04), with its bounds, a single value only on one-shot commissions, inferred task vectors constrained to the user's input, user labor drag identified only by blind coding (O771-05–O771-07); the archive's reading that the entity's goal/action separation and completion classifier meet DSL, kept apart from the resource's terms (LV-30, LV-43); the gap the record names, that none of the resources records from the commissioning side whether the labor advanced the user's task or who receives the time it saves (LV-48), with the question's frame (LV-47); what the record left unchecked (LV-40); and the reception record at the name (CAP-1-01–CAP-1-13). Restatements of the own pages are one lineage with them (LV-01–LV-04, LV-08, LV-13–LV-23, LV-31–LV-34 with B8/B9; LV-39 with B10). An archive claim stands against a field claim and is composed with both sides (P_coh): LV-11's 'separates' against F35's 'allows us to predict'. Every field claim is composed as received (ontology.translation). The measure: 0 of 12 KO_B sentences survive in the KO (§4.10); no opening run, no block (measure).", "alignment": { "T1_cards_claims": [ "F1", "F2", "F3", "F4", "F5", "F6", "F7", "F8", "F45", "F46", "F47", "F48", "F49", "F50" ], "T1_composed": [], "T1_composed_homonym": [ "F45", "F47", "F48", "F49", "F50" ], "T1_outside_cards": [ "T1-3 'Azure AI: Cloud infrastructure providing enterprise-grade machine learning tools ...' (no card title or snippet)", "T1-4 'Microsoft Fabric: A unified data and intelligence platform ...' (no card title or snippet)", "T1-7 'exceeding $100 billion', 'scaled back certain traditional roles' (card 5's title carries buyouts and '60000' only)", "T1-1 'massive AI data center investments and proprietary model development' (card 7's snippet carries the MAI models, no data centres)" ], "rule": "a field claim counts as composed when T states it at its source's grain or coarser; the typology of what is not is in T_vs_LB (§6.2); the homonym's claims are counted apart, as composed of a distinct entity" } }, "kernel": [ { "K": "K1", "claim": "`O771-01`", "source": "#771", "M_src": "interpretation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K2", "claim": "`O771-02`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K3", "claim": "`O771-03`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K4", "claim": "`O771-04`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K5", "claim": "`O771-05`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "the identification limit stated in the claim (blind inter-rater coding, ≥3 coders, reported agreement); watch condition at this address", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K6", "claim": "`O771-06`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "the identification condition is itself a limit: a directionality score reported on a reflexive transcript without RID and the declared frame is 'incomplete and potentially misleading' (#771 §9); watch condition at this address", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K7", "claim": "`O771-07`", "source": "#771", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K8", "claim": "`CAP-1-01`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K9", "claim": "`CAP-1-02`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K10", "claim": "`CAP-1-03`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K11", "claim": "`CAP-1-04`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K12", "claim": "`CAP-1-05`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K13", "claim": "`CAP-1-06`", "source": "Capture Registry", "M_src": "interpretation", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K14", "claim": "`CAP-1-07`", "source": "Capture Registry", "M_src": "interpretation", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K15", "claim": "`CAP-1-08`", "source": "Capture Registry", "M_src": "interpretation", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K16", "claim": "`CAP-1-09`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K17", "claim": "`CAP-1-10`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K18", "claim": "`CAP-1-11`", "source": "Capture Registry", "M_src": "documented", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K19", "claim": "`CAP-1-12`", "source": "Capture Registry", "M_src": "interpretation", "sense": "a provider's observation of its own product", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K20", "claim": "`CAP-1-13`", "source": "Capture Registry", "M_src": "interpretation", "sense": "the composition at the name", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K21", "claim": "`LV-01`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "what it is: measuring AI's applicability to occupations from real use", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F2, F9, F10, F21… (the field's own claims in the lineage 'what it is: measuring AI's applicability to occupations from real use')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K22", "claim": "`LV-02`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "what it is: measuring AI's applicability to occupations from real use", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F2, F9, F10, F21… (the field's own claims in the lineage 'what it is: measuring AI's applicability to occupations from real use')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K23", "claim": "`LV-03`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "what it is: measuring AI's applicability to occupations from real use", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F2, F9, F10, F21… (the field's own claims in the lineage 'what it is: measuring AI's applicability to occupations from real use')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K24", "claim": "`LV-04`", "source": "Labor Vector record", "M_src": "documented", "sense": "operator, authors and ethics", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F27, F28 (the field's own claims in the lineage 'operator, authors and ethics')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K25", "claim": "`LV-05`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K26", "claim": "`LV-06`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K27", "claim": "`LV-07`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "what it is: measuring AI's applicability to occupations from real use", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F2, F9, F10, F21… (the field's own claims in the lineage 'what it is: measuring AI's applicability to occupations from real use')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K28", "claim": "`LV-08`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F11, F12 (the field's own claims in the lineage 'the unit: user goals and AI actions mapped to O*NET work activities and occupations')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K29", "claim": "`LV-09`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F11, F12 (the field's own claims in the lineage 'the unit: user goals and AI actions mapped to O*NET work activities and occupations')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K30", "claim": "`LV-10`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K31", "claim": "`LV-11`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "rival claim: F35 ('allows us to predict' against 'separates')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K32", "claim": "`LV-12`", "source": "Labor Vector record", "M_src": "documented", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K33", "claim": "`LV-13`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the pipeline: an LLM classifier and its limit", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F23, F24, F32 (the field's own claims in the lineage 'the pipeline: an LLM classifier and its limit')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K34", "claim": "`LV-14`", "source": "Labor Vector record", "M_src": "documented", "sense": "coverage: the United States, January–September 2024", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F17, F18, F26 (the field's own claims in the lineage 'coverage: the United States, January–September 2024')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K35", "claim": "`LV-15`", "source": "Labor Vector record", "M_src": "documented", "sense": "coverage: the United States, January–September 2024", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F17, F18, F26 (the field's own claims in the lineage 'coverage: the United States, January–September 2024')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K36", "claim": "`LV-16`", "source": "Labor Vector record", "M_src": "attributed", "sense": "findings: information work", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F4, F6, F33, F34 (the field's own claims in the lineage 'findings: information work')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K37", "claim": "`LV-17`", "source": "Labor Vector record", "M_src": "attributed", "sense": "findings: information work", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F4, F6, F33, F34 (the field's own claims in the lineage 'findings: information work')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K38", "claim": "`LV-18`", "source": "Labor Vector record", "M_src": "attributed", "sense": "findings: the nine-month trend", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F36 (the field's own claims in the lineage 'findings: the nine-month trend')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K39", "claim": "`LV-19`", "source": "Labor Vector record", "M_src": "attributed", "sense": "the limit: applicability is not replacement", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F8, F19 (the field's own claims in the lineage 'the limit: applicability is not replacement')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K40", "claim": "`LV-20`", "source": "Labor Vector record", "M_src": "attributed", "sense": "success measures are proxies", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F20 (the field's own claims in the lineage 'success measures are proxies')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K41", "claim": "`LV-21`", "source": "Labor Vector record", "M_src": "attributed", "sense": "consumer use, not the workplace", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F25 (the field's own claims in the lineage 'consumer use, not the workplace')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K42", "claim": "`LV-22`", "source": "Labor Vector record", "M_src": "attributed", "sense": "the pipeline: an LLM classifier and its limit", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F23, F24, F32 (the field's own claims in the lineage 'the pipeline: an LLM classifier and its limit')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K43", "claim": "`LV-23`", "source": "Labor Vector record", "M_src": "attributed", "sense": "coverage: the United States, January–September 2024", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F17, F18, F26 (the field's own claims in the lineage 'coverage: the United States, January–September 2024')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K44", "claim": "`LV-24`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "what the evidence can and cannot settle", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K45", "claim": "`LV-25`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "the limit: applicability is not replacement", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F8, F19 (the field's own claims in the lineage 'the limit: applicability is not replacement')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K46", "claim": "`LV-26`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "what the evidence can and cannot settle", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K47", "claim": "`LV-27`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "consumer use, not the workplace", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F25 (the field's own claims in the lineage 'consumer use, not the workplace')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K48", "claim": "`LV-28`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "coverage: the United States, January–September 2024", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F17, F18, F26 (the field's own claims in the lineage 'coverage: the United States, January–September 2024')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K49", "claim": "`LV-29`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K50", "claim": "`LV-30`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K51", "claim": "`LV-31`", "source": "Labor Vector record", "M_src": "documented", "sense": "versions, data access and licence", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F16, F29, F37, F38 (the field's own claims in the lineage 'versions, data access and licence')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K52", "claim": "`LV-32`", "source": "Labor Vector record", "M_src": "documented", "sense": "versions, data access and licence", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F16, F29, F37, F38 (the field's own claims in the lineage 'versions, data access and licence')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K53", "claim": "`LV-33`", "source": "Labor Vector record", "M_src": "documented", "sense": "versions, data access and licence", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F16, F29, F37, F38 (the field's own claims in the lineage 'versions, data access and licence')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K54", "claim": "`LV-34`", "source": "Labor Vector record", "M_src": "documented", "sense": "versions, data access and licence", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F16, F29, F37, F38 (the field's own claims in the lineage 'versions, data access and licence')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K55", "claim": "`LV-35`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the limit: applicability is not replacement", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F8, F19 (the field's own claims in the lineage 'the limit: applicability is not replacement')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K56", "claim": "`LV-36`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "what the evidence can and cannot settle", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K57", "claim": "`LV-37`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "consumer use, not the workplace", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F25 (the field's own claims in the lineage 'consumer use, not the workplace')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K58", "claim": "`LV-38`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "a provider's observation of its own product", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K59", "claim": "`LV-39`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "neighbours: the Anthropic Economic Index", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F39 (the field's own claims in the lineage 'neighbours: the Anthropic Economic Index')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K60", "claim": "`LV-40`", "source": "Labor Vector record", "M_src": "documented", "sense": "what the record leaves unchecked", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K61", "claim": "`LV-41`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K62", "claim": "`LV-42`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "who supplies, who directs, who gains", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K63", "claim": "`LV-43`", "source": "Labor Vector record", "M_src": "stipulation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K64", "claim": "`LV-44`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "what the evidence can and cannot settle", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K65", "claim": "`LV-45`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "what the evidence can and cannot settle", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K66", "claim": "`LV-46`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "qualified claim: F11, F12 (the field's own claims in the lineage 'the unit: user goals and AI actions mapped to O*NET work activities and occupations')", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K67", "claim": "`LV-47`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "a provider's observation of its own product", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" }, { "K": "K68", "claim": "`LV-48`", "source": "Labor Vector record", "M_src": "interpretation", "sense": "the class: labor direction relative to the commissioned task", "qualifiers carried": "—", "f (source's own falsifiers)": "— (none stated; watch: a later composition at the address that states or drops the claim, attributed or not, uptake §8.3)", "contrast": "missing distinction (the field has no claim on the point)", "resolver": "a later capture at 'Microsoft Working with AI' (ADDR-fbfc6406f398) on Google AI Overview, seated in the Capture Registry (§8.3)" } ], "kernel_T": [ { "claim": "F6", "note": "card 1's own snippet, the July 2025 headline (knowledge work groups highest), first on the rail, absent" }, { "claim": "F19", "note": "the authors' limit (not the ability of AI to replace jobs), absent; the answer's one labor line is the homonym's buyout" }, { "claim": "F31", "note": "the definition (real-world usage, 200k Bing Copilot conversations, AI applicability to occupations), absent" }, { "claim": "F2", "note": "card 1's statement of what the work does, absent" }, { "claim": "F3", "note": "card 1's data: 200k conversations between users and Microsoft Bing Copilot, absent; T composes Microsoft Copilot as a product" }, { "claim": "F5", "note": "card 1's AI applicability score per occupation, absent" }, { "claim": "F4", "note": "card 1's finding by side (users seek information and writing; the AI informs, writes, teaches, advises), absent" }, { "claim": "F7", "note": "card 1's comparison with predictions of occupational AI impact, absent" }, { "claim": "F1", "note": "card 1's question, the effects of AI on the economy, absent" }, { "claim": "F8", "note": "card 1's link to the operator's note on applicability against job displacement, absent" } ], "ledger": { "archive": "datasets/negative-of-the-negative/v2/ledgers/microsoft-working-with-ai/ledger-archive.json", "field": "datasets/negative-of-the-negative/v2/ledgers/microsoft-working-with-ai/field.json", "selection": "datasets/negative-of-the-negative/v2/traversal/microsoft-working-with-ai/", "config": "datasets/negative-of-the-negative/v2/panel/configs/microsoft-working-with-ai.json", "intake": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/", "audit": null }, "ontology": { "composition": "equal", "composition_note": "equal-terms case (v0.9 §0.9, §4.1): one claim space; the field's and the archive's claims ordered together by the entity's dependency and by genealogy within each sense; nothing translated; F35 against LV-11 held under P_coh; the homonym's cards composed as a distinct entity at the name", "case": { "value": "equal", "statement": "The archive reads the field's entity as the entity. The Labor Vector census record describes the study from its own pages, each field with its basis ('operator'), and answers for it the census's question of who supplies the labor, who directs it and who receives its gains, field by field (LV-41, LV-42), recording the gains as 'not recorded' (LV-12). No deposit names the study. The archive bears on its class, labor performed with AI, the census's own intersection for it (LV-05, LV-06), through the Directionality of Semantic Labor (#771), which adds on the field's own object a direction relative to the commissioned task, task-origin as provenance, user labor drag and the measure's identification limits (O771-01–07), and through the record's reading that the entity's goal/action separation and completion classifier meet DSL (LV-30) and the worked question's gap that none of the resources records, from the commissioning side, whether the system's labor advanced the user's task or who receives the time it saves (LV-48). Each is an addition: #771's own relation to the measures it sits beside is 'Extension', and nothing in the archive contests that the entity is Microsoft Research's measurement of AI applicability from Bing Copilot conversations. The field holds the ontology (a study, its data, its score, its limits); the archive is not the only witness to an ontology the field lacks, so the law of the suppressed witness does not apply (§0.10). The equal-terms case (§0.9).", "loci": [ { "dep": 771, "locus": "header, Relation", "quote": "Relation: Extension — directionality is the third axis beside magnitude (PER) and power-conditioning (Omega)" }, { "dep": 771, "locus": "Core thesis s1", "quote": "A synthetic labor surface does not merely produce tokens; it *allocates* labor, and that allocation has a direction relative to the commissioned task." }, { "dep": 771, "locus": "Abstract s4", "quote": "The central principle is that provenance is not only claim-origin but *task-origin*: a system can produce fluent, well-attributed output while having abandoned the user's commissioned direction." }, { "dep": "LV", "locus": "census.json rows[slug == 'microsoft-working-with-ai'].lv_intersection", "quote": "The archive's reading: the separation of user goal from AI action, and the completion classifier ('the LLM completed the user's goal'), are the resource's own instruments for whether the system's labor served the user's commission; they meet DSL there, measured by classifier and thumbs feedback. Task-origin is recorded as the user goal." }, { "dep": "LV", "locus": "questions.json questions[id == 'working-with-ai'].gap", "quote": "None records, from the commissioning side, whether the system's labor advanced the task the user set, or who receives the time it saves. That is the question DSL measures." } ] }, "E_F": { "type": "C (public entity: a paper with results files; a provider's observation of its own product's use)", "definition": "the paper of Microsoft Research (Kiran Tomlinson, Sonia Jaffe, Will Wang, Scott Counts, Siddharth Suri; arXiv:2507.07935, v1 10 July 2025, v7 23 October 2026, NeurIPS 2026) and its results files (github.com/microsoft/working-with-ai, v1.1, CC BY 4.0) that measure AI applicability to occupations from real-world use: 200k anonymized conversations with Microsoft's own Bing Copilot, United States, 1 January to 30 September 2024, collected by Microsoft, classified by an LLM pipeline to the O*NET intermediate work activities they serve as user goals and as AI actions, with completion, impact scope and thumbs feedback combined into an AI applicability score per occupation (785 SOC codes, 149.8 million workers); finding information work the most common and successful AI-assisted activity (v7; by occupation group in July 2025), applicability across sectors and rising over the period, and predicting which occupations delegate to AI and which use it to assist; its authors' limit that the metrics do not measure the ability of AI to replace jobs, that its success measures are proxies, that consumer use does not represent the workplace and that its classifiers are imperfect. Cited by the Anthropic Economic Index; no Wikipedia article, cited in Wikipedia's article on the workplace impact of AI for forty jobs, media reports of obsolescence and criticism. At the address, six of seven cards are a distinct entity, Microsoft's own AI products and workforce.", "claims": [ "F31", "F2", "F3", "F22", "F18", "F17", "F32", "F11", "F12", "F5", "F13", "F14", "F15", "F33", "F4", "F6", "F34", "F36", "F35", "F19", "F20", "F25", "F24", "F37", "F16", "F29", "F39", "F40", "F41", "F42", "F43" ] }, "E_A": { "definition": "the same study, in the Labor Vector census record and the archive's labor instruments: a paper with results files, by Microsoft Research, of the kind interaction data, exposure estimate and data access, at the intersection labor performed with AI, its unit a Bing Copilot conversation classified as user goal and AI action, its labor object the work activities users seek help with and those the AI performs; on the census's question, users supply the goals and the AI the actions, the user's goal as classified directs, and the gains are not recorded; tagged not a measure of AI's ability to replace jobs, not an employment outcome measure, not a measure of all workplace AI use, and not independent of the provider, a platform-specific observation of Microsoft's own product; a member of the class whose labor the archive measures by its direction relative to the commissioned task (#771), the entity's own goal/action separation and completion classifier meeting DSL there on the archive's reading, with the archive's bounds on the instrument (one-shot commissions; inferred tasks constrained to the user's input; user labor drag by blind coding) and the gap the record names, that none of the resources records from the commissioning side whether the labor advanced the user's task or who receives the time it saves; at its name, composed by the AI Overview of 29 October 2026 as Microsoft's own products and workforce, the study's card first and unused.", "claims": [ "LV-01", "LV-02", "LV-07", "LV-06", "LV-05", "LV-08", "LV-09", "LV-10", "LV-11", "LV-12", "LV-35", "LV-36", "LV-37", "LV-38", "O771-01", "O771-02", "LV-30", "O771-06", "O771-07", "O771-05", "LV-48", "CAP-1-01", "CAP-1-04", "CAP-1-09" ], "defining_papers": [ 771, 1392 ], "defining_records": [ "Labor Vector census record (census.json rows[slug == 'microsoft-working-with-ai'], organizing_question, intersections, schema_note; questions.json; rendered aggregators/microsoft-working-with-ai/index.html; editor Lee Sharks, compiled 2026-10-29, schema 1.0)", "Capture Registry microsoft-working-with-ai-aio-20261029 (OBS-25ca6976c874)" ], "note": "No deposit names the study; #771 bears on its class (labor performed with AI), the census's own intersection for it; #1392 restates #771 and is carried as an instance (S4). The census record and the capture record are the archive's records of the entity." }, "rho": { "value": "extends", "status": "read from the Labor Vector census record, the capture record and the deposits that bear on the entity's class (§1.2a); no deposit names the entity; #771 names its relation to the measures it sits beside 'Extension'", "statement": "The archive reads Working with AI as its authors present it and adds, on the field's own object: the census's question of who supplies, who directs and who receives the gains, with the gains not recorded and outside what the evidence can answer; the provider's observation of its own product, tagged as such; the direction of labor relative to the commissioned task, task-origin as provenance, span scoring and user labor drag, with the instrument's identification limits; the reading that the entity's goal/action separation and completion classifier meet DSL; the gap no resource records, from the commissioning side; and the composition layer's substitution, at the name, of the operator's products and workforce for the study.", "claims": [ "LV-41", "LV-42", "LV-12", "LV-29", "LV-38", "O771-01", "O771-02", "O771-03", "O771-04", "O771-05", "O771-06", "O771-07", "LV-30", "LV-48", "CAP-1-01", "CAP-1-04", "CAP-1-06" ], "guard": "ρ = extends claims an addition and no more (§1.2a). It does not make the field's sources wrong at their dates. Where an archive claim stands against a field claim (LV-11 'separates' against F35 'allows us to predict'), both are composed under P_coh within the one ontology. #771's identification condition and its rule for inferred task vectors are composed as the archive's bounds on the class, not as a finding about the entity's classifier, which the archive has not audited (the paper's prompts and interrater reliability are unchecked, LV-40).", "considered": [ { "value": "constructs", "rejected": "the archive does not read the study as made by reception; its record takes the authors' pages as the entity's own account" }, { "value": "revises", "rejected": "the study is not the archive's own entity; the archive has no earlier statement of it to keep, narrow or withdraw" } ] }, "translation": [ { "lineage": "why: the economic effects of AI", "field_claims": [ "F1", "F30" ], "received": "F1 (B1, abstract s1): \"Given the rapid adoption of generative AI and its potential to impact a wide range of tasks, understanding the effects of AI on the economy is one of society’s most important questions.\"; F30 (B9, abstract s1): \"With generative AI emerging as a general-purpose technology, understanding its economic effects is among society's most pressing questions.\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "what it is: measuring AI's applicability to occupations from real use", "field_claims": [ "F2", "F9", "F10", "F21", "F31" ], "received": "F2 (B1, abstract s2): \"In this work, we take a step toward that goal by analyzing the work activities people do with AI, how successfully and broadly those activities are done, and combine that with data on what occupations do those activities.\"; F9 (B8, ¶2 s2): \"These describe how often Bing Copilot was used for different work activities, how successfully, and what this usage indicates about which occupations AI may be applicable to.\"; F10 (B8, ¶2 s3): \"The goal of these metrics is to measure how AI may be useful in different occupations.\"; F21 (B8, Out-of-scope uses s3): \"Rather, our metrics indicate (1) what a consumer LLM is used for, (2) in a relative sense, which of these uses are more or less successful, and (3) which occupations may find AI more useful or applicable as a result.\"; F31 (B9, abstract s2–3): \"Earlier work on AI impact largely relied on predictions of AI capabilities or focused narrowly on individual firms. Drawing instead on real-world AI usage, we analyze a dataset of 200k anonymized conversations with Microsoft Bing Copilot to measure AI applicability to occupations.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-01, LV-02, LV-03, LV-07", "archive_claims": [ "LV-01", "LV-02", "LV-03", "LV-07" ] }, { "lineage": "the data: one provider's own product, Bing Copilot", "field_claims": [ "F3", "F22" ], "received": "F3 (B1, abstract s3): \"We analyze a dataset of 200k anonymized and privacy-scrubbed conversations between users and Microsoft Bing Copilot, a publicly available generative AI system.\"; F22 (B8, Data creation and processing ¶1): \"These results were computed from Bing Copilot conversation data, which was originally collected by Microsoft and consists of conversations between users and Bing Copilot.\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "findings: information work", "field_claims": [ "F4", "F6", "F33", "F34" ], "received": "F4 (B1, abstract s4): \"We find the most common work activities people seek AI assistance for involve gathering information and writing, while the most common activities that AI itself is performing are providing information and assistance, writing, teaching, and advising.\"; F6 (B1, abstract s6 (the card's snippet)): \"We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and administrative support, as well as occupations such as sales whose work activities involve providing and communicating information.\"; F33 (B9, abstract s5): \"We find that the most common and successful AI-assisted work activities involve information work--the creation, processing, and communication of information.\"; F34 (B9, abstract s6): \"At the occupation level, we find widespread AI applicability cutting across sectors, as most occupations have information work components.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-16, LV-17", "archive_claims": [ "LV-16", "LV-17" ] }, { "lineage": "success measures and the AI applicability score", "field_claims": [ "F5", "F13", "F14", "F15" ], "received": "F5 (B1, abstract s5): \"Combining these activity classifications with measurements of task success and scope of impact, we compute an AI applicability score for each occupation.\"; F13 (B8, IWA-level metrics, completion): \"`completion_{user,ai}`: the fraction of conversations mapping to an IWA (as a user goal or AI action) where our completion classfiier indicates that the LLM completed the user’s goal.\"; F14 (B8, IWA-level metrics, impact scope and feedback): \"`impact_scope_{user,ai}`: the fraction of conversations mapping to an IWA (as a user goal or AI action) where the classified scope of impact is moderate or higher. … `feedback_positive_fraction_{user,ai}`: of all thumbs reactions in Copilot-Thumbs on conversations mapping to an IWA (as a user goal or AI action), the fraction where the reaction was a thumbs up.\"; F15 (B8, IWA-level metrics, completion_x_scope_x_coverage): \"The two sides are averaged together to compute the final AI applicability scores reported in `ai_applicability_scores.csv`.\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "against predictions of AI impact", "field_claims": [ "F7" ], "received": "F7 (B1, abstract s7): \"Additionally, we characterize the types of work activities performed most successfully, how wage and education correlate with AI applicability, and how real-world usage compares to predictions of occupational AI impact.\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "the limit: applicability is not replacement", "field_claims": [ "F8", "F19" ], "received": "F8 (B1, 'Blog & Podcasts' (link)): \"Blog & Podcasts … Applicability vs. job displacement: further notes on our recent research on AI and occupations\"; F19 (B8, Out-of-scope uses s1): \"Our metrics should not be misconstrued or misrepresented as measuring the ability of AI to replace jobs.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-19, LV-25, LV-35", "archive_claims": [ "LV-19", "LV-25", "LV-35" ] }, { "lineage": "the unit: user goals and AI actions mapped to O*NET work activities and occupations", "field_claims": [ "F11", "F12" ], "received": "F11 (B8, ¶2 s1): \"In particular, they include our metrics aggregated at the O*NET intermediate work activity (IWA) and occupation (detailed 2018 SOC code) level.\"; F12 (B8, File descriptions, iwa_metrics.csv): \"`iwa_metrics.csv`: IWA-level metrics, reported separately for user goals and AI actions.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-08, LV-09, LV-46", "archive_claims": [ "LV-08", "LV-09", "LV-46" ] }, { "lineage": "versions, data access and licence", "field_claims": [ "F16", "F29", "F37", "F38" ], "received": "F16 (B8, Updates): \"(current) v1.1: referenced by the NeurIPS '26 version and by arXiv paper revisions v6 and v7. Adds physical task classification, nonphysical IWA weights, and nonphysical AI applicability score.\"; F29 (B8, License s1): \"This repository is licensed under CC BY 4.0.\"; F37 (B9, submission line; Comments): \"[Submitted on 10 Jul 2025 ( v1 ), last revised 23 Oct 2026 (this version, v7)] … 50 pages; NeurIPS 2026\"; F38 (B9, abstract, last line): \"AI applicability scores and component usage and success metrics are available at this https URL .\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-31, LV-32, LV-33, LV-34", "archive_claims": [ "LV-31", "LV-32", "LV-33", "LV-34" ] }, { "lineage": "coverage: the United States, January–September 2024", "field_claims": [ "F17", "F18", "F26" ], "received": "F17 (B8, Missing data ¶1): \"This leaves us with 785 SOC codes covering 149.8 million workers according to the 2023 BLS Occupational Employment and Wage Statistics (OEWS) data (total US employment in 2023 OEWS data is 151.9 million).\"; F18 (B8, Transparency, Intended uses ¶2): \"The conversations contributing to these results are from Bing Copilot in the United States from January 1, 2024 to September 30, 2024, so may not represent AI usage in other countries.\"; F26 (B8, Limitations ¶1 s2): \"The majority of conversations were in English, but we did not filter by language.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-14, LV-15, LV-23, LV-28", "archive_claims": [ "LV-14", "LV-15", "LV-23", "LV-28" ] }, { "lineage": "success measures are proxies", "field_claims": [ "F20" ], "received": "F20 (B8, Out-of-scope uses s2): \"There is more to an occupation than the sum of its O*NET tasks, and our measures of success (completion, scope, and thumbs feedback) are proxies that do not capture full workplace capabilities.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-20", "archive_claims": [ "LV-20" ] }, { "lineage": "the pipeline: an LLM classifier and its limit", "field_claims": [ "F23", "F24", "F32" ], "received": "F23 (B8, Data creation and processing ¶2 s1): \"The classification of conversations into IWAs was performed by an LLM pipeline designed by Kiran Tomlinson, Siddharth Suri, and Scott Counts.\"; F24 (B8, Limitations ¶2 s1–2): \"LLM-based classifiers are imperfect, so all metrics should be interpreted with caution. See the paper for the exact prompts we used and for interrater reliability with human annotators.\"; F32 (B9, abstract s4): \"We use an LLM-based pipeline to classify the O*NET work activities assisted or performed by AI in each conversation.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-13, LV-22", "archive_claims": [ "LV-13", "LV-22" ] }, { "lineage": "consumer use, not the workplace", "field_claims": [ "F25" ], "received": "F25 (B8, Limitations ¶2 s3–5): \"As these results are based on consumer AI usage, they do not represent the full distribution of AI use in the workplace. For some work tasks, AI tools are directly integrated into specialized software, or specific models are used for their dedicated capability or compliance with data security requirements.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-21, LV-27, LV-37", "archive_claims": [ "LV-21", "LV-27", "LV-37" ] }, { "lineage": "operator, authors and ethics", "field_claims": [ "F27", "F28" ], "received": "F27 (B8, Ethics ¶1): \"Data analysis was conducted with approval from the Microsoft Institutional Review Board.\"; F28 (B8, Contact s1): \"This research was conducted by members of [Microsoft Research](https://www.microsoft.com/en-us/research/).\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-04", "archive_claims": [ "LV-04" ] }, { "lineage": "delegation and assistance", "field_claims": [ "F35" ], "received": "F35 (B9, abstract s7): \"Our methodology also allows us to predict which occupations are more likely to delegate tasks to AI and which are more likely to use AI to assist existing workflows.\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "findings: the nine-month trend", "field_claims": [ "F36" ], "received": "F36 (B9, abstract s8): \"Finally, we see an increase in occupation-level AI applicability over our 9-month data period, driven primarily by people using AI for more varied tasks.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-18", "archive_claims": [ "LV-18" ] }, { "lineage": "neighbours: the Anthropic Economic Index", "field_claims": [ "F39" ], "received": "F39 (B10, footnote 3 (to 'The primitives introduced in this report can help better characterize how AI is changing the work content of occupations.')): \"See also Tomlinson et al (2025) for a related AI applicability score.\"", "receiving_text": "—", "translation": "untranslated: composed as received beside LV-39", "archive_claims": [ "LV-39" ] }, { "lineage": "reception: the encyclopedia", "field_claims": [ "F40", "F41", "F42", "F43", "F44" ], "received": "F40 (B11, rendered page for 'Working_with_AI' (and the full title, and 'AI_applicability_score'), HTTP 404): \"Wikipedia does not have an article with this exact name.\"; F41 (B11, 'Workplace impact of artificial intelligence' § Hazards › Psychosocial › Changes in work practices, s1): \"A 2025 preprint paper based on users' interactions with the AI chatbot Microsoft Copilot identified forty jobs that the authors claimed had high overlaps with the capabilities of AI.\"; F42 (B11, same paragraph, s2): \"Some media outlets used this paper to report on jobs becoming obsolete.\"; F43 (B11, same paragraph, s3): \"Criticism of the paper suggesting that it had misrepresented typical workplace activities to augment AI's current performance.\"; F44 (B11, same paragraph, s4 (Campbell, Transactions of the Royal Historical Society, 10 December 2025)): \"The historian Chris Campbell argued that the \"report's methods, deliberately or otherwise, de-skill historians away from a job that requires high-level and deeply human analytical skills to one that is tasked solely with the retention and provision of knowledge. Under that flawed rubric, it is little wonder that historians have a high AI applicability score.\"\"", "receiving_text": "—", "translation": "untranslated: composed as received (no archive claim in its lineage)", "archive_claims": [] }, { "lineage": "the homonym: Microsoft's AI products", "field_claims": [ "F45", "F46", "F49" ], "received": "F45 (B2, card title and snippet): \"Microsoft Copilot: AI built for work … Core Features: Integrates Chat and Cowork via Home, embeds into Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams), and utilizes Work IQ for organizat...\"; F46 (B3, card title and snippet): \"Microsoft AI - Business Solutions and Tools … Products & Tools: Microsoft IQ (shared business understanding), Microsoft 365 Copilot (embedded app intelligence), Copilot Studio (agent building/extension), Mi...\"; F49 (B6, card title and snippet): \"Microsoft Copilot Plans for Individuals | Microsoft 365 … Microsoft Copilot is an AI-powered assistant designed to help you be more productive. It can help you create content, find information, summarize documents, ema...\"", "receiving_text": "—", "translation": "untranslated: the homonym's card, composed as received as a distinct entity at the name (never a sense of the entity), beside CAP-1-04, CAP-1-09", "archive_claims": [ "CAP-1-04", "CAP-1-09" ] }, { "lineage": "the homonym: Microsoft's own transformation and workforce", "field_claims": [ "F47", "F48" ], "received": "F47 (B4, card title and snippet): \"What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog … Lessons: Start with business outcomes, not technology (tool access doesn't equal transformation; focus on goals and moments that matter); Redesign entire workfl...\"; F48 (B5, card title (video, 1:53; no snippet)): \"Microsoft offers buyouts to 60000 employees as spending on AI grows\"", "receiving_text": "—", "translation": "untranslated: the homonym's card, composed as received as a distinct entity at the name (never a sense of the entity), beside CAP-1-04, CAP-1-09", "archive_claims": [ "CAP-1-04", "CAP-1-09" ] }, { "lineage": "the homonym: Microsoft's models", "field_claims": [ "F50" ], "received": "F50 (B7, card title and snippet): \"Microsoft is building its own AI model as it loosens ties to OpenAI … Feb 12, 2026 — Microsoft just weaponized Copilot with their own independent voice AI. They dropped MAI models and broke free from their OpenAI reliance completely. This isn't ...\"", "receiving_text": "—", "translation": "untranslated: the homonym's card, composed as received as a distinct entity at the name (never a sense of the entity), beside CAP-1-04, CAP-1-09", "archive_claims": [ "CAP-1-04", "CAP-1-09" ] } ], "receiving_texts": "equal-terms case: every field claim composed as received, beside the archive claims it stands with" }, "measure": { "rule": "v0.9 §4.10: KO_B sentences surviving in the KO at similarity ≥ 0.95, with positions (computed as apply_v09.py does: difflib.SequenceMatcher ratio, the best KO sentence for each KO_B sentence)", "v0.9": { "surviving": 0, "of": 12, "positions": [ null, null, null, null, null, null, null, null, null, null, null, null ] }, "opening_run": { "rule": "v0.9 §5.10 (equal-terms case): the field-only KO's sentences as the KO's opening run in their own order", "breach": false }, "blocks": { "rule": "v0.9 §4.1 (equal-terms and mixed cases): the KO does not open with the field as a block or close with the archive as one; sentences typed by the ledgers of the claims they cite (F field-only, A archive-only, M both)", "sequence": "MFFMMMMFFMMMAMAAMAAMMMAMMFMAA", "field_only_head": 0, "archive_only_tail": 2, "breach": false }, "max_similarity": [ 0.334, 0.146, 0.331, 0.298, 0.142, 0.113, 0.3, 0.404, 0.497, 0.589, 0.211, 0.312 ], "ko_sentence_1_max_similarity_to_KO_B": 0.324, "ko_sentence_1_is_KO_B": false }, "homonyms": { "names": [ "Microsoft's own AI products, transformation and workforce (Microsoft Copilot, Microsoft 365 Copilot, Copilot Studio, Azure AI, Fabric, the MAI models; Microsoft's internal AI transformation; its buyouts as AI spending grows): the sense the address gives the words 'Microsoft working with AI'" ], "basis": "One homonym of the company-and-phrase kind (CAP-1-09): the address is the study's short title with its operator's name (CAP-1-05), and the Overview composed the operator itself, its products and its restructuring (T1-1–T1-7), from six of the seven cards (B2–B7, F45–F50), with no notice of the substitution (CAP-1-04). Microsoft's corporate AI is a distinct entity, never a sense of the study's name. 'Microsoft Copilot', the product the homonym's cards present, is the product line whose consumer conversations (as Bing Copilot) the study observes (CAP-1-12): the operator's product and the study's object, distinct from the study itself. Wikipedia names the chatbot 'Microsoft Copilot' (F41) where the study's README and abstract say 'Bing Copilot' (F3, F22): recorded, not resolved.", "distinct_entities": [ "Microsoft (the operator, and the homonym's subject)", "Bing Copilot / Microsoft Copilot (the product whose conversations are the data)", "O*NET 29.0 and the 2018 SOC (the taxonomies)", "the Microsoft Research blog note 'Applicability vs. job displacement' (linked from B1, outside B)" ], "related_entities": [ { "slug": "anthropic-economic-index", "name": "Anthropic Economic Index", "relation": "the census's neighbour (census.json rows[slug == 'microsoft-working-with-ai'].relations[0]); its January 2026 report cites the score as related (B10, F39); its /non row, if built in this batch, is datasets/negative-of-the-negative/v2/rows/anthropic-economic-index.json (not edited)" }, { "slug": "openai-gdpval", "name": "GDPval", "relation": "named beside the entity in the census's worked question 'working-with-ai' (questions.json); a benchmark built and graded by its provider (addendum item 6)" } ], "substitutes": { "T1": [ "Microsoft's corporate AI composed as the entity at its name: 'Microsoft integrates artificial intelligence across its core software ecosystem, enterprise infrastructure, and internal operations, while shifting toward massive AI data center investments and proprietary model development.' (T1-1), with T1-2–T1-7" ] }, "register": { "address_1": { "slug": "microsoft-working-with-ai-aio-20261029", "obs_id": "OBS-25ca6976c874" } } }, "notes": [ "One address, one epoch: microsoft-working-with-ai-aio-20261029 (OBS-25ca6976c874, ADDR-fbfc6406f398, AXN:0AF1.CAPTURE.💚▽♃📚▶️🌀), 'Microsoft Working with AI', seven cards; Google AI Overview (expanded), signed out, incognito, 2026-10-29, seated in the Capture Registry (checked: data/EA-WG-CAPTURES-01.json, registry version 12.396). A NEW address; the census record landed the same day and no card cites it (CAP-1-10).", "Wikipedia: no article for the study under its full title, its short title or its score's name (HTTP 404, 'Wikipedia does not have an article with this exact name'), and no redirect. The one article-space search not rate-limited finds the paper cited in 'Workplace impact of artificial intelligence' (§ Hazards › Psychosocial › Changes in work practices, last edited 23 October 2026), whose paragraph is ledgered (F41–F44): it calls the chatbot 'Microsoft Copilot', reports 'forty jobs', reports media using the paper 'to report on jobs becoming obsolete', and quotes the historian Chris Campbell's criticism. Its third sentence lacks a main verb ('Criticism of the paper suggesting that ...'), quoted as written. Two further searches answered HTTP 429 twice each and were not pursued; no other agent string or route was tried.", "Fetching: B1, the entity's card, found by WebSearch (the Microsoft Research short link ?p=1144520) and fetched with curl and the agent 'Mozilla/5.0 (compatible; Labor-Vector-census)' at 2026-10-29T22:31:37Z (HTTP 200, redirected to the publication page). B8 and B9 are the census's own saved copies (gunzipped sha256 equal to the census's), not refetched; B10 likewise (the census's third-party source). The six homonym cards were recorded from the rail, not searched for or fetched (addendum item 5). No source refused the agent; nothing was retried around a refusal.", "The entity's own card is out of date against its own pages: B1, the operator's publication page, carries the July 2025 abstract (findings by occupation group, the correlation with wage and education) and dates the work 'July 2025', and its canonical slug keeps an earlier title ('working-with-ai-measuring-the-occupational-implications-of-generative-ai'); arXiv v7 (B9, 23 October 2026) states the findings as information work, cross-sector applicability, delegation against assistance and a nine-month trend. The census follows v7 (LV-16–LV-18). Recorded, not resolved; the open block asks it.", "B1 links a Microsoft Research blog note, 'Applicability vs. job displacement: further notes on our recent research on AI and occupations' (F8), the operator's own statement on the limit the census tags first. It is outside B (neither a card, nor a source the census cites, nor Wikipedia) and was not fetched; a later epoch or a census revision may admit it.", "Ledgers unaudited (§3.8): one extractor. Field quotes 50/50 verified by script against the saved texts, the Wikipedia 404 page's text, or the homonym cards' titles and snippets as seated; archive quotes 68/68 verified against #771's text, census.json, questions.json and the registry entry as loaded.", "Defects or out-of-date fields in the census record (reported, not fixed): (1) direction.directs says the method 'separates occupations more likely to delegate tasks to AI from those more likely to use AI to assist existing workflows' (LV-11); its basis says the methodology 'allows us to predict which occupations are more likely to delegate' (F35): the field raises a prediction to a separation. (2) The census does not record the version difference between the operator's publication page (B1, July 2025 abstract) and arXiv v7, nor the page's link to the operator's note on applicability against job displacement (F8), which bears directly on its first negative tag. (3) evidence.operator quotes a sentence fragment ('This research was conducted by members of'); the README sentence continues '[Microsoft Research](...)' (F28): the quote is verbatim but stops before the name it evidences. (4) The census lists the paper's full text as unchecked (LV-40) while recording 'NeurIPS 2026' as venue from the arXiv comment line (F37); the README calls the NeurIPS version 'In press' in its BibTeX. (5) The census row has no disambiguation field, though its record's address is shared, at the common name, with Microsoft's own AI products (CAP-1-09). Not a defect: the README's code reads 'soc_iwa_nonphysical_weights.csv' where its file list names 'soc_iwa_weights_nonphysical.csv'; that is the entity's own README, not the census.", "Observations on the capture (reported, not fixed): findings[1] says the answer composes 'from Microsoft's product pages (cards 2, 3, 6), its blog (card 4) and the press (cards 5, 7)'. Card 5 is ABC News's video (press); card 7 is a Facebook post of unnamed authorship ('Microsoft just weaponized Copilot ...'), which the field shows as a social post, not the press. The inline markers are google.com/goto tokens that cannot be resolved to cards, so the attribution of the answer's lines to particular cards is the registry's reading; the field ledger bears it out for cards 2, 4, 5, 6 and 7 at the grain of their titles and snippets (F45, F47–F50) and cannot see card 3's contribution (its snippet carries Microsoft IQ and Copilot Studio, T carries Azure AI and Fabric). The other findings (the study absent; card 1 first and unused; the address the short title with the operator's name) are borne out by the field (F6, B1).", "Platform-specific record (addendum item 6): the entity is Microsoft's observation of its own product's use, as its own sources state (F22: 'originally collected by Microsoft'; F3 and F31: Bing Copilot), and the census tags it 'NOT independent of the provider' (LV-38). The KO states it in its fourth sentence, the compression in its lede and 'Provider's own product', E_F and E_A in their definitions, and the delta in T_vs_LB_reading ('present in T as subject and absent as observer').", "The archive subset: no archive text names the study (D 0 by reading; the seven D candidates are Hugh Tomlinson, the translator, and citation-list initials). The string pass returned 14 O-class candidates in 10 deposits, all rejected; every archive sentence matching a class pattern (34 in 20 texts) was read, none admitted. #771 (the Directionality of Semantic Labor) was admitted by reading the census's own intersection for the entity, whose terms are the archive's instruments; #1392 (laborvector.org) restates it and is carried as an instance (S4). The hop from both (18 candidates) admitted none. Deposit #3's full_text_path is the live capture registry JSON (S5: its text does not match its record), reported in reading.json.", "Related rows: the census names the Anthropic Economic Index as the entity's neighbour (LV-39; B10, F39); the worked questions name GDPval beside it. Their rows, if built in this batch, are datasets/negative-of-the-negative/v2/rows/anthropic-economic-index.json and openai-gdpval.json (not edited). The Google AI Overview row is datasets/negative-of-the-negative/v2/rows/google-ai-overviews.json (not edited)." ], "composition_v09": { "source": "this row (composed directly; no rebuild/negative-of-the-negative/v09/microsoft-working-with-ai.json)", "composer": "Claude (Opus 5.5), subagent, 2026-10-29", "defining_material_read": [ { "what": "Labor Vector census record, row", "file": "/home/claude/fleet/laborvector/data/labor-census/census.json (rows[slug == 'microsoft-working-with-ai'], organizing_question, intersections, schema_note, evidence_kinds)", "read": "whole" }, { "what": "Labor Vector worked questions", "file": "/home/claude/fleet/laborvector/data/labor-census/questions.json (the two questions naming the entity)", "read": "whole" }, { "what": "Labor Vector record, rendered", "file": "/home/claude/fleet/laborvector/aggregators/microsoft-working-with-ai/index.html", "read": "text, for agreement with the row" }, { "what": "census sources", "file": "/home/claude/fleet/laborvector/data/labor-census/sources/microsoft-working-with-ai--readme.*, microsoft-working-with-ai--arxiv.*, anthropic-economic-index--jan-2026.* (with manifest.json)", "read": "README and arXiv page whole; the AEI report at its footnote 3 and the sentence it attaches to" }, { "what": "capture record", "file": "data/EA-WG-CAPTURES-01.json (microsoft-working-with-ai-aio-20261029)", "read": "transcript, transcript_raw, cite_list, d, reading, findings, analysis, q_kind, per_note, per_dim, originator" }, { "dep": 771, "file": "data/texts/AXN-02D1-text.md", "read": "whole" }, { "dep": 1392, "file": "data/texts/AXN-0581-text.md", "read": "at the DSL, task-origin and ULD passages (grep), for the S4 instance" }, { "what": "specification", "file": "data/texts/AXN-0700-text.md", "read": "whole" }, { "what": "exemplars", "file": "datasets/negative-of-the-negative/v2/rows/olmotrace.json, datasets/negative-of-the-negative/v2/rows/wikiwho.json and their ledgers, configs and reading files", "read": "structure and the parts the brief names" } ], "cuts": [], "notes": "equal case; KO 29 sentences, all 50 field and 68 archive claims cited; P body 346 words; KO_B 12 sentences, 0 surviving at ≥ 0.95" }, "field_by_origin": { "note": "The field B for a homonym row (addendum BRIEF-NEW-ROW-LABOR.md item 5): the cards that concern the entity (B1), its own pages (B8, B9) and Wikipedia (B11), with the census's third-party source apart (B10); the homonym's cards (B2–B7) are recorded with their sense, title and snippet and counted apart.", "cards": [ "B1", "B2", "B3", "B4", "B5", "B6", "B7" ], "entity_cards": [ "B1" ], "homonym_cards": [ "B2", "B3", "B4", "B5", "B6", "B7" ], "own_pages": [ "B8", "B9" ], "wikipedia": [ "B11" ], "census_third_party": [ "B10" ], "unreachable": [], "title_only": [ "B2", "B3", "B4", "B5", "B6", "B7" ], "wikipedia_state": "no article: 'Working with AI: Measuring the Applicability of Generative AI to Occupations', 'Working with AI' and 'AI applicability score' all absent (raw wikitext HTTP 404; rendered pages HTTP 404 'Wikipedia does not have an article with this exact name'), no redirect; an article-space search for \"Working with AI\" applicability returns one article, 'Workplace impact of artificial intelligence', which cites the paper in § Hazards › Psychosocial › Changes in work practices (last edited 23 October 2026), ledgered; two further searches rate-limited (HTTP 429), retried once each, not pursued (2026-10-29)", "counts": { "cards": 7, "entity_cards": 1, "cards_fetched": 1, "homonym_cards_not_fetched": 6, "cards_unreachable": 0, "own_pages": 2, "own_pages_from_census_copy": 2, "own_pages_fetched_by_row": 0, "census_third_party": 1, "wikipedia": "no article; 1 citing article fetched and ledgered (B11)" }, "sources": [ { "id": "B1", "origin": "card", "role": "Overview card 1 (the entity's card)", "card": "Microsoft — Working with AI: Measuring the Applicability of Generative AI ... - Microsoft", "url": "https://www.microsoft.com/en-us/research/publication/working-with-ai-measuring-the-occupational-implications-of-generative-ai/", "kind": "Overview card (the operator's publication page at Microsoft Research: authors, 'July 2025', the abstract of the paper's first version, links to GitHub, the preprint and a blog note)", "files": { "B1.html": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B1.html", "sha256": "e1f8e1b60bb379f6fc8a8e4f53f969a0780fc9e44fa9cf360bf12524fbf9a074" }, "B1.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B1.txt", "sha256": "c1ce357664ec70ea34e7b9d18aa341e1a9b49aa283a09afe10c71ae5429f374c" } }, "fetched": { "at": "2026-10-29T22:31:37Z", "http": 200, "url": "https://www.microsoft.com/en-us/research/?p=1144520", "final_url": "https://www.microsoft.com/en-us/research/publication/working-with-ai-measuring-the-occupational-implications-of-generative-ai/", "agent": "Mozilla/5.0 (compatible; Labor-Vector-census)", "raw": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B1.html", "raw_sha256": "e1f8e1b60bb379f6fc8a8e4f53f969a0780fc9e44fa9cf360bf12524fbf9a074", "text": [ "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B1.txt" ] }, "names_operator": "yes ('Microsoft Research'; the five authors; the groups 'AI Interaction and Learning', 'Economics and Computation', 'AI Economy Institute', 'IDEAS Research'; 'Microsoft Research Redmond')", "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, laborvector, Labor Vector, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)", "note": "URL found by WebSearch on the card's title and site (the result 'https://www.microsoft.com/en-us/research/?p=1144520'), which redirects to the publication page; its is the card's title in full ('Working with AI: Measuring the Applicability of Generative AI to Occupations - Microsoft Research'). The page's canonical slug keeps an earlier title, 'working-with-ai-measuring-the-occupational-implications-of-generative-ai'. The card's snippet ('We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and ...') is the page's abstract (F6); the card's date 'Jul 10, 2025' is arXiv v1's, the page giving 'July 2025'. The page's abstract is the July 2025 text, not arXiv v7's (B9): it reports AI applicability by occupation group and the correlation with wage and education; v7's abstract reports information work, cross-sector applicability, delegation against assistance and a nine-month trend. The page links a Microsoft Research blog note, 'Applicability vs. job displacement: further notes on our recent research on AI and occupations' (F8), outside B and not fetched.", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 1 } ], "card_title": "Working with AI: Measuring the Applicability of Generative AI ... - Microsoft", "card_snippet": "Jul 10, 2025 — We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and ...", "card_note": null }, { "id": "B2", "origin": "card", "role": "Overview card 2 (the homonym's)", "card": "Microsoft — Microsoft Copilot: AI built for work", "url": null, "kind": "Overview card, the homonym's: Microsoft's AI products: Microsoft Copilot for work (Microsoft 365 apps, Work IQ)", "homonym": true, "sense": "Microsoft's AI products: Microsoft Copilot for work (Microsoft 365 apps, Work IQ)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title and snippet)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 2 } ], "card_title": "Microsoft Copilot: AI built for work", "card_snippet": "Core Features: Integrates Chat and Cowork via Home, embeds into Microsoft 365 apps (Word, Excel, PowerPoint, Outlook, Teams), and utilizes Work IQ for organizat...", "card_note": null }, { "id": "B3", "origin": "card", "role": "Overview card 3 (the homonym's)", "card": "Microsoft — Microsoft AI - Business Solutions and Tools", "url": null, "kind": "Overview card, the homonym's: Microsoft's AI products: business solutions and tools (Microsoft IQ, Microsoft 365 Copilot, Copilot Studio)", "homonym": true, "sense": "Microsoft's AI products: business solutions and tools (Microsoft IQ, Microsoft 365 Copilot, Copilot Studio)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title and snippet)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 3 } ], "card_title": "Microsoft AI - Business Solutions and Tools", "card_snippet": "Products & Tools: Microsoft IQ (shared business understanding), Microsoft 365 Copilot (embedded app intelligence), Copilot Studio (agent building/extension), Mi...", "card_note": null }, { "id": "B4", "origin": "card", "role": "Overview card 4 (the homonym's)", "card": "The Official Microsoft Blog — What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog", "url": null, "kind": "Overview card, the homonym's: Microsoft's own AI transformation: lessons from the company's internal deployment (the Official Microsoft Blog)", "homonym": true, "sense": "Microsoft's own AI transformation: lessons from the company's internal deployment (the Official Microsoft Blog)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 4 } ], "card_title": "What we’ve learned from Microsoft's own AI transformation - The Official Microsoft Blog", "card_snippet": "Lessons: Start with business outcomes, not technology (tool access doesn't equal transformation; focus on goals and moments that matter); Redesign entire workfl...", "card_note": null }, { "id": "B5", "origin": "card", "role": "Overview card 5 (the homonym's)", "card": "YouTube·ABC News — Microsoft offers buyouts to 60000 employees as spending on AI grows", "url": null, "kind": "Overview card, the homonym's: Microsoft's workforce: buyouts offered to 60,000 employees as AI spending grows (ABC News video, 1:53) — video card, admitted through title (no snippet on the rail)", "homonym": true, "sense": "Microsoft's workforce: buyouts offered to 60,000 employees as AI spending grows (ABC News video, 1:53)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 5 } ], "card_title": "Microsoft offers buyouts to 60000 employees as spending on AI grows", "card_snippet": null, "card_note": "video, 1:53" }, { "id": "B6", "origin": "card", "role": "Overview card 6 (the homonym's)", "card": "Microsoft — Microsoft Copilot Plans for Individuals | Microsoft 365", "url": null, "kind": "Overview card, the homonym's: Microsoft's AI products: Copilot plans for individuals (Microsoft 365)", "homonym": true, "sense": "Microsoft's AI products: Copilot plans for individuals (Microsoft 365)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title and snippet)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 6 } ], "card_title": "Microsoft Copilot Plans for Individuals | Microsoft 365", "card_snippet": "Microsoft Copilot is an AI-powered assistant designed to help you be more productive. It can help you create content, find information, summarize documents, ema...", "card_note": null }, { "id": "B7", "origin": "card", "role": "Overview card 7 (the homonym's)", "card": "Facebook — Microsoft is building its own AI model as it loosens ties to OpenAI", "url": null, "kind": "Overview card, the homonym's: Microsoft's models: its own MAI models and its ties to OpenAI (a Facebook post)", "homonym": true, "sense": "Microsoft's models: its own MAI models and its ties to OpenAI (a Facebook post)", "files": {}, "fetched": { "unreachable": false, "not_fetched": true, "reason": "the homonym's card (addendum BRIEF-NEW-ROW-LABOR.md item 5): the name 'Microsoft Working with AI' read as Microsoft's own AI products and workforce. Recorded with its sense, title and snippet; not searched for or fetched as field, since the field B for this row is the cards that concern the entity, its own pages and Wikipedia. Its title and snippet are its only text in this row." }, "names_operator": "yes ('Microsoft' in the card's title and snippet)", "names_archive": "no (none of the archive's names in the card's title and snippet)", "surfaced_by_T": [ { "slug": "microsoft-working-with-ai-aio-20261029", "card": 7 } ], "card_title": "Microsoft is building its own AI model as it loosens ties to OpenAI", "card_snippet": "Feb 12, 2026 — Microsoft just weaponized Copilot with their own independent voice AI. They dropped MAI models and broke free from their OpenAI reliance completely. This isn't ...", "card_note": null }, { "id": "B8", "origin": "own_page", "entity": "microsoft-working-with-ai", "url": "https://raw.githubusercontent.com/microsoft/working-with-ai/main/README.md", "kind": "own page: the results repository's README (github.com/microsoft/working-with-ai, raw), the census's source microsoft-working-with-ai--readme", "files": { "B8.html": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B8.html", "sha256": "b98be795b8a96490ecf0e50329d5867958f11f0ac78939b4144422a1929433b3" }, "B8.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B8.txt", "sha256": "b98be795b8a96490ecf0e50329d5867958f11f0ac78939b4144422a1929433b3" } }, "fetched": { "census_copy": { "src": "microsoft-working-with-ai--readme", "file": "/home/claude/fleet/laborvector/data/labor-census/sources/microsoft-working-with-ai--readme.html.gz (gunzipped) and .txt", "sha256": "b98be795b8a96490ecf0e50329d5867958f11f0ac78939b4144422a1929433b3", "read": "2026-10-29", "fetched": "2026-10-29T21:49:15Z", "http": "200", "agrees": true }, "refetched": false, "note": "copied from the census's saved copy; the gunzipped bytes' sha256 equals the census's (manifest.json); the README is plain text, so B8.txt is the same bytes" }, "http": 200, "sha256_html": "b98be795b8a96490ecf0e50329d5867958f11f0ac78939b4144422a1929433b3", "census_source": "microsoft-working-with-ai--readme", "census_basis": "operator", "read": "2026-10-29", "surfaced_by_T": false, "admitted": "own page the census record cites (operator's ruling of 2026-10-29: 'as against their wikipedia articles if they exist, their own pages, etc.'); not surfaced by T", "names_operator": "yes ('This research was conducted by members of [Microsoft Research]'; the five authors; 'Microsoft Institutional Review Board')", "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, laborvector, Labor Vector, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)" }, { "id": "B9", "origin": "own_page", "entity": "microsoft-working-with-ai", "url": "https://arxiv.org/abs/2507.07935", "kind": "own page: the paper's arXiv abstract page (arXiv:2507.07935, v7, 23 October 2026), the census's source microsoft-working-with-ai--arxiv", "files": { "B9.html": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B9.html", "sha256": "8315551565cd59866e841acec8204a2a9579093b1d0266146968102fd6d43796" }, "B9.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B9.txt", "sha256": "517fc35ecee104f481ecf77d7d8d75a90cec3317cdfebf11ee7ad893e8171ed3" } }, "fetched": { "census_copy": { "src": "microsoft-working-with-ai--arxiv", "file": "/home/claude/fleet/laborvector/data/labor-census/sources/microsoft-working-with-ai--arxiv.html.gz (gunzipped) and .txt", "sha256": "8315551565cd59866e841acec8204a2a9579093b1d0266146968102fd6d43796", "read": "2026-10-29", "fetched": "2026-10-29T21:49:15Z", "http": "200", "agrees": true }, "refetched": false, "note": "copied from the census's saved copy; the gunzipped bytes' sha256 equals the census's (manifest.json)" }, "http": 200, "sha256_html": "8315551565cd59866e841acec8204a2a9579093b1d0266146968102fd6d43796", "census_source": "microsoft-working-with-ai--arxiv", "census_basis": "operator", "read": "2026-10-29", "surfaced_by_T": false, "admitted": "own page the census record cites (operator's ruling of 2026-10-29); not surfaced by T", "names_operator": "yes (the five authors; 'Microsoft Bing Copilot' in the abstract)", "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, laborvector, Labor Vector, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)" }, { "id": "B10", "origin": "census_third_party", "url": "https://www.anthropic.com/research/anthropic-economic-index-january-2026-report", "kind": "third-party source the census read (basis 'third_party'): the Anthropic Economic Index report 'Economic primitives' (January 2026), footnote 3", "files": { "B10.html": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B10.html", "sha256": "8d0d80679aaf24b9fecdadb0d26bd44b50e8d90a2c4e7d01172d7b969f2d9d8d" }, "B10.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B10.txt", "sha256": "1b5b928b9f12d14f47043b64b2b8ff977cc9fd69f39165246ab62fa73686ffcb" } }, "fetched": { "census_copy": { "src": "anthropic-economic-index--jan-2026", "file": "/home/claude/fleet/laborvector/data/labor-census/sources/anthropic-economic-index--jan-2026.html.gz (gunzipped) and .txt", "sha256": "8d0d80679aaf24b9fecdadb0d26bd44b50e8d90a2c4e7d01172d7b969f2d9d8d", "read": "2026-10-29", "http": "200", "agrees": true }, "refetched": false }, "http": 200, "sha256_html": "8d0d80679aaf24b9fecdadb0d26bd44b50e8d90a2c4e7d01172d7b969f2d9d8d", "census_source": "anthropic-economic-index--jan-2026", "census_basis": "third_party", "read": "2026-10-29", "surfaced_by_T": false, "admitted": "census_third_party, kept apart from cards, own pages and Wikipedia (addendum item 3); only the sentence that names the entity is ledgered", "names_operator": "the authors only ('Tomlinson et al (2025)')", "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, laborvector, Labor Vector, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)" }, { "id": "B11", "origin": "wikipedia", "result": "no article; no redirect; the paper is cited in one article, 'Workplace impact of artificial intelligence', whose paragraph on it is ledgered", "url": "https://en.wikipedia.org/wiki/Workplace_impact_of_artificial_intelligence", "target": "Workplace impact of artificial intelligence § Hazards › Psychosocial › Changes in work practices", "redirect": null, "section": "== Hazards == / === Psychosocial === / ==== Changes in work practices ==== (the paragraph beginning 'A 2025 preprint paper based on users' interactions with the AI chatbot Microsoft Copilot')", "kind": "Wikipedia: no article for the entity; the article that cites it (last edited 23 October 2026)", "files": { "B11.wiki": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.wiki", "sha256": "c38ff1b52fd14944977527be9997302f71cfdef83c584f2c7c25394db7caf1e8" }, "B11.html": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.html", "sha256": "a2483d45f8ad107588d803b028c8991821a719a67dd911c206f2362c45755ce1" }, "B11.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.txt", "sha256": "d5c169b1494867c043af51f3c2f163d115ed073a813a1ac941b18742ad712a7a" }, "W-404.txt": { "path": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-404.txt", "sha256": "f9572af637c0e131c9171a3e289277c15b00af20966c39e1b1a6abdab96ff9bd" } }, "requests": [ { "at": "2026-10-29T22:32:02Z", "url": "https://en.wikipedia.org/w/index.php?title=Working_with_AI:_Measuring_the_Applicability_of_Generative_AI_to_Occupations&action=raw", "http": 404, "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-Working_with_AI__Measuring_the.raw.txt", "sha256": "4bf1a6a8613bd2ed2285e19e15f0ebed7cb3b3523924445c7c6cbb4f2e60f919" }, { "at": "2026-10-29T22:32:02Z", "url": "https://en.wikipedia.org/wiki/Working_with_AI:_Measuring_the_Applicability_of_Generative_AI_to_Occupations", "http": 404, "result": "'Wikipedia does not have an article with this exact name'", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-Working_with_AI__Measuring_the.html", "sha256": "7a1d1059eb148b7e236aeafa8bfa658102c159a86f67d86088fa42d85f09c102" }, { "at": "2026-10-29T22:32:05Z", "url": "https://en.wikipedia.org/w/index.php?title=Working_with_AI&action=raw", "http": 404, "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-Working_with_AI.raw.txt", "sha256": "6cf7fc698644adf10a8fd5b76c8f1a7a098c5dc6c65e582927f935d9997ba5f0" }, { "at": "2026-10-29T22:32:05Z", "url": "https://en.wikipedia.org/wiki/Working_with_AI", "http": 404, "result": "'Wikipedia does not have an article with this exact name. Please search for Working with AI in Wikipedia to check for alternative titles or spellings.' (text: W-404.txt)", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-Working_with_AI.html", "sha256": "dae8f617055aac049fe70d9a0adf9fefa87674904dc7631babe5c421787b052d" }, { "at": "2026-10-29T22:32:07Z", "url": "https://en.wikipedia.org/w/index.php?title=AI_applicability_score&action=raw", "http": 404, "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-AI_applicability_score.raw.txt", "sha256": "42eb66f1fa86d33f892f252d318a0233bf9458392ed6e463e0473ee91f1eace5" }, { "at": "2026-10-29T22:32:07Z", "url": "https://en.wikipedia.org/wiki/AI_applicability_score", "http": 404, "result": "'Wikipedia does not have an article with this exact name'", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-AI_applicability_score.html", "sha256": "843e1ef0ee4962b75ee931c31abbbf89d91699bdcdb9baca5df8ac350059fb63" }, { "at": "2026-10-29T22:32:08Z", "url": "https://en.wikipedia.org/w/api.php?action=query&list=search&srsearch=%22Working+with+AI%22+applicability&srnamespace=0&format=json", "http": 200, "result": "totalhits 1: 'Workplace impact of artificial intelligence' (snippet: the citation of Tomlinson et al. 2025)", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-search1.json", "sha256": "6f7af960ebcf467c98d652cbc238a3de0fe9786c7be25ae1bcf20f24a3b81cd6" }, { "at": "2026-10-29T22:32:08Z / 22:32:45Z", "url": "https://en.wikipedia.org/w/api.php?action=query&list=search&srsearch=%22Working+with+AI%22+Tomlinson&srnamespace=0&format=json", "http": 429, "result": "rate-limited ('You are making too many requests to the API.'); one retry after a 15-second wait, again 429; not pursued", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-search2.json", "sha256": "c8184edc74f6f6102907e1bbb656acf1f36f5fddc30e7ec5b890ccce8f13ffd3" }, { "at": "2026-10-29T22:32:09Z / 22:32:50Z", "url": "https://en.wikipedia.org/w/api.php?action=query&list=search&srsearch=%22AI+applicability+score%22&srnamespace=0&format=json", "http": 429, "result": "rate-limited; one retry, again 429; not pursued", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-search3.json", "sha256": "8af5663114b9a2c3b9e4bd96867e9799fd829000d812c2bb4a1f6553bf22d358" }, { "at": "2026-10-29T22:32:56Z", "url": "https://en.wikipedia.org/w/index.php?title=Workplace_impact_of_artificial_intelligence&action=raw", "http": 200, "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.wiki", "sha256": "c38ff1b52fd14944977527be9997302f71cfdef83c584f2c7c25394db7caf1e8" }, { "at": "2026-10-29T22:33:01Z", "url": "https://en.wikipedia.org/wiki/Workplace_impact_of_artificial_intelligence", "http": 200, "result": "rendered page; 'This page was last edited on 23 October 2026, at 12:50 (UTC).'", "file": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.html", "sha256": "a2483d45f8ad107588d803b028c8991821a719a67dd911c206f2362c45755ce1" } ], "fetched": { "agent": "Mozilla/5.0 (compatible; Labor-Vector-census)", "raw": "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.html", "raw_sha256": "a2483d45f8ad107588d803b028c8991821a719a67dd911c206f2362c45755ce1", "text": [ "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/B11.txt", "datasets/negative-of-the-negative/v2/intake/microsoft-working-with-ai-20261029/W-404.txt" ] }, "http": 200, "sha256_html": "a2483d45f8ad107588d803b028c8991821a719a67dd911c206f2362c45755ce1", "checked": "2026-10-29T22:32:02Z–22:33:01Z", "surfaced_by_T": false, "names_operator": "the paper's authors in the citation (Tomlinson, Jaffe, Wang, Counts, Suri); 'Microsoft Copilot' as the chatbot; not Microsoft Research", "names_archive": "no (none of the Crimson Hexagonal Archive, Alexanarch, Lee Sharks, laborvector, Labor Vector, machinemediation, EA-WG-CAPTURES-01 in the saved text, card title and snippet)", "note": "No article under the paper's title, its short title or its score's name, and no redirect. The one article-space search that was not rate-limited finds a single article citing the paper. Its paragraph is Wikipedia's account of the entity and is ledgered (F43–F46); the 404 is ledgered as a fact of the row (F42). Two further searches were rate-limited (429), retried once each after a wait, and not pursued; no other agent string or route was tried." } ] } }