Left politics and labor / Labor Vector
Labor Vector
What it is
DSL measures across five points of a transaction, "capacity, task-origin, retrieval, output, attribution (C → T → R → O → A)", and shows that erasure can occur at each: "Erasure at the first layer is deepest; at the last, most visible." Its principle is "Provenance is not only claim-origin. Provenance is task-origin." Its core is the Directionality Stack (DS-6), fixed at six places, and the site bounds its quick prompt as "a screening estimate, not a measurement".
From its /non article, sentences 7, 10, 8, 11, 12
The site and its record
It was deposited in the archive on 2026-06-05 as software and executable apparatus, its source recovered from the live surface on 2026-08-04. Its colophon names deposit #771, the specification of DSL, and records the page as "drafted with Claude (TACHYON), MANUS-approved". Its field measurement reads the Self-Audit Module's dissolution by Google AI Overview on 13 June 2026 as "DSL measured at 1.00".
From its /non article, sentences 1, 3, 4, 6, 14
The meta-aggregator
Its census asks one question of every record: "who supplies the labor, who directs it, and who receives its gains — and what kind of evidence does this resource bring to that question?" Its intersections run from labor producing AI to worker power ("bargaining, organizing, consent, refusal, contestation and enforceable protections"), and it publishes the census as data. It states its limits: "A record states only what its resource measures or reports." and "No record in the census measures the distribution of productivity gains between workers, employers, platforms and model providers."
From its /non article, sentences 19, 20, 21
Whose record, who holds it, the losing side
It keeps the records of 15 labor-and-AI resources, held by Lee Sharks under CC BY 4.0, and records no losing side: its organizing question is who supplies, directs and receives the gains of labor. The archive reads its worker power intersection as where it meets the left census, and states "It shares this census's editor, so it is no independent holder in #1725's sense." At the address "laborvector" on 31 October 2026 the Overview gave "THE FRAMEWORK, ATTRIBUTED", and the meta-aggregator was absent: "the site is composed as its definition surface alone."
From its /non article, sentences 25, 26, 29
This article is derived from the entity's /non article (EA-NEGONT-02, the knowledge object composed under v0.9): every fact in it is stated there, and each paragraph names the sentences it rests on.
| Kind | the canonical definition surface for Directionality of Semantic Labor and the interoperable metric program, carrying a meta-aggregator on labor and AI (a census of 15 resources) |
|---|---|
| Operator | Lee Sharks (Semantic Economy Institute / Crimson Hexagonal Archive), the editor of this census |
| Unit of record | in the meta-aggregator: a resource that observes an intersection of labor and AI, formalized as an entity and tagged by intersection and kind of evidence; on the home page: a DSL reading of a transcript |
| What it records | resources that observe the intersections of labor and AI, from data workers' own inquiries and contract inventories to employment trackers and records of how people use AI systems, read for who supplies the labor, who directs it and who receives its gains, and on what evidence |
| Whose record it keeps | the census's records of 15 labor-and-AI resources (Data Workers' Inquiry, Turkopticon, Fairwork Cloudwork Ratings, Negotiating Tech, the ILO platform policy tracker and others), each with its own unit, operator and evidence kind |
| Who holds it | Lee Sharks, under CC BY 4.0; the census is published as data (census.json and questions.json) with every saved source |
| The losing side | not recorded: its organizing question is who supplies, directs and receives the gains of labor, and each record states what its resource cannot answer |
| Method | a census compiled from saved sources, each resource formalized as an entity and read against one question; worked questions name the resources that hold the evidence and what none of them records |
| Coverage | geography: not stated as a scope; the resources range from global trackers to U.S. contract inventories; period: compiled 2026-10-29 |
| Count | value: 15 resources across five intersections; as_of: 2026-10-29 |
| Holdings | 15 resource records across five intersections (labor producing AI, AI governing labor, labor performed with AI, gains and losses, worker power) and twelve kinds of evidence; worked questions with the resources that answer them and what none of them records; the DSL metric and its interoperable program on the home page |
| Stated limits | 'A record states only what its resource measures or reports.'; 'No record in the census measures the distribution of productivity gains between workers, employers, platforms and model providers.' |
| Data access | public pages; the census as data (census.json, questions.json) with saved sources |
| Licence | CC BY 4.0 |
| Lifecycle | state: active; detail: the DSL surface deposit #771; the meta-aggregator compiled 2026-10-29 with 15 resources |
| Founded | 2026 (DSL defined by Lee Sharks) |
What its record can settle
- which labor-and-AI resources hold which kind of evidence (exposure estimate, employment outcome, worker testimony, negotiated provision and the rest)
- which resources record organizing and negotiated protections (its worker power intersection)
What it cannot settle
- how a left organization holds or loses internal critique
- the record of left movements, parties or strikes outside labor and AI
Kept apart
- NOT labor economics: its page reads 'DSL measures semantic, not wage, labor'
- NOT a census of left politics: its census covers labor and AI
- NOT independent of this census: it has the same editor
Captures: the composition layer's account
- «laborvector» on Google AI Overview, 2026-10-31the composition layer's account of the entity at its name, seated in the Capture Registry (AXN:0B63)
The /non row
Evidence
- kind (the resource's own statement):
Labor Vector · laborvector.org — canonical definition surface for Directionality of Semantic Labor and the interoperable metric program.
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - operator (the resource's own statement):
Defined by Lee Sharks (2026) · ORCID · Semantic Economy Institute / Crimson Hexagonal Archive
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - license (the resource's own statement):
© Lee Sharks 2026 · CC BY 4.0 · ORCID 0009-0000-1599-0703
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - distinct (the resource's own statement):
Distinct from: labor economics (DSL measures semantic, not wage, labor)
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - deposit (the resource's own statement):
laborvector.org deposit #771
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - observed (the resource's own statement):
surface_observed_at: 2026-08-15T09:40:32Z
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - aggregator_home (the resource's own statement):
15 resources across five intersections, compiled 2026-10-29.
https://www.laborvector.org/ · read 2026-10-31 · http 200 · sha256 68b21e12c0668d46… - record_object (the resource's own statement):
The resources that observe the intersections of labor and AI, from data workers' own inquiries and contract inventories to employment trackers and records of how people use AI systems, each formalized as an entity.
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - question (the resource's own statement):
Every record answers one 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?
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - count (the resource's own statement):
15 resources, compiled 2026-10-29.
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - worker_power (the resource's own statement):
Worker power : bargaining, organizing, consent, refusal, contestation and enforceable protections
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - limit_record (the resource's own statement):
A record states only what its resource measures or reports.
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - limit_gains (the resource's own statement):
No record in the census measures the distribution of productivity gains between workers, employers, platforms and model providers.
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - data (the resource's own statement):
The census as data: census.json and questions.json , with every saved source in data/labor-census/sources .
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba… - intersections (the resource's own statement):
Five intersections Labor producing AI
https://www.laborvector.org/aggregators/ · read 2026-10-31 · http 200 · sha256 912d0e8aafbf93ba…
Unchecked
census.json and questions.json themselves (not among the saved sources); the Questions and Captures pages
Cite: Lee Sharks, ed. "Labor Vector." Mary Lee Labor meta-aggregator: left politics and labor, maryleelabor.org, compiled 2026-10-31. https://www.maryleelabor.org/aggregators/labor-vector