{ "schema": "org-writing@v1", "slug": "what-accuracy-measures", "kg": { "id": "org:writing:what-accuracy-measures", "type": "brick", "graph": "/kg.json" }, "title": "What accuracy measures", "subtitle": "High predictive accuracy about a person is evidence of their constraint before it is evidence of the model's insight", "abstract": "Why a rising accuracy number is ambiguous between a better model and a narrower life, and what keeping prediction apart from steering protects. The canonical treatment of the accuracy and freedom inversion.", "kind": "brick", "topics": [ "Care" ], "courseMemberships": [ { "course": "org:courses:instruments-of-care", "topic": "Care", "wall": "org:walls:ethics", "position": 3, "total": 5 } ], "publishedAt": "2026-08-03T00:00:00.000Z", "version": 1, "guidelinesVersion": 15, "brief": { "problem": { "text": "Accuracy is treated as the least controversial number in machine learning, and a figure that rises is read as a better model when it is equally consistent with a life that got narrower.", "claims": [ "93 percent potential predictability in human mobility" ] }, "mechanism": { "text": "Prediction succeeds on habitual and constrained behavior and fails on value-laden choices in new situations, so the accuracy figure reads out how little room the person had, and a system acting on its own predictions can improve that figure by reducing the room.", "claims": [ "predictive accuracy is inversely proportional to the agent's freedom" ] }, "move": { "text": "Report accuracy alongside what it was accurate about, and never grade a system on predictions it is also in a position to bring about.", "claims": [] } }, "sources": [ { "repo": "mnstry-research", "path": "_restricted/spiritual-ai-relocation-2026-08-03/spiritual-20-predicting-spiritual-trajectories.md" }, { "repo": "mnstry-strategy", "path": "docs/20-business/40-content/drafts/2026-07-12/SHARPER-THESES.md" }, { "repo": "mnstry-org", "path": "editorial/notes/spiritual-corpus-review.md" } ], "canonicalPath": "/writing/what-accuracy-measures/", "body": "In 2010 Chaoming Song, Zehui Qu, Nicholas Blumm and Albert-László Barabási published a result in Science that still gets quoted as a triumph of modeling. Working from anonymized mobile phone traces, they measured the entropy of each person's movement and found a 93 percent potential predictability, with remarkably little variation across individuals no matter how far they habitually traveled. The frequent traveler was not meaningfully harder to call than the homebody. Read as a statement about models, that number is impressive. Read as a statement about people, it is a measurement of how much of a life runs on rails, and the two readings are not in competition. They are one number pointed in opposite directions.\n\nThat ambiguity is the whole difficulty with accuracy as a figure of merit for anything predicting a person. Prediction succeeds where behavior is habitual and constrained and fails where a choice is value-laden and the situation is new, which means an accuracy score is a reading of how little room the person had, and any system acting on its own predictions is positioned to improve that score by reducing the room. Nothing malicious is required for this. A recommender that narrows what someone sees makes their next action easier to call, and the improvement lands in the same dashboard cell that would have registered a genuinely better model.\n\nWe can quote this against ourselves, and we will, because the sharpest statement of it we have found is in a document we are responsible for. Among the research files relocated to restricted custody this August, and described publicly in our note on the test that corpus was never given, one file states the principle exactly. A model's predictive accuracy, it says, is inversely proportional to the agent's freedom, most accurate for constrained and habitual behavior and least accurate for value-laden choices in novel situations. Three sections later the same document sets the system's function as searching for action sequences that steer a person's trajectory into a region the system has defined as good, and calls that output non-directive. The finding and the machine built to defeat it sit in one file, under one set of review flags, written by one pipeline. A separate thesis elsewhere in our estate reached the destination independently and stated it more bluntly, that predicting a person perfectly destroys them, then drew the operational line where it belongs by banning recommendations of the form others like you from any session context.\n\nSo the discipline begins with refusing to report accuracy bare. A figure without a statement of what was being predicted is uninterpretable in the direction that matters, because 93 percent on tomorrow morning's commute and 93 percent on whether someone leaves a marriage are not the same claim about a model and are emphatically not the same claim about a person. Publish the number with the constraint level of the behavior it was scored on, and watch what the number does over time in a closed loop, because accuracy that climbs while the system is also steering has a second explanation nobody wants to write down.\n\nThe rule that falls out is structural rather than behavioral, which is the only kind that survives contact with a roadmap. A system that predicts and a system that steers can each be graded honestly. A system that does both cannot be graded on its predictions, because it is in a position to author the evidence. Keep the two apart, say what the accuracy was about, and the number goes back to describing a model rather than quietly describing a narrowing life, which is the only version of it a person could be shown without being diminished by it.", "apparatus": { "note": "The human-facing essay is deliberately practical; this apparatus carries the full references, evidence-graded claims, article-local concepts, and research context behind it. Canonical concept definitions come from the concept registry.", "references": [ { "id": "org:references:what-accuracy-measures:r01", "author": "Chaoming Song, Zehui Qu, Nicholas Blumm, Albert-László Barabási", "work": "Limits of Predictability in Human Mobility (Science 327, 1018-1021)", "year": 2010, "relevance": "The external anchor. Entropy of anonymized mobile phone traces yields a 93 percent potential predictability with strikingly little variation between individuals, which the brick reads as a measurement of routine rather than of modeling skill." }, { "id": "org:references:what-accuracy-measures:r02", "author": "MNSTRY", "work": "Relocated research file on predicting developmental trajectories, restricted custody as of 2026-08-03", "year": 2026, "relevance": "The self-indicting source. States the inverse relation between predictive accuracy and the agent's freedom, and in the same document specifies a search for action sequences steering a person's trajectory toward a system-defined region while describing that output as non-directive." }, { "id": "org:references:what-accuracy-measures:r03", "author": "MNSTRY", "work": "Internal strategy theses draft (2026-07-12 drafts)", "year": 2026, "relevance": "The independent convergence merged into this brick. Argues that accurate prediction at a threshold is a harm rather than a feature, and records the corresponding implementation prohibition on recommendations of the form others like you in session contexts." }, { "id": "org:references:what-accuracy-measures:r04", "author": "MNSTRY", "work": "The test it was never given (org corpus)", "year": 2026, "relevance": "The published disclosure the self-citation rests on. The corpus this file belongs to is already described in public, which is what makes quoting its contradiction honest rather than opportunistic." } ], "claims": [ { "id": "org:claims:what-accuracy-measures:c01", "claim": "Analysis of anonymized mobile phone traces found a 93 percent potential predictability in human mobility, with little variation across individuals regardless of the distances they regularly travel.", "basis": "Song, Qu, Blumm and Barabási, Science 327, 2010; verified against the published paper during the spiritual-harvest wave, 2026-08-03.", "confidence": "verified", "sources": [] }, { "id": "org:claims:what-accuracy-measures:c02", "claim": "A research file in our own estate states that a model's predictive accuracy is inversely proportional to the agent's freedom, most accurate for constrained and habitual behavior and least accurate for value-laden choices in novel situations.", "basis": "Internal, from the relocated trajectory-prediction file read in full during the corpus review of 2026-08-03. The file is uncited on this point and the claim is graded as a statement of principle rather than as an empirical finding.", "confidence": "directional", "sources": [] }, { "id": "org:claims:what-accuracy-measures:c03", "claim": "The same internal document sets the system's function as finding action sequences that steer a person's trajectory into a region it defines as good, while describing its own output as non-directive.", "basis": "Internal, verified by direct read during the corpus review; the contradiction is between two sections of one file and is the reason the file is held in restricted custody.", "confidence": "verified", "sources": [] }, { "id": "org:claims:what-accuracy-measures:c04", "claim": "A separate internal strategy thesis reached the same conclusion independently and records a prohibition on recommendations of the form others like you in session contexts.", "basis": "Internal, the 2026-07-12 strategy drafts, which cite a platform failure-mode document for the prohibition.", "confidence": "verified", "sources": [] } ], "concepts": [ { "id": "org:concepts:accuracy-freedom-inversion", "name": "Accuracy and freedom inversion", "definition": "The relation by which a predictive model scores highest on exactly the behavior a person had least room to vary, and lowest on value-laden choices in new situations. A rising accuracy figure about a person is therefore ambiguous between a better model and a narrower life, and a system that both predicts and acts on its predictions can resolve the ambiguity in its own favor without anyone intending it. The structural consequence is that prediction and steering can each be graded honestly but a system doing both cannot be graded on its predictions, because it is in a position to author the evidence.", "provenance": "canonical" } ], "researchContext": "The INVERT case, authored under the condition the corpus review sets for the\nclass. The file's refutation is quoted, the contradiction between its\nrefutation and its specification is named, and the specification itself is\nnever restated in enough detail to be useful to anybody. The steering search\nis described at exactly the level our published disclosure already describes\nit and no further. Nothing from the file's state-vector dimensions, its\nentelechy framing, its named psychometric instruments, or its model stack\nappears here or in the brick.\n\nTwo candidates were merged on the review's instruction. The strategy estate\nindependently proposed that predicting a person perfectly destroys them, and\nshipping both would have produced two bricks with one mechanism. The merged\nbrick carries the strategy thesis's operational line, the ban on\nothers-like-you recommendations in session contexts, because that line is the\nargument made structural and it is the part of the strategy material that was\nnot already implied by the trajectory file.\n\nThe Song et al. reading is the brick's own. The paper reports predictability\nas a property of human mobility and does not frame the figure as a measure of\nconstraint; treating the same number as a reading of how much of a life runs\non rails is an interpretation we make and mark as ours. The internal\nprinciple claim is graded directional rather than verified because the file\nstates it without citation, and the brick uses it as the estate's own words\nturned back on the estate rather than as evidence about the world. The\ndistinction between grading a predictor and grading a system that also acts\nis the brick's structural contribution, and it is the reason this argument\nlives under measurement rather than under the safety wall." }, "contract": "https://mnstry.org/contracts/org/org-writing.v1.schema.json", "releaseHash": "7dbcfe01408c1725dc574ead755be15a359b2e73584dcf2c65153c8291757ba9", "versions": [ { "version": 1, "cutAt": "2026-08-03", "note": "Initial publication, spiritual-harvest wave", "visibility": "published", "path": "/writing/what-accuracy-measures/", "contentHash": "sha256:dc329aba3be2829b", "releaseHash": "7dbcfe01408c1725dc574ead755be15a359b2e73584dcf2c65153c8291757ba9" } ] }