hatred
Enable an AI agent to recognise reported or evidenced hatred, assess its persistence and consequences without equating feeling with intent, and identify proportionate responses.
Research draft, second pass
A second pass drafted this model: the structure a model of this thing needs, and what is known about it in the world. The line under this one says how the second half was obtained - researched against sources, or recalled without web access, in which case nothing here was read anywhere and every claim is a lead to verify. Unreviewed either way.
Researched by: Codex
Purpose and description
Enable an AI agent to recognise reported or evidenced hatred, assess its persistence and consequences without equating feeling with intent, and identify proportionate responses.
It can be Record hatred as a qualified attribution linked to statements, observations and competing interpretations.; Clarify the target and distinguish hostility toward conduct from rejection of a person or group.; Track recurrence, intensification, generalisation and reduction using comparable observations.; Separate hostile feelings and wishes from intentions, preparations and acts, linking concrete concerns to an appropriate safety process.; Identify opportunities for voluntary distance, interruption of hostile rehearsal, support or changes in interaction.; Reassess attribution and outcomes without requiring forgiveness, renewed contact or agreement with the target..
Distinguishing features
Test whether hostility extends beyond condemning a particular act to a broader rejection of its target; disagreement or moral criticism alone does not establish hatred.
Test whether the orientation recurs outside an immediate provocation; persistence helps distinguish it from an anger episode but is not an established necessary condition.
Test whether avoidance reflects hostility toward the target rather than fear, sensory aversion or a protective boundary alone.
Test whether words such as 'hate' express serious target-directed hostility, exaggeration, quotation, satire or an unwanted intrusive thought.
Record hatred, prejudice, harmful intent and harmful conduct separately: evidence of any one does not automatically establish the others.
Scope
+ Who reportedly hates whom or what, including uncertainty about the holder and target
+ Whether hostility concerns particular conduct or extends to the target's perceived identity or existence
+ Reported experience, recurring appraisals and changes in hatred over time
+ Expressions of hatred and their relationship to desires, intentions and conduct
+ Social reinforcement, target generalisation and conditions affecting persistence
+ Boundaries and interventions relevant to reducing harm without demanding reconciliation
- Anger episodes and general emotion regulation as independent phenomena
- Disgust, contempt, fear and aversion without evidence of hatred
- Prejudice, discrimination and ideologies as complete social systems
- Threat assessment, violence planning and incident management as independent processes
- Legal classification of hate speech or hate crimes
- Clinical diagnosis and treatment of mental disorders
Characteristics
- Attribution basis
- self-report | attributed by another | inferred from expression or conduct | disputed | unknown; multiple values allowed Distinguishes an expressed experience from an observer's interpretation of another person's internal state.
- Holder and collective attribution
- identified person, organisational position or bounded collective claim, with evidence and limits of attribution Prevents one speaker's hostility from being assigned to every member of a group.
- Target and target scope
- self | individual | group | institution | practice | idea | other; record specificity and changes in scope Distinguishes hatred of an actor from hostility generalised to associated people or categories.
- Hostile appraisal
- recorded appraisals concerning conduct, character, identity, perceived threat or perceived irredeemability; unresolved where unclear Captures what the holder rejects without endorsing the appraisal as true.
- Experienced intensity
- holder's stated intensity or a declared scale with anchors, date and method; unknown when unavailable Supports comparison over time without treating an unvalidated score as a measure of dangerousness.
- Temporal pattern
- newly reported | episodic | recurrent | sustained | diminishing | no longer reported | unknown, with observation period Separates current activation from a continuing orientation and preserves uncertainty during silence.
- Desired outcome for the target
- distance | changed conduct | accountability | exclusion | humiliation | suffering | destruction | mixed | unstated; preserve the holder's wording Distinguishes wishes associated with hatred from concrete intentions or plans.
- Expression and action relationship
- reported feeling only | expressed hostility | stated wish | stated intention | preparation reported | conduct observed | unknown; categories may coexist Keeps evidence about feeling, communication and action distinct while allowing relevant links.
- Restraints and boundaries
- reported commitments, protective distance, support relationships and situational constraints, with evidence of current relevance Records ways a holder limits conduct without assuming that restraint means hatred has ended.
Also called
Where this came from
wikidata · CC0 1.0
Drafted structure
Bundle to layer to finding to question, as the second pass will find it: 5 bundles · 10 layers · 10 findings · 20 questions.
Recognition and attribution Establishes what supports describing an orientation as hatred and whose description it is.
Hatred is an interpretation of experience and orientation that cannot safely be read directly from a word or isolated act.
Attribution evidence
Separates first-person reports, outside allegations and agent inferences.
Basis for hatred attribution
Record the supporting statement or observation, its source and circumstances, and whether the attribution is contested.
- Who calls this hatred, and what exact report or observation supports that description? provenance
- Could the apparent expression instead be quotation, exaggeration, performance or an unwanted intrusive thought? boundary
Neighbouring experiences
Tests the interpretation against anger, contempt, disgust, fear and principled opposition.
Hostility beyond an episode
Record whether the holder describes broader target rejection and how that differs from their immediate reaction to particular conduct.
- What does the holder mean by hatred here, beyond being angry, afraid, disgusted or opposed? definition
- Does the hostile orientation remain when the immediate provocation is absent, and what evidence establishes that? measurement
Target and hostile meaning Identifies the object of hatred and the appraisals through which the holder understands it.
Response choices depend on whether hostility concerns an act, a whole person, the self or a generalised category.
Target boundaries
Records the target's identity and the extent of any generalisation.
Specific and generalised targets
Preserve distinctions between a named target, its conduct and associated people or categories.
- Is hatred directed at the self, a particular actor, a group, an institution, a practice or an idea? definition
- Has hostility spread to people merely associated with the original target, and where does the holder draw exceptions? boundary
Appraisals and grievances
Records the holder's account of wrongdoing, threat or intolerability without adopting it as fact.
Grounds and revisability
Distinguish reported events from interpretations of the target's character, identity and capacity to change.
- Which experiences, accounts or claims does the holder identify as grounds for hatred, and which have independent support? provenance
- Does the holder describe anything that could change their appraisal, or regard the target as irredeemable? boundary
Persistence and reinforcement Records how hatred varies over time and what reportedly renews or weakens it.
A model must distinguish temporary activation, persistent orientation and apparent reduction caused only by lack of observation.
Temporal course
Tracks onset, recurrence, intensity and periods of reduced activation.
Recurrence and change
Record dated observations and self-reports using consistent descriptions or explicitly defined scales.
- When was hatred first reported, how often does it recur, and over what observation period? measurement
- What supports an increase or decrease in hatred rather than a change in willingness or opportunity to express it? measurement
Renewal contexts
Examines reported links with reminders, repeated grievances and social encouragement.
Hostile rehearsal and social reward
Record whether revisiting grievances, consuming target-focused material or receiving approval accompanies renewed hostility; keep causal explanations tentative.
- Which encounters, reminders or repeated narratives reportedly reactivate hatred? provenance
- Do particular audiences reward expressions of hatred or discourage moderation, and what observations support that interpretation? provenance
Expression and action Distinguishes experienced hatred from its communication, desired outcomes and relationship to conduct.
Hatred alone does not establish harmful intent, while concrete intentions or preparations require attention beyond emotion description.
Communication and wishes
Records what is expressed, to whom and with what stated desired consequence for the target.
Expression context and target outcome
Distinguish private disclosure, public denunciation, targeted abuse and calls for others to act, preserving ambiguity about purpose.
- What was communicated, to which audience, and was the target directly addressed? provenance
- Does the holder seek distance, accountability, exclusion, humiliation, suffering or another outcome for the target? definition
Intent, conduct and restraint
Links hatred to separately evidenced intentions and conduct without treating those links as automatic.
Transition evidence and action limits
Record explicit intentions, preparations, observed acts and stated restraints as distinct evidence requiring their own interpretation.
- What evidence, if any, distinguishes a hostile wish from an intention, preparation or completed act? boundary
- What restraints are reported or observed, and do concrete threats or preparations warrant referral to a separate safety process? action
Response and reassessment Identifies permissible responses and evaluates change without making reconciliation the default objective.
An agent needs to support reduced harm and informed choices while respecting protective boundaries and uncertainty about internal states.
Response selection
Connects the holder's goals, the target's boundaries and the agent's authority to possible interventions.
Voluntary change and protective distance
Record whether help is sought with distress, hostile expression, unwanted rehearsal or conduct, and whether continued distance serves protection.
- What change does the holder want, and which actions fall within the agent's role and authorisation? action
- Would contact, mediation or reconciliation conflict with either party's boundaries or expose someone to further harm? boundary
Outcomes and revision
Evaluates changes in experience, appraisal and conduct separately.
Reduced hatred versus reduced expression
Distinguish less reported hatred, narrower target scope, reduced hostile conduct and simple absence of new evidence.
- Which outcome changed: reported hatred, appraisal of the target, expression, desired harm or actual conduct? measurement
- What new evidence would justify revising or withdrawing the attribution of hatred? boundary
What the second pass must settle
- Which researched recognition criteria best distinguish hatred from sustained anger, contempt and aversion, and how do those criteria vary across languages and contexts?
- Can self-hatred, interpersonal hatred and hatred of practices or ideas share this model without obscuring differences that require specialised extensions?
- What evidence justifies attributing hatred to a collective beyond identifiable speakers, official communications or coordinated acts?
- Which measures of intensity, persistence and target generalisation are sufficiently validated for an agent's intended use?
- Which interventions reduce distress or harmful conduct in different contexts, and when might contact, confrontation or pressure to forgive worsen outcomes?