AI Vulnerability Study Dimension

Routine vs. Novel

One of 4 dimensions that determine AI vulnerability archetype assignment

straightenWhat This Dimension Measures

The Routine vs. Novel dimension captures how predictable or patterned a person's daily work is. Routine work follows recognizable sequences: the steps are known, the inputs are familiar, and the output format is established. Novel work requires combining information from sources that do not normally connect, navigating ambiguity, and approaching each task as a unique problem rather than an instance of a known category. Most knowledge work contains elements of both, but the ratio between them varies significantly across roles.

This dimension is critical for AI vulnerability because current AI systems are fundamentally pattern-matching engines. They excel at tasks that can be decomposed into known sequences, where training data provides abundant examples of similar work. Routine work is precisely what machine learning models optimize for: given enough examples of how a task has been done before, the model can generate a competent version. Novel work, by contrast, requires the kind of lateral thinking, contextual judgment, and cross-domain synthesis that current systems handle poorly.

The relationship between AI capability and routine work is accelerating. Each generation of language models handles a broader range of tasks that were previously considered too complex for automation. Tasks that felt novel five years ago may now fall within AI's routine handling capability. This means the boundary between routine and novel is not fixed; it moves as AI improves. People whose work currently feels novel should consider whether it will still feel novel in two to three years as AI capabilities expand. The relevant question is not whether the work is novel today, but whether the novelty is structural or merely a function of current tool limitations.

The scoring mechanism combines three tradeoff pairs that probe the routine-novel balance with two scenario sub-questions that test how the person approaches new situations. T4 asks directly about pattern recognition in daily work. T5 probes whether the first steps of a project are known in advance or require discovery. T6 tests whether tedium comes from familiar tasks or from pre-work scoping. The scenario contributions from S1a and S1c add behavioral signals about how the person responds to real situations that require either routine execution or novel investigation.

swap_horizThe Spectrum

Routine Novel
syncRoutine

Scoring toward Routine indicates that daily work follows recognizable patterns with predictable steps. The person knows what to do before being briefed, and the most tedious aspect of work is performing familiar tasks. AI excels at pattern-based execution, making routine-heavy roles among the most exposed to automation.

lightbulbNovel

Scoring toward Novel indicates that daily work requires combining information from disconnected sources and navigating situations without established templates. Understanding what is actually needed often requires multiple conversations before work can begin. Novel work resists automation because each instance presents a unique combination of inputs that defies standardized processing.

circle Middle Position

A moderate position on this dimension indicates work that involves both established patterns and novel problem-solving. Some tasks follow predictable procedures while others require fresh thinking and cross-domain synthesis. This balance provides partial protection because AI handles the routine components while the novel components remain human-dependent.

scienceHow It's Measured

This dimension is measured through three tradeoff pairs (T4, T5, T6) that probe the routine-novel balance, plus two scenario sub-questions (S1a, S1c) that contribute at half weight each.

T4 tradeoff
Tests whether work follows recognizable patterns with known steps (routine) or requires combining information from sources that do not normally connect (novel).
calculateCore tradeoff. Option A pulls toward Routine; Option B pulls toward Novel.
T5 tradeoff
Tests whether the first steps of a new project are known before anyone provides a brief (routine) or whether understanding what is needed requires talking to several people first (novel).
calculateCore tradeoff. Option A pulls toward Routine; Option B pulls toward Novel.
T6 tradeoff
Tests whether the most tedious part of work is doing familiar tasks that could be automated (routine) or figuring out what needs to happen before starting (novel).
calculateCore tradeoff. Option A pulls toward Routine; Option B pulls toward Novel.
S1a scenario
In the briefing scenario, tests whether the first action is to update and deliver quickly (routine execution) or to consult others about how the strategic context has changed (novel investigation).
calculateWeight: 0.5. Option A pulls toward Routine; Option B pulls toward Novel.
S1c scenario
If an AI tool could generate a solid first draft in five minutes, tests whether the response is to produce more frequently (routine scaling) or to redirect time toward understanding whether the deliverable is the right one (novel judgment).
calculateWeight: 0.5. Option A pulls toward Routine; Option B pulls toward Novel.
infoThe raw score is calculated as the mean of T4, T5, and T6 (each ranging from -2 to +2), plus 0.5 times the S1a scenario value and 0.5 times the S1c scenario value. This produces a raw range of approximately -3.5 to +3.5. The raw score is then normalized to a 0 to 100 scale, where 0 represents the strongest possible Routine orientation and 100 represents the strongest possible Novel orientation.

device_hubWhere Archetypes Cluster

Archetypes distribute across this dimension with a clear gradient: The Exposed archetypes cluster toward Routine, The Transitioning archetypes occupy the middle range, and The Durable archetypes cluster toward Novel.

Leans Routine (0-40)

Archetypes at this end perform work that follows established patterns and predictable procedures. The routine nature of the work is what makes it most susceptible to AI automation. These roles often feel productive and efficient, which can mask the fact that the efficiency is precisely what AI replicates.

The Accelerated Producer The Template Specialist The Volume Player The Efficiency Amplifier The Confident Automator The Acceleration Navigator The Confident Explorer
Balanced (40-55)

Archetypes in the middle range experience a mix of routine tasks and novel challenges. The routine components are being absorbed by AI while the novel components are becoming more prominent. This transitional position is where most knowledge workers will find themselves as AI adoption accelerates.

The Institutional Memory The Selective Curator The Dual Navigator
Leans Novel (55-100)

Archetypes at this end handle work that resists standardization. Each task presents a unique combination of inputs requiring fresh synthesis, cross-domain thinking, and contextual judgment. AI can contribute components, but the assembly and direction remain deeply human.

The Judgment Concentrator The Knowledge Translator The Context Bridge The Orchestrator The Catalyst The Sense-Maker The Relationship Architect The Cautious Stronghold

account_treeInteractions with Other Dimensions

The Routine vs. Novel dimension interacts with the other three dimensions to shape vulnerability profiles. Its interaction with Creation vs. Curation is the strongest predictor of overall vulnerability.

call_split
Creation vs. Curation

Routine Creation is the most automatable work pattern in the study. Producing predictable outputs from known procedures following established templates is what AI does best. Novel Curation is the least automatable: evaluating unique situations using experiential judgment to determine what matters. The diagonal between these two corners of the dimensional space accounts for the majority of variance in the Vulnerability Index.

Common pattern: Routine creation is the epicenter of AI exposure. Novel curation is the epicenter of durability.
call_split
Individual vs. Coordination

Routine Individual work is highly automatable because it involves one person performing predictable tasks in isolation. Routine Coordination work is somewhat protected because the social overhead of coordination resists automation even when the tasks themselves are routine. Novel Coordination work, where unique situations require cross-boundary navigation, represents the strongest combination on these two dimensions.

Common pattern: Routine individual work is easiest to automate. Novel coordination is hardest.
call_split
Explicit vs. Tacit

Routine work grounded in Explicit knowledge is the most accessible target for AI systems, as both the procedures and the knowledge base are documented and replicable. Novel work drawing on Tacit knowledge presents the deepest challenge for automation, requiring pattern recognition from experience and contextual judgment that cannot be extracted from documentation alone.

Common pattern: Routine + Explicit = most accessible to AI. Novel + Tacit = most resistant.

targetWhy This Dimension Matters

The Routine vs. Novel dimension provides the clearest lens for understanding how quickly AI might affect a specific role. Roles at the Routine end face near-term displacement pressure measured in years, not decades. Roles at the Novel end face a longer timeline, though the boundary between routine and novel shifts as AI capabilities expand. Understanding where current work falls on this spectrum enables proactive planning rather than reactive adjustment.

The practical implication is that investing in novelty-generating skills provides the most durable career protection. Learning to handle ambiguity, synthesize across domains, navigate situations without templates, and ask questions that reframe problems are all investments in the Novel end of the spectrum. These skills compound over time and become more valuable as AI handles an expanding range of routine tasks.

quizSee Where You Fall

The AI Vulnerability Study takes approximately 6 minutes. It produces a personalized archetype based on all 4 dimensions.

Take the Assessment

exploreAll Dimensions

The AI Vulnerability Study measures 4 dimensions. Each contributes to the archetype assignment.

arrow_forward Creation vs. Curation
circle Routine vs. Novel
arrow_forward Individual vs. Coordination arrow_forward Explicit vs. Tacit