FutureJobPath logo
The career map for the AI era
This page explains how the Durability Score is built — the components, the evidence behind each one, and the named sources. For who this work fits and what a career path through it looks like, see the Deep Read. For your personalized match, take the free quiz.
Where the 50 comes from.

Three components - Automation Resistance, Structural Moat, and Demand - add up to 50.

Data note

Federal labor data does not count AI safety or alignment engineers separately; the wage, workforce, openings, and AI-exposure numbers use Data Scientists as the public comparison. That gives a large modeling base, while actual alignment hiring is much narrower and more concentrated.

FJP Durability Score
50/100
Automation Resistance
16/40

Automation resistance is moderate because research judgment remains human, while experiment scaffolding is easy to accelerate. Routine experiments, summaries, and test generation are tool-friendly; resistance comes from research taste and evidence standards for difficult model behavior.

Sub-components
Substitution Resistance
6/30

Substitution resistance is limited for routine experiments and summaries, but stronger for research direction and evidence judgment.

Sources feeding this sub-component
Anthropic labor-market impacts → Observed exposure for the Data Scientists occupation category is 46.05%.
Tufts American AI Jobs Risk Index → Median modeled job-loss pressure for the occupation category is 37.18%.
Augmentation Leverage
10/10

Augmentation leverage is high because AI can help generate tests, write code, summarize papers, and organize results.

Sources feeding this sub-component
IMF Staff Discussion Notes on AI and labor markets → Links AI-related skills with wage premiums in exposed labor markets.
Structural Moat
14/35

The moat is research credibility, math, programming, and trust in a small field rather than a protected credential. The barrier is built through math, programming, machine-learning depth, fellowships, papers, evaluations, and trusted feedback in a small research community.

Sub-components
Physical & Environmental
1/10

Physical and environmental protection is minimal because the work is mostly computational and desk-based.

Regulatory Moat
1/12

Regulatory protection is low; safety concern can raise attention, but no license reserves the job.

Robotics Resistance
8/8

Robotics are not the main threat because the work is analytical and research-focused.

Sources feeding this sub-component
Credential Depth
4/5

Credential depth is moderate through advanced study, fellowships, papers, evaluations, and strong technical portfolios.

Sources feeding this sub-component
O*NET Online occupation summary → Lists this occupation in Job Zone 4, a higher-preparation category.
Demand
20/25

Demand is narrow and lab-centered even though the broader data-science public row is much larger. The available labor scale comes from data science, while real alignment hiring clusters around labs, fellowships, evaluation teams, and safety nonprofits.

Sub-components
Volume
9/10

Volume is high only in the broader data-scientist row; actual alignment roles are much narrower.

Sources feeding this sub-component
Bureau of Labor Statistics Employment Projections → Data Scientists: 245.9K jobs, 33.5% growth, and 23.4K annual openings.
Source Quality
6/8

Source quality is decent but imperfect because data scientists are a broad anchor for a specialized research path.

Resilience
5/7

Resilience is fair because frontier-model risk keeps attention on the work, but funding and roles can remain concentrated.

What would move the score
Scenario 1
Research support gets automated

The case weakens if labs need fewer junior researchers because AI can run routine evaluations, draft writeups, and reproduce common experiments. The title would remain important, but entry access would narrow. The training risk is losing the small tasks that help new researchers learn how evidence is built.

Direction
down
Components affected
Automation Resistance, Demand
Scenario 2
Model evaluations become central

The case strengthens if labs, buyers, and governments require credible behavioral evidence before major model releases. That would support more roles for people who can design tests and interpret uncertain results. That would support more roles for people who can design tests, audit results, and explain uncertainty to decision-makers.

Direction
up
Components affected
Demand
Scenario 3
Safety splits into research and assurance

A mixed outcome needs review if a few research roles stay elite while more jobs move toward evaluation checklists and compliance support. The important question would be whether the worker sets evidence standards or only runs them. Readers should watch whether new jobs ask for original research judgment or mainly checklist execution.

Direction
neutral
Components affected
Automation Resistance, Demand
Personalized job matches →
Want to find the careers that fit your specific profile? Take the free FJP quiz — 3 personalized matches.
Last reviewed June 2026 · Next September 2026