← UK Occupations
🇬🇧 UKSOC 2115· Science, Research, Engineering and Technology Professionals

Social and humanities scientists

Social and humanities scientists study and analyse human behaviour and the origin, structure and characteristics of language undertake research in areas such as sociology, economics, politics, archaeology, history, philosophy, literature, the arts organise the collection of qualitative and quantitative information and perform subsequent analyses.

AnthropologistArchaeologistCriminologistEpidemiologistGeographerGIS analystHistorianPolitical researcherSocial scientist
£38,591
median salary
£28,393 – £46,192
44,400
currently employed
47,002
projected 2035
Strong growth
7 tasks scored1 AI Automates4 AI Assists2 Human-Led

Task Bundle Breakdown

How AI interacts with this occupation's task bundle

14%
57%
29%
14%
AI automates
57%
AI assists
29%
Human-led

Task-by-Task Breakdown

Every task scored individually — sorted highest to lowest AI exposure

78

identifies, compiles and analyses economic, demographic, legal, political, social and other data to address research objectives

AI Automates

AI tools like Python-based ML libraries, Power BI Copilot, and ChatGPT Advanced Data Analysis can rapidly compile, clean, and analyse large multi-source datasets, significantly accelerating this task.

67

traces the evolution of word and language forms, compares grammatical structures and analyses the relationships between ancient, parent and modern languages

AI Assists

NLP tools and large language models can assist with comparative linguistic analysis, identifying grammatical patterns, and tracing etymological relationships across corpora at scale.

67

undertakes analyses of data, presents results of research to sponsors, the media and other interested organisations

AI Assists

AI tools like ChatGPT, Tableau, and Power BI Copilot can automate data visualisation, draft summaries, and identify patterns, though researchers must interpret findings and communicate nuance.

56

addresses conferences and publishes articles detailing the methodology and results of research undertaken

AI Assists

AI like Grammarly and Claude can assist with drafting, structuring, and editing academic papers and conference presentations, but original argumentation and scholarly authority remain human-led.

44

administers questionnaires, carries out interviews, organises focus groups and implements other social research tools

AI Assists

AI can assist with survey design, transcription (Otter.ai), and thematic coding of qualitative data, but conducting interviews and focus groups requires human rapport and adaptive questioning.

33

studies society and the manner in which people behave and impact upon the world

Human-Led

AI can surface literature and synthesise existing findings, but the interpretive, theoretical, and observational work of studying society requires deep human judgment and contextual understanding.

33

undertakes research across the humanities that furthers understanding of human culture and creativity

Human-Led

AI tools like Claude can assist with literature reviews and summarising cultural texts, but original humanistic inquiry, interpretation, and creative insight remain fundamentally human endeavours.

ROI Calculator

Social and humanities scientists

13.4hrs
saved per week
642.9hrs
per year ≈ 17.1 weeks
£13,782
salary-equivalent capacity freed per year

This is not a pay rise — it's 642.9 hours of productive capacity that AI could return to you, valued at your current rate.

£27,565
value at risk — 71% of role exposed to AI

The portion of your current salary tied to tasks AI can automate or augment. Not a prediction of job loss — a signal of where your role is changing fastest.

How is this calculated?▾

Task weights — each task is weighted by its O*NET frequency rating. Where frequency data is absent, equal weighting is applied.

AI time reduction — tasks scored as "AI Automates" assume 80% time saving; "AI Assists" assumes 40%; "Human-Led" assumes 5%.

Salary equivalent — annual hours saved × (salary ÷ 1,800 working hours/year).

Value at risk — salary × share of role time in AI-automate or AI-assist tasks.

Figures are indicative estimates for planning purposes, not guarantees of productivity gain or job loss risk.