Chapter 2 · Tasks and jobs
What people bring to AI is not who they are.
Anthropic classifies observed Claude conversations into occupational task groups. That shows activity on one platform. It does not show users' actual jobs, and it does not measure what AI does to employment.
The verified version of the thesis. In Anthropic's sample of U.S. Claude.ai conversations (May 2026), 21.1% were classified to computer and mathematical tasks, the largest of 22 groups. This describes activity on one platform, not users' occupations or the effects of all AI.
U.S. Claude.ai conversations by occupational task group
Bars: May 2026. Grey: April 2026.
Computer and mathematical tasks lead at 21.13%, followed by arts, design and media (12.91%) and education (11.92%). Construction (0.14%) and farming (0.05%) are at the bottom.
Caveat. Shares describe what sampled conversations were about, mapped to O*NET tasks and SOC groups. They are not users' jobs, not all AI use, and not economic effects. Shares sum to 98.67% because small cells are suppressed.
Inspect the evidence
- Users' actual occupations or industries.
- AI used inside other software, machines or industrial systems.
- Whether the task was completed faster or better.
- Other AI products (ChatGPT, Gemini, Copilot).
- aei_claude_ai_2026-06-26.csv
- Filter geo_id=USA, category_name=soc_occupation, hierarchy_level=1, metric_id=pct
- No reweighting
Text description. Computer and mathematical tasks lead at 21.13%, followed by arts, design and media (12.91%) and education (11.92%). Construction (0.14%) and farming (0.05%) are at the bottom.
Developer API traffic by occupational task group (global)
API traffic leans further to computer and mathematical tasks (28.17%).
Caveat. 1P API excludes Claude Code and third-party clouds. Global only; no U.S. split published.
Inspect the evidence
Text description. API traffic leans further to computer and mathematical tasks (28.17%).
What U.S. Claude.ai conversations produced
Explanations and documents lead; code fixes are a small share of consumer chat.
Caveat. An artifact is the most prominent output, classified by a model. Producing a document does not mean it was used.
Inspect the evidence
Text description. Explanations and documents lead; code fixes are a small share of consumer chat.
Work
of U.S. conversations classified as work
Personal
classified as personal
Automation
directive or feedback-loop pattern (AI does the task)
Augmentation
learning, iteration or validation (human leads)
U.S., Claude.ai, May 2026. Self-reported productivity in Anthropic's linked survey (~9,700 respondents, not representative): faster 86%, broader scope 82%, higher quality 69%. Self-reports are kept separate from measured outcomes.
Workers
Occupations behind the headlines, May 2025.
Employment and pay from the BLS occupational survey. This is a level, not a trend: the BLS API serves only the latest year, and the historical files refused this client (logged in the gaps register). So this page does not claim that any occupation grew or shrank because of AI.
Radiologists are the canonical test. Since 2016 they have been cited as the job AI would replace first. In May 2025 BLS counted 26,770 radiologists earning a mean $381,530 a year. The trend that would settle the question is the next acquisition.
| Occupation | SOC | Employment | Mean annual wage | Median annual wage |
|---|---|---|---|---|
| Home health and personal care aides | 31-1120 | 4,305,810 | $36,120 | $35,800 |
| Registered nurses | 29-1141 | 3,379,720 | $101,420 | $97,550 |
| Customer service representatives | 43-4051 | 2,595,750 | $46,590 | $44,770 |
| Heavy and tractor-trailer truck drivers | 53-3032 | 2,062,040 | $59,710 | $58,640 |
| Software developers | 15-1252 | 1,687,890 | $148,100 | $135,980 |
| Accountants and auditors | 13-2011 | 1,449,500 | $94,750 | $83,680 |
| Elementary school teachers | 25-2021 | 1,388,390 | $72,650 | $63,970 |
| Secondary school teachers | 25-2031 | 1,065,210 | $76,320 | $72,040 |
| Electricians | 47-2111 | 757,220 | $71,490 | $63,190 |
| Plumbers, pipefitters, steamfitters | 47-2152 | 465,840 | $72,170 | $63,800 |
| HVAC mechanics and installers | 49-9021 | 409,670 | $64,780 | $61,010 |
| Paralegals and legal assistants | 23-2011 | 392,880 | $69,700 | $62,890 |
| Physicians, all other | 29-1229 | 342,720 | $262,040 | $265,930 |
| Nurse practitioners | 29-1171 | 323,040 | $137,300 | $132,300 |
| Network and computer systems administrators | 15-1244 | 314,340 | $103,680 | $99,130 |
| Machinists | 51-4041 | 287,050 | $59,320 | $58,750 |
| Data scientists | 15-2051 | 262,440 | $126,800 | $120,230 |
| Radiologic technologists and technicians | 29-2034 | 230,490 | $83,840 | $80,110 |
| Graphic designers | 27-1024 | 197,830 | $70,560 | $62,960 |
| Medical records specialists | 29-2072 | 194,720 | $56,790 | $51,140 |
| Family medicine physicians | 29-1215 | 107,510 | $255,820 | $244,180 |
| Interpreters and translators | 27-3091 | 52,060 | $66,360 | $60,170 |
| Technical writers | 27-3042 | 45,500 | $96,970 | $90,390 |
| Radiologists | 29-1224 | 26,770 | $381,530 | $420,860 |
Task exposure is not job loss. An occupation can have many AI-suitable tasks and still grow.
Observed task use is not automation. Most U.S. conversations are split roughly evenly between automation-style and augmentation-style patterns.
A new task is not a new job. Roles are called new here only with employment evidence that persists over time.