Scrub through 327 years of this role's history — from when it first emerged, through every wave of technology that reshaped it, to the cited projections for where it's heading next.
Printing press — movable type (Gutenberg, c. 1440)
Drag the dot, click anywhere on the track, or use ← → arrow keys (Shift for 10-year jumps, PgUp/PgDn for 25).
2026
Known today as Writer / Author / Content Creator
US Employment
135K
BLS OEWS May 2024 estimate (via O*NET). The jump from 51,070 (2023) to 135,400 (2024) is a survey-methodology revision, not an actual tripling of the employed writer population — likely a reclassification of digital-media workers and content creators into 27-3043. Treat with caution; compare to the ACS household survey figure of 183,220 (Data USA, 2024 ACS). Median annual wage: $72,270. Projected growth 2024-2034: +4% ("about as fast as average"), ~13,400 annual openings.
Median Annual Wage
$72,270
Source: BLS-OEWS
Tool of the era · Claude + GPT-4 — long-form AI writing at publishable quality; WGA wins AI protections
By mid-2023, Claude (Anthropic) and GPT-4 (OpenAI) could produce long-form writing — reported journalism, narrative nonfiction, short fiction, scripts — that required substantial editorial expertise to distinguish from human-authored work. The Anthropic Economic Index (January 2026) found that Arts, Design, Entertainment, Sports, and Media tasks accounted for 11% of all Claude.ai traffic in November 2025, growing as Claude was used in an increasing share of conversations for copyediting and the writing and refinement of fictional pieces. The WGA strike (May 2 – September 27, 2023) was the first labor action in any industry to win specific AI protections: AI cannot be used to write or rewrite literary material, cannot undermine writer credit, and studios must disclose when AI-generated material is given to a writer.
The Authors Guild filed a class-action suit against OpenAI (September 19, 2023) with 17 named authors including George R.R. Martin, John Grisham, Jodi Picoult, and David Baldacci. The New York Times filed separately against OpenAI and Microsoft (December 27, 2023), alleging billions of dollars in damages from unauthorized use of published articles as training data. By 2024, a wave of content-licensing deals — News Corp ($250M+), Vox Media, The Atlantic, Reddit — signaled that content had value as AI training data even as the underlying profession contracted.
Projection cone · present → 2034
What credible sources project
Scrub the slider past now to anchor each scenario on the scrubber. The spread you see below is the range of futures credible sources project for this role.
BLS Occupational Outlook 2024-34
→ 2034
+4%
BLS Employment Projections — industry-occupation matrix + labor productivity assumptions. The current OOH for 27-3043 (2024-34 cycle) projects +4% ("about as fast as average"), with ~13,400 annual openings. This is counter-intuitive given the severity of the 2022–2024 AI disruption — but BLS captures formally employed writers in the establishment survey. The freelance contraction (where AI substitution is most acute), mass content-mill displacement, and the implosion of digital-media publishing jobs (which operate largely on contract) are not fully visible in the BLS headline number. The true headcount trajectory for the actual working-writer population is almost certainly more negative.
Frey & Osborne (2013) — pre-LLM estimate
→ 2033
-4%
Gaussian-process classifier on O*NET task features. Frey & Osborne rated Writers and Authors at approximately 0.038 probability of computerisation (based on secondary-literature citation convention) — placing them among the LEAST at-risk occupations of 702 studied, roughly at the same safety level as dentists and physicians. Their framework's key bottleneck variables were "creativity" and "originality" — both rated as engineering blockers for automation that they believed would persist for the foreseeable future. This is arguably the paper's most consequential prediction failure: within a decade, large language models had made writing not just automatable but effortlessly so. The -4% figure represents the implied employment-change direction from F&O's risk assessment; it is included in the cone precisely to show the magnitude of the miss.
Anthropic Economic Index (live observational)
→ 2026
-11%
Direct measurement of Claude API usage by task category, January 2026 report. Arts, Design, Entertainment, Sports, and Media tasks — which includes writing, copyediting, and fiction refinement — accounted for 11% of all Claude.ai traffic in November 2025, growing between August and November 2025. The 11% figure is Claude.ai consumer traffic; first-party API traffic shows 6% for the same category. Unlike the graphic design case (where generative image models do most of the actual displacement), language models are the primary substitution vector for writing — so the Anthropic API figures more directly represent the actual displacement surface. Reported as -11% to represent current observational share, not a permanent employment forecast.
McKinsey Global Institute (2023)
→ 2030
-25%
McKinsey's July 2023 "Generative AI and the Future of Work in America" report. McKinsey identifies "creatives and arts management" as facing a 25 percentage-point shift in automation potential due to generative AI — one of the largest single-category jumps in the report. Writers, creatives, lawyers, and consultants are specifically named as roles that "will all need to work differently because parts of their jobs will be affected by generative AI." The net employment effect on writers is uncertain at the sector level — demand for written content may expand even as per-word output costs collapse — but the -25% figure reflects the share of tasks now automatable.
Goldman Sachs (March 2023)
→ 2030
-26%
Goldman Sachs "Potentially Large Effects of AI on Economic Growth" (March 2023) maps O*NET work-activity importance scores to LLM capability ratings. Arts, Design, Entertainment, Sports, and Media occupations — the broader category that includes Writers and Authors — are assigned approximately 26% task automation exposure by current LLM capabilities. As with Eloundou, this is share of tasks automatable, not jobs eliminated; Goldman notes that demand expansion could produce net employment growth even with high task automation. The -26% figure is the category-level task automation share applied here as the cone lower-bound.
Eloundou et al. — "GPTs are GPTs" (2023)
→ 2025
-80%
GPT-4 task-by-task labeling against O*NET task statements. Writers and Authors is among the very highest-exposure occupations in the Eloundou dataset — writing tasks are literally what LLMs do. The study finds approximately 80% of Writers and Authors' tasks directly exposed to LLM capability (γ ≈ 0.80). This is one of the study's most unambiguous findings: unlike occupations where the LLM exposure is partial, writing is structurally LLM-native. The -80% figure represents γ (any task exposure at all); β (E1 + 0.5×E2, the more conservative measure) would be lower. As with all Eloundou numbers, this is capability not substitution — LLMs can do these tasks; how many jobs actually disappear depends on adoption rates and demand expansion.
Today, in this role
What's shifting in the work right now
The historical view above shows how this role has moved. This is the present-day detail: which AI tools are picking up which tasks, where the edge still is, and the natural directions this work can grow.
What's changing in your day
Three parts of your work where AI is already doing real lifting — and what stays yours.
AI is taking this on
Produce SEO blog content at volume — keyword-targeted articles, FAQ pages, and programmatic content — using AI drafting pipelines in Jasper or Writer.com, then reviewing for E-E-A-T signals, factual accuracy, and human voice before publication.[7],[6]
Volume SEO content is the most displaced segment of writing work — freelance rates for standard blog posts dropped 60-75% from 2022 to 2025 as AI flooded the market. Survive by owning E-E-A-T signals: your byline expertise, cited sources, and editorial judgment are what Google's Helpful Content system now rewards over word count.
AI is taking this on
Generate first-draft marketing copy — product descriptions, email campaigns, landing page headlines, and ad variations — using Jasper or Copy.ai, then edit for brand voice, accuracy, and regulatory compliance before delivery to clients.[8],[9],[6]
Generic first-draft copy is now nearly fully automated at scale. Your defensibility here is brand voice enforcement and fact-checking — clients pay for accuracy and consistency, not word count. Learn to operate AI writing platforms as a production lead who reviews and calibrates, not a line-drafter.
AI is sitting alongside you here
Edit and fact-check AI-generated content drafts — reviewing LLM output from ChatGPT, Claude, or Gemini for hallucinated citations, inaccurate statistics, trademark issues, and tone mismatches before client or publication delivery.[5],[1]
AI-editing is now a distinct and growing role: the ability to efficiently catch LLM hallucinations, enforce house style, and calibrate voice is increasingly what clients want from a human writer. Develop a systematic fact-verification workflow and become expert in the failure modes of the specific LLMs your clients use.
Natural next steps for someone with your foundation — not exits, evolutions.
A direction you could grow
Marketing Managers
Content writers who understand audience psychology, messaging strategy, and channel distribution already possess the core intuitions of Marketing Managers — what they lack is budget ownership, team management, and paid channel expertise. With AI automating content production, Marketing Managers increasingly need people who can brief AI tools well and evaluate output quality, which is exactly what a strong writer brings. This pivot typically requires 1-2 years in a content strategist or content marketing manager role as a bridge step.
What you'd add
· Paid media fundamentals (Google Ads, Meta Ads, programmatic display)