COOKED
The core deliverable — drafting prose to a brief, on deadline, in a specified voice — is the single task current language models perform most competently and most cheaply, and the commodity end of the market (SEO copy, product descriptions, ghostwritten business books, content marketing, corporate scripts) is where most of the paid work actually sits. What survives is not writing skill in the abstract but authorship: a name readers seek out, original reporting or lived access nobody can synthesize, and being the person a client trusts to own the message. Score reflects the modal working writer, not the bylined novelist or the staff feature writer; the profession is splitting into a small identity-driven tier and a shrinking, price-crushed production tier.
Roughly flat across the period, with year-to-year wobble.
Median pay $63,200 → $76,910 -2.6% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+3.6%
Percentage only. The projection counts a different population from the 47,940 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +3.6% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~13,400 openings a year on average, including replacing people who leave.
PoetAuthorWriterBloggerHumoristLyricistNovelistDramatistBiographerCopywriterGag WriterLibrettistPlaywrightSongwriterBlog WriterFilm WriterGame AuthorGhostwriterPlay WriterPoem WriterBlurb WriterComic WriterMovie WriterMusic Critic
Holding it up: trust premium . Weakest point: liability shield .
Core tasks are already automatable At 4, the day-to-day — turning a content brief, style guide and word count into finished copy, then revising to editorial notes — is exactly the generate-and-revise loop models do at near-zero marginal cost; the points that remain reflect the minority of assignments requiring an interview, a site visit, or reading a 300-page source nobody has digitized.
Fully desk- and screen-based A 2 acknowledges the laptop-and-deadline reality of the job: apart from occasional travel to interview a subject or attend an event, nothing in drafting, researching, or delivering a manuscript requires your body in a particular place.
No licence, no signature requirement There is no licence, no board, no continuing-education requirement to call yourself a writer; a publisher's indemnity clause pushes libel and plagiarism exposure back onto you but grants no monopoly, which is why this sits at 1 rather than 0 — you carry risk without the credential that would restrict who competes for the work.
Executes defined procedures on defined inputs At 6, the consequential calls — what the piece argues, whether it runs, what gets cut for legal — are made by an editor, publisher or client, and your discretion is real but bounded to structure, framing and sentence-level choices within an approved brief.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (4/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (1/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (6/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 15 of this occupation's 21 points (71%).
Embodiment (2/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
Public Relations Specialists EXPOSED
Actors EXPOSED
Music Directors and Composers EXPOSED
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 43/100 — EXPOSED.
Genuine two-tier collapse: the production tier (SEO, product copy, corporate scripts) automates out, leaving work whose input is non-synthesizable — original reporting, interview access, archival and FOIA-derived material, lived testimony, and negotiated on-record sourcing. The measured score rises because the denominator shrinks, not because the tasks got harder. This is a smaller occupation at a higher score, not relief for current incumbents.
The Authors Guild's 'Human Authored' certification mark (launched 2025, tied to the Guild's registry) becoming a condition of shelf placement or platform surfacing — e.g. a major retailer or Audible/KDP filtering or badging by it, or a Big Five imprint contractually requiring certification. Buyer-visible provenance labeling is what converts diffuse preference-for-humans into a price differential.
Copyright registrability as a de facto human-signature requirement: the Copyright Office's 2023 registration guidance and Thaler v. Perlmutter (D.C. Cir. 2025) mean AI-generated text is unregistrable. If publishers and studios harden this into standard contractual warranties — author personally warrants and indemnifies that the text is human-authored, backed by E&O insurers refusing to cover unwarranted manuscripts — a named human becomes legally necessary to make the asset ownable and insurable.
Platform-level saturation backlash producing paid subscriber demand for named humans: Substack/Patreon-style direct-pay where the purchase is the person, plus disclosure rules (e.g. EU AI Act Art. 50 transparency obligations phasing in Aug 2026) forcing synthetic text to be marked, making unmarked human work a visible category rather than the invisible default.
Defamation and product-liability exposure attaching to a named human: cases like Walters v. OpenAI put unsettled liability on model outputs, and Section 230 is widely read not to cover generated text. If courts or publisher insurers settle on 'a natural person must be identifiable as author of record for indemnity to attach', bylines become a liability instrument rather than a courtesy.
Formalized editorial accountability standards — e.g. an outlet or publisher policy (several newsrooms and the AP already have AI use policies) naming a specific human as accountable for every published claim, with correction and retraction authority resting on that person. If trade bodies or a professional standard (IPSO-style, or an expanded Authors Guild code) make named accountability a condition of publication, the role owns the consequential call explicitly.
The limit. Even with every lever, this stays a low-to-mid score occupation and the gains are concentrated in the identity tier. Certification marks and human-authorship warranties protect the writer who already has a name; they do nothing for the anonymous production tier, whose work is not sought by name and where no buyer is paying for provenance. The realistic outcome is a higher score attached to far fewer people.
| New York-Newark-Jersey City, NY-NJ | 7,220 | — |
| Chicago-Naperville-Elgin, IL-IN | 2,120 | $75,940 -1% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 2,060 | $108,460 +41% |
| San Francisco-Oakland-Fremont, CA | 1,560 | — |
| Miami-Fort Lauderdale-West Palm Beach, FL | 1,270 | $66,470 -14% |
| Boston-Cambridge-Newton, MA-NH | 1,130 | $85,260 +11% |
| Denver-Aurora-Centennial, CO | 860 | $69,100 -10% |
| Dallas-Fort Worth-Arlington, TX | 850 | $64,520 -16% |
| Waterbury-Shelton, CT | 50 | $127,350 +66% |
| San Jose-Sunnyvale-Santa Clara, CA | 400 | $118,810 +54% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 2,060 | $108,460 +41% |
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.