Writers and Authors
27-3043.00Median wage $76,910/yr47,940 employed (US)Rank #70 of 923 scored · top 8% by substitution
Originate and prepare written material, such as scripts, stories, advertisements, and other material.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution53
Exposure44
Augmentation83
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
20 rated tasks, binned by substitution score.
Position among all scored occupations
Distribution of 923 occupation scores; the marker is this occupation.
Tasks with substitution ≥ 70
15%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%43
panel mean rating 2.7/5 → substitution pressure 43/100
Technical feasibility todayw 20%45
panel mean rating 2.8/5 → substitution pressure 45/100
Cost vs. human wagew 15%60
panel mean rating 3.4/5 → substitution pressure 60/100
Adoption barriersw 20%inverted — strong barriers lower the score70
panel mean rating 2.2/5 (barrier strength) → substitution pressure 70/100
Sector adoption velocityw 10%54
panel mean rating 3.2/5 → substitution pressure 54/100
Task breakdown (20 tasks)
Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.
Write advertising material for use by publication, broadcast, or internet media to promote the sale of goods and services.
82CI 79–86 · exposure 75 · augmentation 100 · click for rater detail
Write advertising material for use by publication, broadcast, or internet media to promote the sale of goods and services.
82| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Advertising and marketing—information-sector, highly digitized—are among the fastest AI adopters. Major platforms, agencies, and retailers are deploying AI copywriting tools in production, with measurable displacement already evident. |
| Sector adoption velocity | claude-sonnet-5 | 5/5 | Marketing and advertising is one of the fastest-adopting sectors for generative AI, with widespread production use of AI copywriting tools across agencies and brands. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically assists human copywriters by generating multiple drafts, variations, and brainstorms instantly, allowing creative professionals to focus on strategy, refinement, and brand fit rather than blank-page composition. Humans remain in the loop while productivity soars. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dramatically speeds up brainstorming, drafting, and iterating on ad copy while writers retain control over brand voice, strategy, and final selection. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can generate promotional copy, slogans, and basic advertising material at scale with minimal human input, meeting or exceeding time-saving thresholds for many routine ad variants. However, brand-specific tone, legal compliance, and ensuring genuine persuasive impact for niche markets may still require human oversight, preventing a perfect 5. |
| Task automatability | claude-sonnet-5 | 4/5 | LLMs can generate ad copy variants for publication, broadcast, or digital media rapidly, and many marketing teams already replace significant drafting effort with AI-generated first drafts requiring only light editing. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard legal or regulatory barriers exist; advertising is not a licensed profession. Light friction remains from brand guidelines, FTC compliance review, and organizational preference for human creative judgment, but nothing prevents substitution. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing or legal requirement for a human to write advertising copy, and no regulatory barrier prevents AI-generated ad content. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost per ad variant is orders of magnitude cheaper than hiring a copywriter, even after accounting for prompt engineering, review, and integration overhead. One AI system can generate thousands of ad variations at near-zero marginal cost. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating multiple ad copy drafts via API costs cents versus hours of copywriter time, an order-of-magnitude cost advantage even after factoring in review costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed AI copywriting tools (GPT-4, Claude, specialized ad platforms) reliably generate advertising material in production for publishers, e-commerce, and agencies. Quality is production-ready for routine campaigns, though complex strategic messaging or high-stakes brand work still shows material error rates. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like Jasper, Copy.ai, and ChatGPT are used in production by marketing agencies and in-house teams to generate ad copy at scale, though final review and brand voice tuning still occur. |
Vary language and tone of messages based on product and medium.
81CI 79–84 · exposure 75 · augmentation 100 · importance 4.4/5 · click for rater detail
Vary language and tone of messages based on product and medium.
81| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Marketing, tech, and professional services firms are rapidly deploying AI for content variation and copywriting. Tools like Jasper and Copy.ai report widespread production use; adoption is measurably accelerating in digitized, information-sector contexts where this task is common. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Marketing, advertising, and content industries have rapidly integrated generative AI tools into copywriting workflows, representing fast adoption typical of information/professional-services sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically enhances writer productivity by generating multiple tone and style variants instantly, enabling rapid A/B testing and audience-specific adaptation. Writers stay in the loop to select and refine, but AI transforms their throughput on this specific task. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is widely used as a drafting and brainstorming aid for tone/style variation, letting writers quickly generate and iterate on options while retaining final creative control. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current LLMs can effectively adjust tone, register, and style across different contexts and mediums (formal/casual, social/professional, audio/text). While a human author might add subtle nuance or brand-specific voice refinement, AI systems can automate 50%+ of the message variation work with minimal oversight, particularly for templated or category-based content variations. |
| Task automatability | claude-sonnet-5 | 4/5 | Modern LLMs can readily generate copy in varied tones and styles tailored to product and medium, meeting or exceeding the 50% time-saving threshold for drafting work, though final polish and brand judgment still benefit from human review. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Minimal legal or regulatory barriers exist for AI-generated message variations. The main friction is organizational (human preference for author review, brand consistency checks) and customer perception, but nothing prevents deployment. No licensing requirement or liability asymmetry blocks adoption. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, regulatory, or human-signoff requirement governs tone/style adaptation in marketing or general writing; adoption is purely a business choice. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | API costs for LLM inference are a fraction of a cent per message variation, while a professional copywriter or content specialist costs $25–75/hour or more. Even with human oversight and integration overhead, AI cost is typically 1–2% of human labor cost per output unit. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating multiple tonal variants via AI costs cents to dollars per piece versus hours of a writer's time, an order-of-magnitude cost advantage for this specific subtask. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (ChatGPT, Claude, Jasper, Copy.ai) reliably perform tone and language adaptation across mediums in production. These systems are actively used by marketing, communications, and content teams. Minor limitations exist in capturing highly distinctive brand voices, but the core task executes dependably at scale. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed products (Jasper, Copy.ai, ChatGPT, Claude) are widely used in marketing and content teams today to produce tone-adapted copy across channels, with reliable output for common formats though occasional need for editing. |
Write articles, bulletins, sales letters, speeches, and other related informative, marketing and promotional material.
81CI 79–84 · exposure 75 · augmentation 100 · importance 3.7/5 · click for rater detail
Write articles, bulletins, sales letters, speeches, and other related informative, marketing and promotional material.
81| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Marketing, SaaS, e-commerce, and publishing sectors are rapidly adopting AI copywriting in production (email campaigns, landing pages, social media, internal bulletins). Adoption is measurably fast in digitally native sectors. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Marketing, advertising, and content industries have rapidly integrated AI writing tools into everyday workflows, among the fastest-adopting knowledge-work sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI drafting tools dramatically accelerate writer productivity on brainstorming, outlining, revision, and multi-variant generation while writers retain editorial control, making augmentation a clear strength of current AI systems. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is heavily used as a drafting and ideation assistant for writers, dramatically speeding up production while humans retain editorial control and final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can generate complete first drafts of articles, marketing copy, speeches, and promotional material at scale with minimal oversight, meeting the 50% time-saving bar. However, nuance, brand voice consistency, and fact-checking still often require human review, preventing a full 5 rating. |
| Task automatability | claude-sonnet-5 | 4/5 | LLMs can draft articles, bulletins, and marketing copy quickly, and for many routine formats this meets or exceeds the 50% time-saving bar, though high-stakes or highly branded content still needs human refinement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or regulatory barriers exist; no licensing requirement to write marketing or informational material. Main friction is organizational (brand governance, fact-checking liability, human preference for bylines), not legal. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or legal requirement mandates a human author for marketing and promotional writing; adoption is purely a business choice. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference and integration cost for generating articles and marketing material is orders of magnitude cheaper than freelance or staff writer wages, especially at volume. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | AI subscription costs are a tiny fraction of a writer's hourly wage for producing comparable volumes of draft marketing copy. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (Claude, GPT-4, specialized marketing AI tools) reliably generate publishable marketing and informational content in production. Error rates on factual claims and tone mismatches remain concerns for some use cases, but broad feasibility is well-established. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like ChatGPT, Jasper, and Copy.ai are widely deployed in production for marketing and content writing, though quality control and brand voice consistency still require human editing. |
Edit or rewrite existing written material as necessary, and submit written material for approval by supervisor, editor, or publisher.
66CI 66–66 · exposure 59 · augmentation 100 · click for rater detail
Edit or rewrite existing written material as necessary, and submit written material for approval by supervisor, editor, or publisher.
66| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Publishing, journalism, and content industries have rapidly adopted AI writing assistants (Grammarly, ChatGPT integration in editorial platforms). Adoption is measurable in production workflows, though some high-prestige literary publishers still resist full automation of editorial judgment. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Publishing, media, and content industries have rapidly adopted AI writing/editing tools as part of standard workflows, reflecting fast adoption patterns typical of information-sector work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI excels as an augmentation layer: writers can use AI to draft initial rewrites, flag inconsistencies, and suggest style improvements while the human author retains control and judgment. This workflow demonstrably boosts productivity and is already widely adopted. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI substantially enhances writer productivity by suggesting edits, restructuring text, and catching errors, while the human retains control over final approval and creative decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate substantial portions of editing (grammar, style, consistency checks) and rewrite sections autonomously, but final judgment about material quality, tone alignment, and approval-readiness typically requires human oversight. The task involves human judgment about what constitutes acceptable output for submission, limiting full end-to-end automation. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can perform substantial editing and rewriting (grammar, style, structure suggestions) at reasonable quality, but the submission-for-approval step and final creative judgment still require human involvement, so only part of the task meets the 50% time-saving bar. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard barriers exist: editors and authors are not legally required to perform their own edits, and AI tools integrate into standard workflows without licensing friction. The main barriers are organizational practice and author/editor preference for human judgment, which are soft and eroding. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for writers/editors, but organizational workflows still require human submission and sign-off from supervisors/editors/publishers, creating some structural friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Current AI inference costs for editing tasks are very low (pennies per document), while loaded costs for human editors are substantial (>$30/hour). Even with integration and human oversight overhead, AI-assisted editing is orders of magnitude cheaper per unit of editing performed. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI editing tools cost a small fraction of a human editor's time for comparable draft revisions, though human oversight is still needed for final quality control. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products (Grammarly, Claude, GPT-4, specialized editing tools) reliably perform editing and rewriting in production today with acceptable error rates. However, approval submission remains a human responsibility in most professional contexts, and error rates on nuanced rewrites can still be material for high-stakes publishing. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like Grammarly, ChatGPT, and Claude are widely deployed in production for editing and rewriting tasks with generally reliable results, though quality varies with genre and nuance. |
Invent names for products and write the slogans that appear on packaging, brochures and other promotional material.
64CI 50–79 · exposure 55 · augmentation 88 · importance 3.5/5 · click for rater detail
Invent names for products and write the slogans that appear on packaging, brochures and other promotional material.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Marketing and advertising sectors show active pilot and early production adoption of generative AI for ideation and copy drafting, but full replacement remains limited by client expectations for human creative judgment; adoption is faster than traditional industries but slower than technical domains. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Marketing and advertising, a fast-adopting professional services sector, has rapidly integrated generative AI for copywriting and ideation tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | LLMs excel at rapidly generating multiple creative alternatives for copywriters to refine, significantly accelerating brainstorming and iteration; human writers report measurable productivity gains when using AI to augment rather than replace their creative process. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dramatically accelerates brainstorming of names and slogans, letting copywriters generate and refine many more options while retaining creative judgment and final selection. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While LLMs can generate product names and slogans at scale, creative output often requires brand alignment, market differentiation, and cultural sensitivity that typically demands human refinement; automation rarely meets the ≥50% time-saving bar when accounting for review cycles and revisions needed to meet quality standards. |
| Task automatability | claude-sonnet-5 | 4/5 | LLMs can generate large batches of candidate names and slogans very quickly, and this generative, low-precision-required task fits well within current text-generation capabilities, though human curation and trademark screening remain necessary. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No legal requirement mandates a licensed human create marketing copy; however, brand risk, client approval workflows, and organizational preference for human creative input create modest adoption friction rather than hard barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing is required to write slogans, though trademark clearance and brand risk create some incentive for human review and legal oversight before final adoption. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | API-based generation of dozens of name/slogan variations costs pennies per iteration, vastly cheaper than paying a copywriter's loaded wage for equivalent volume; even with integration and human review time factored in, AI generation provides substantial cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating hundreds of name/slogan candidates via an LLM costs pennies compared to hours of a copywriter's or naming consultant's billable time. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed LLM products can generate candidate names and slogans with acceptable speed, but material limitations exist: brand consistency enforcement, legal/trademark checking, and cultural appropriateness require human oversight; no production system fully autonomously owns this end-to-end. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed tools (ChatGPT, Copy.ai, Jasper, naming-generator products) are routinely used in marketing workflows today to brainstorm names and slogans, though final selection and legal vetting still involve humans. |
Follow appropriate procedures to get copyrights for completed work.
64CI 43–85 · exposure 66 · augmentation 75 · click for rater detail
Follow appropriate procedures to get copyrights for completed work.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Legal tech and publishing platforms have begun integrating copyright filing automation, but adoption remains piecemeal; most individual writers and smaller publishers still file manually or through traditional legal services, reflecting slower digitization in creative industries. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Writers and authors as a group show slow, uneven AI adoption for administrative/legal tasks like copyright filing, which remains a niche, infrequent task with limited tool specialization. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI substantially assists by auto-populating forms, checking completion requirements, and scheduling filings, allowing a human writer to focus on content rather than administrative logistics, with the human maintaining simple supervisory role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully help authors understand copyright requirements, draft application materials, and organize necessary documentation, streamlining a task that is otherwise unfamiliar and tedious for many. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Copyrighting work involves filling out standardized forms, submitting them to registrars with required documentation, and tracking responses—tasks that AI can execute end-to-end today using form-filling agents and email automation, achieving >50% time savings versus manual submission and tracking. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft copyright applications, fill forms, and explain registration procedures, but the actual filing and legal decision-making steps still require human review and submission through official channels. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Copyright registration is a ministerial, not discretionary, task and requires no licensed human signature; while the Copyright Office must receive the filing and conduct its own review, the submission procedure itself has no legal barrier to automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Copyright registration involves official government processes (e.g., U.S. Copyright Office) with legal accountability, requiring accurate human-verified submissions, though no license is strictly required to file. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Copyright filing via AI-driven agents costs only API fees and minimal oversight, roughly $1–10 per filing, compared to hiring a lawyer or spending several hours of a writer's billable time ($50–200+), yielding at least 10× cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI assistance for drafting and understanding procedures is cheap, but the overall task still requires human time for filing and verification, keeping total cost roughly comparable to doing it manually with some AI help. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed AI products and legal tech platforms can automate copyright registration workflows, including form generation and filing coordination; while human review of final submissions is often prudent, the underlying procedural task is reliably handled by production systems. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | No deployed product autonomously handles end-to-end copyright registration; existing tools (legal chatbots, form-fillers) assist but don't reliably complete the process without human oversight. |
Plan project arrangements or outlines, and organize material accordingly.
62CI 59–66 · exposure 50 · augmentation 100 · click for rater detail
Plan project arrangements or outlines, and organize material accordingly.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Publishing, content agencies, and media organizations are actively using AI for outline generation and project organization in production workflows. Adoption is accelerating in information-sector jobs where planning is routine and iterative feedback is feasible. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Publishing and content creation have seen notable AI tool adoption for brainstorming and outlining, but many professional writers still resist full delegation of structural planning, keeping adoption at a moderate pace. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI excels at augmenting writers during planning: rapid brainstorming, generating alternative outlines, organizing sources, and suggesting structure refinements. Writers routinely report that AI planning assistance accelerates their workflow while they retain creative direction and final decisions. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is widely and effectively used by writers to generate outline drafts, alternative structures, and organizational suggestions, significantly speeding up the planning phase while the author retains creative control. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can generate outlines and organize material into structured formats, but typically requires significant human refinement for coherence, strategic vision, and domain-specific prioritization. Current systems handle partial automation (outline generation, categorization) well but struggle with the full end-to-end planning that integrates multiple project constraints and creative intent. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate outlines and structural plans quickly, but a writer's project arrangement often reflects unique creative vision, voice, and strategic choices that require significant human judgment and revision to reach equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No licensing, authorization, or regulatory requirement mandates human involvement in planning and outline creation. Organizational friction exists mainly around writers' preferences to maintain creative control, but no hard legal or contractual barrier prevents substitution. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There are no licensing, legal, or regulatory requirements around outlining creative or nonfiction works, so no formal barrier prevents AI assistance or substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | API costs for outline generation and material organization are low relative to skilled writer time spent on planning. A single query can generate initial structure at near-zero marginal cost, making AI-assisted planning substantially cheaper than paying a human for the same task from scratch. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Generating draft outlines via AI costs a small fraction of a writer's time-based fee, though the human review and refinement needed keeps it from being a full order-of-magnitude savings for the whole task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Multiple products (Claude, GPT-4, specialized outline tools) can generate drafts and organize content reliably, but they often require human iteration and judgment to finalize plans that reflect authentic project goals. Production use is real but typically in an assistive mode rather than fully autonomous. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Writing tools and LLMs are routinely used by authors to brainstorm and draft outlines, but they still require substantial human curation and often miss nuanced narrative or thematic goals, so reliability varies by project type. |
Write to customers in their terms and on their level so that the script, story, or advertisement message is more readily received.
62CI 52–71 · exposure 47 · augmentation 88 · click for rater detail
Write to customers in their terms and on their level so that the script, story, or advertisement message is more readily received.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Marketing, advertising, and content teams in digital-first sectors (tech, e-commerce, media) are rapidly deploying AI copywriting tools in production; adoption is visible and accelerating in professional services and information industries. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Marketing, advertising, and content industries have rapidly adopted generative AI tools for audience-tailored writing, with widespread production use across agencies and in-house teams. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at rapidly generating multiple audience-targeted drafts, tone variations, and A/B copy alternatives, meaningfully amplifying a human writer's productivity while the writer retains creative direction and final judgment. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly effective as a drafting and brainstorming assistant, helping writers quickly generate audience-adapted variations, test tones, and speed revision cycles while the writer retains final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate text adapted to different registers and audiences, the task requires understanding subtle contextual customer psychology and brand voice nuances that AI systems today struggle to match consistently without human oversight and iteration. |
| Task automatability | claude-sonnet-5 | 3/5 | LLMs can draft audience-tailored copy quickly and adjust tone/register on request, but achieving truly resonant, brand-consistent, and creatively distinctive writing at equal quality across contexts still often requires human revision.5There is meaningful time savings for drafting but not full end-to-end replacement at equal quality for all writing genres. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or licensing barriers exist; this is primarily business-driven adoption. Marketing departments and content teams face minimal regulatory constraints on automation, though brand risk and customer trust concerns create some organizational friction. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or regulatory requirement mandates a human write marketing copy or stories; the main constraint is quality/brand fit rather than legal barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-generated initial drafts and variations are inexpensive per iteration, and with prompt engineering and templates, the per-output cost is substantially below hiring a copywriter, though human review still adds cost. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting costs pennies per piece versus a writer's hourly rate, making it far cheaper for first-draft or iterative copy generation, though human review still adds some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI writing tools and GPT-based systems can produce audience-adapted copy, but deployed products often require significant human editing to ensure authentic voice, cultural appropriateness, and strategic alignment—they are not yet reliably autonomous end-to-end. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like ChatGPT, Jasper, and Copy.ai are widely deployed in marketing and content teams today for audience-targeted copywriting, with real production use, though quality control and brand voice consistency remain imperfect. |
Review advertising trends, consumer surveys, and other data regarding marketing of goods and services to determine the best way to promote products.
61CI 61–61 · exposure 50 · augmentation 88 · importance 3.7/5 · click for rater detail
Review advertising trends, consumer surveys, and other data regarding marketing of goods and services to determine the best way to promote products.
61| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Marketing and advertising are information-heavy, digitized sectors with rapid AI adoption; analytics platforms, data dashboards, and AI-assisted market intelligence are deployed widely in large and mid-market organizations. Production use of AI for trend analysis and data-driven marketing recommendations is common and accelerating. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Marketing and advertising are among the faster-adopting sectors for AI, with widespread use of AI-driven analytics and content tools already embedded in agency and in-house workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists marketing professionals by rapidly synthesizing large volumes of survey and trend data, generating candidate insights, and surfacing patterns that would take humans weeks. These tools augment human strategists' productivity substantially while the human retains oversight and final decision authority. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI excels at augmenting this task by rapidly analyzing consumer data, summarizing trends, and generating promotional angle options, significantly speeding up the research phase while writers retain final creative and strategic control. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can analyze structured marketing data, surveys, and trends at scale to identify patterns and generate insights faster than humans, but requires significant human judgment on strategy formulation, creative direction, and deciding which findings are actionable. The end-to-end task of determining 'the best way to promote' involves strategic synthesis that typically falls short of 50% time savings at equal quality without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can synthesize survey data, trend reports, and market research into summarized recommendations quickly, but determining the 'best way to promote' still often requires human strategic judgment, brand knowledge, and creative synthesis that current tools only partially replicate. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard regulatory or legal barriers exist for using AI to analyze marketing data; this work does not require licensing or mandatory human sign-off. Organizational adoption faces some friction from preference for human judgment in strategic marketing decisions, but these are softer barriers than in regulated fields. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates human involvement in this analytical task, though organizations may prefer human judgment for brand-sensitive positioning decisions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven market analysis and trend detection is substantially cheaper than hiring analysts or consultants to manually review surveys and trend reports. Inference and data processing costs are low relative to professional labor rates, though integration and human oversight add overhead. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI can process and summarize large volumes of survey and trend data far faster and cheaper than a human analyst or writer, though some data integration and interpretation still requires paid tools or human oversight. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed AI tools (analytics platforms, LLMs with data analysis capabilities) can process advertising trends and survey data reliably, but production systems handling this task show material limitations in nuanced interpretation, contextual application, and strategic recommendation quality. Such tools exist and are in use but typically as assistants rather than end-to-end solvers. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like ChatGPT, Claude, and specialized marketing analytics tools (e.g., Jasper, HubSpot AI) are deployed for trend analysis and content strategy suggestions, but they still require human validation and are not fully autonomous decision-makers in production marketing workflows. |
Conduct research to obtain factual information and authentic detail, using sources such as newspaper accounts, diaries, and interviews.
59CI 55–64 · exposure 55 · augmentation 75 · click for rater detail
Conduct research to obtain factual information and authentic detail, using sources such as newspaper accounts, diaries, and interviews.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Journalism and publishing have begun experimenting with AI for research and drafting (e.g., newsroom automation, content generation aids), but adoption remains cautious due to accuracy concerns and editorial standards. Pilots are common in tech-forward outlets, but widespread production deployment is still emerging. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Writing and publishing sectors are adopting AI research assistants moderately, with common pilot use for drafting and background research but human-led primary research still standard. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at augmenting writers by rapidly gathering candidate sources, summarizing documents, and highlighting key facts, allowing writers to focus on analysis, synthesis, and creative framing. This assistant role significantly raises productivity for the human researcher without removing the human from the loop. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up background research, summarization, and organizing secondary sources, though it cannot replace human-conducted interviews or original archival discovery. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate significant portions of research—gathering and summarizing text from publicly available sources, extracting facts from documents, and synthesizing information from multiple sources. However, authentic detail often requires judgment about source credibility, context-specific knowledge, and identifying which details truly matter for the work, which still requires human oversight and verification. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can rapidly gather and summarize information from text corpora and search results, but verifying authenticity, sourcing original diaries, or conducting real interviews still requires human effort and judgment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist to automating research gathering itself; however, professional norms in journalism and serious writing still favor human verification and byline accountability, and some publications require human fact-checking. These are soft barriers rather than hard legal requirements. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but journalistic/literary standards for authenticity and primary-source verification create moderate friction against pure automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-powered research tools are now substantially cheaper than paying researchers or writers to conduct manual research, with inference costs minimal and integration straightforward. The all-in cost per unit of research output is likely an order of magnitude lower than hiring a person to do the same research. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply do preliminary literature/web research, but human fact-checking, interviewing and verification remain costly, keeping overall cost roughly comparable for authentic, high-quality research. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products like ChatGPT, Claude, and specialized research tools reliably retrieve and synthesize information from available sources today. Search-augmented AI systems can locate relevant newspaper accounts and diaries; however, they occasionally hallucinate citations and may miss nuanced or obscure sources that human researchers would find. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Deployed tools (search-augmented LLMs, research assistants) reliably summarize secondary sources today, but they cannot conduct primary interviews or access many archival/physical sources like diaries. |
Prepare works in appropriate format for publication, and send them to publishers or producers.
55CI 37–72 · exposure 45 · augmentation 63 · click for rater detail
Prepare works in appropriate format for publication, and send them to publishers or producers.
55| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption remains slow; most authors and small/mid-size publishers manage submissions manually or with basic document tools. Publishing is a traditionally slow-moving sector with high manual workflow integration, and AI-driven submission automation is not yet mainstream in production systems. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Publishing and media sectors show moderate AI tool adoption for formatting and submission assistance, but many authors and small presses still rely on manual or semi-manual processes, keeping adoption in the pilot-to-moderate range. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by auto-formatting documents to publisher specs, generating metadata, organizing submission files, and checking format compliance—genuinely useful assistance that saves time. However, final judgment about suitability, submission strategy, and publisher selection remain human tasks. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools significantly speed up formatting, style-checking, and metadata preparation, letting authors focus on content while automation handles the mechanical publication-prep steps. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with formatting and document preparation (converting files, applying templates, basic metadata), the task requires judgment about the specific publication format, editorial standards, and context-appropriate submission that still demands human oversight. End-to-end automation with 50% time saving at equal quality is not reliably achieved today. |
| Task automatability | claude-sonnet-5 | 4/5 | Formatting manuscripts to publisher/producer specifications (style guides, manuscript formatting, metadata, file conversion) is largely mechanical and AI tools can handle most of this with templates and automated formatting checks, though final submission logistics and relationship management remain human tasks.“}, |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Publishers have specific submission requirements and gatekeeping processes, and human contact with editors is often preferred or required. However, there is no legal or licensing barrier preventing submission by automated means if format and content requirements are met, creating moderate organizational and process friction rather than hard barriers. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing or regulatory requirement that a human format and submit a manuscript; it's purely a technical/administrative task with minimal legal or professional barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | The AI cost for document formatting and submission support is very low (fractions of a dollar in inference), while the human time saved on formatting tasks is modest (typically minutes per submission), making automation highly cost-effective relative to the human wage for that formatting labor. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated formatting tools and templates cost very little compared to a human spending hours reformatting a manuscript, making AI-assisted formatting substantially cheaper at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No production system reliably handles the full submission workflow (format selection, compliance with publisher-specific requirements, metadata completeness, submission portal navigation) without human review. Tools exist for individual steps (document conversion, formatting templates) but not for integrated, end-to-end submission management. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Word processors, formatting software, and AI writing assistants (e.g., Grammarly, Vellum, Scrivener integrations) reliably handle formatting today, but full end-to-end preparation and submission workflows still require human oversight for publisher-specific quirks and personal correspondence. |
Write fiction or nonfiction prose, such as short stories, novels, biographies, articles, descriptive or critical analyses, and essays.
52CI 45–59 · exposure 38 · augmentation 88 · click for rater detail
Write fiction or nonfiction prose, such as short stories, novels, biographies, articles, descriptive or critical analyses, and essays.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Publishing, marketing, and media firms are piloting AI for commodity content (headlines, summaries, SEO articles), but significant literary fiction and serious nonfiction remain author-driven. Adoption is faster in low-stakes, high-volume segments; slower in prestige and craft-heavy sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Publishing, journalism, and content industries show substantial AI tool adoption for drafting and idea generation, but human-authored final products remain the norm in professional/literary contexts, so adoption is uneven. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI provides substantial assistance: drafting, ideation, editing feedback, research synthesis, and prose polishing. Many writers actively use these tools to accelerate composition and revision, boosting productivity while maintaining human creative control and judgment. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is heavily used by writers for brainstorming, drafting, editing, and overcoming writer's block, significantly boosting productivity while the author retains creative control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can generate prose quickly, but end-to-end autonomous fiction/nonfiction meeting publication quality and achieving 50% time savings over skilled human writers remains rare. AI drafts typically require substantial human revision for coherence, originality, and thematic depth. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft substantial prose quickly, but producing publishable-quality fiction/nonfiction with voice, structure, and originality still requires heavy human revision and judgment, so full end-to-end automation at equal quality isn't yet reliable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No legal requirement for a licensed human; copyright and attribution concerns exist but do not block deployment. Market and reputational barriers are real (audience preference for human authorship, editorial trust), but these are soft constraints rather than hard regulatory or liability blocks. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but publishing norms, authorship attribution, copyright concerns, and readers' preference for human-authored work create moderate friction against full substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Inference cost for generating prose is very low (pennies per article or story), while human writer compensation ranges from $30–$200+ per hour. Even accounting for quality oversight and revision, AI-generated text is substantially cheaper at scale than commissioned human writing. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Generating draft prose via AI costs a tiny fraction of a writer's hourly rate, though human editing/oversight is still needed, keeping it below a full 5. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI writing tools exist (GPT-based, etc.) but deployed products are primarily assistive rather than autonomous; they produce drafts requiring heavy editorial oversight. Production use cases show AI generates content, but error rates in fact-checking, narrative consistency, and voice prevent reliable autonomous completion. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like ChatGPT, Claude, and specialized writing tools are widely used for drafting articles, outlines, and short pieces, but longer-form professional-quality work (novels, investigative articles) still shows inconsistency, factual errors, and generic style. |
Revise written material to meet personal standards and to satisfy needs of clients, publishers, directors, or producers.
49CI 36–61 · exposure 38 · augmentation 88 · click for rater detail
Revise written material to meet personal standards and to satisfy needs of clients, publishers, directors, or producers.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Information and publishing sectors show growing adoption of AI-assisted drafting and editing tools, but primarily for augmentation rather than replacement. Pilots are common; autonomous full revision remains rare in production. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Publishing, media, and content industries have rapidly adopted AI writing/editing tools, with widespread use of AI-assisted drafting and revision workflows already common. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI writing assistants substantially augment human revisers by flagging grammar, consistency, and style issues, generating alternative phrasings, and accelerating iterative cycles. The human writer maintains judgment while AI productivity multipliers are significant. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly effective as a revision aid—offering suggestions, style edits, and restructuring—while the human retains control over final creative and client-driven decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with surface-level revision (grammar, style suggestions) but cannot reliably assess or enforce personal standards, client intent, or thematic coherence without substantial human oversight. The task requires judgment about subjective quality and stakeholder satisfaction that current systems cannot perform end-to-end. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft revisions, tighten prose, and suggest edits well, but final revision requires human judgment on voice, client-specific intent, and creative standards, so only partial automation meets the equal-quality bar. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While there are no strict licensing barriers, significant friction exists through client preference for human judgment, liability concerns over attribution and errors, and organizational reluctance to cede creative control entirely to automation. Publishers and producers typically demand human sign-off. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for revising creative or professional writing, though client/publisher preference for human authorial voice and contractual authorship expectations create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI-powered writing tools have dropped in cost and can reduce revision time for routine edits, but integrating them into a professional workflow with quality oversight remains labor-intensive, making the all-in cost roughly comparable to experienced human revision. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted revision is very cheap per pass compared to professional editor or writer time, though human review/oversight is still typically needed, moderating full order-of-magnitude savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI writing assistants (Grammarly, ChatGPT) exist in production, they operate as suggestion tools requiring human review rather than autonomous revision engines. Deployed systems lack the contextual understanding and accountability to independently satisfy diverse client standards. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like ChatGPT, Grammarly, and specialized editing tools are used in production for revision assistance, but reliability on nuanced creative or client-specific requirements varies and often needs human correction. |
Conduct research and interviews to determine which of a product's selling features should be promoted.
41CI 32–50 · exposure 33 · augmentation 75 · importance 3.6/5 · click for rater detail
Conduct research and interviews to determine which of a product's selling features should be promoted.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Marketing and content teams are experimenting with AI research tools and drafting assistants, but actual decision-making about which features to promote remains largely human-driven. Adoption is in the pilot and early-adoption phase, not yet at production scale. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Marketing and content teams are adopting AI tools for research and drafting at a moderate pace, with pilots common but full workflow automation for feature prioritization still uneven. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at augmenting this task: it can rapidly synthesize competitive research, suggest feature angles, summarize customer feedback, and help draft talking points—all while the human strategist retains control over final promotion decisions. This is a strong case for AI-assisted productivity gain. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up preliminary research, competitive analysis, and drafting interview questions or summaries, meaningfully boosting a writer's productivity while they retain control over interviews and final feature selection. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with research synthesis and initial competitive analysis, but determining *which* features to promote requires nuanced judgment about market positioning, customer psychology, and brand strategy that current systems cannot reliably execute end-to-end. Interviews themselves remain difficult for AI to conduct authentically. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can synthesize existing research and market data quickly, but conducting actual interviews and forming original judgment about which features to promote requires human interaction and strategic insight that current AI cannot fully replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Marketing and brand strategy decisions typically require human accountability and creative judgment; most organizations still expect a human writer or strategist to own feature-promotion choices. However, no formal licensing barrier exists, so adoption is organizationally/culturally gated rather than legally mandated. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists, but organizational preference for human relationship-building in interviews and brand strategy judgment creates moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for research aggregation and data synthesis are cheap, but the core task—deciding which features to promote—still requires significant human expertise and judgment. The cost savings are modest relative to the loaded wage of a skilled writer/strategist. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply handle desk research and data synthesis, but human interviewers and strategic judgment still add significant cost, making the blended task cost roughly comparable to a human doing it alone with AI assistance. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can gather and summarize existing market data and competitor information, but no deployed product reliably conducts original interviews or makes trustworthy promotion-strategy recommendations without human oversight. The human-facing research component remains largely manual. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI research assistants and summarization tools are deployed for background research, but the interview component and final synthesis into promotable feature selection is not something production AI systems reliably perform independently today. |
Work with staff to develop script, story, or advertising concepts.
39CI 32–45 · exposure 25 · augmentation 75 · click for rater detail
Work with staff to develop script, story, or advertising concepts.
39| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Marketing and advertising sectors are actively piloting AI for copy and concept generation, but most still treat it as an assistant tool rather than autonomous development; adoption is accelerating but remains mixed across media and publishing. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Media, advertising, and entertainment sectors are adopting AI writing assistants at a moderate pace, with pilots and partial integration in ideation workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at augmenting writers by rapidly generating multiple concept variations, providing drafts for critique, and brainstorming—keeping the human in the creative loop while dramatically accelerating the exploration and iteration phase of script and story development. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can generate brainstorming material, alternate story angles, or ad copy variations that meaningfully speed up and enrich staff collaboration sessions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can generate rough story concepts and advertising copy quickly, but collaborative script/story development with staff requires iterative creative judgment, stakeholder alignment, and domain-specific vision that current AI cannot reliably handle end-to-end without extensive human direction and revision. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires live collaborative negotiation, creative judgment, and interpersonal team dynamics that current AI cannot fully replicate end-to-end, though it can generate draft ideas to discuss. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Creative work carries reputational and IP risk; many organizations require human creative sign-off and are reluctant to fully automate concept development for legal and brand-safety reasons, though no hard licensing barrier exists. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but organizational and creative-team culture favors human collaboration and creative ownership, creating moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference for concept generation is very cheap per unit, but integration into a real collaborative workflow with staff coordination and oversight reduces the ratio advantage; still likely 5–10× cheaper than a human writer's labor for the raw generation phase. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI drafting is cheap, the collaborative/interpersonal component still requires paid human staff time, limiting overall cost savings for this specific task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI writing tools exist and can produce draft concepts, no production system reliably performs the full collaborative development task—which requires understanding organizational context, integrating feedback from multiple stakeholders, and maintaining coherent creative direction across iterations. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI writing tools can suggest concepts or drafts, but no deployed product actually 'works with staff' in real collaborative meetings to develop concepts reliably. |
Discuss with the client the product, advertising themes and methods, and any changes that should be made in advertising copy.
35CI 32–38 · exposure 25 · augmentation 75 · importance 4.2/5 · click for rater detail
Discuss with the client the product, advertising themes and methods, and any changes that should be made in advertising copy.
35| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Advertising and marketing firms are piloting AI-assisted copywriting and analysis, but autonomous client discussions remain rare in production; adoption is middling, concentrated in copy generation rather than strategic consultation. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Advertising and marketing sectors are moderately fast adopters of AI tools for drafting and analysis, but client-facing consultative work still involves humans predominantly. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can significantly assist by generating multiple advertising themes, analyzing copy performance, and proposing revisions in real time, allowing human writers to conduct client discussions more efficiently and creatively while retaining full control over strategic decisions and relationships. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can help prepare talking points, analyze ad performance data, summarize themes, and draft copy variations to support the writer during and after client discussions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate advertising copy and analyze themes, the interactive discussion element—understanding client preferences, negotiating changes, and building consensus—requires human judgment and relationship management that AI cannot reliably perform end-to-end with equal quality outcomes. |
| Task automatability | claude-sonnet-5 | 2/5 | This task is a live, relational discussion requiring real-time interpretation of client intent, negotiation, and adaptive dialogue, which current AI cannot fully replace end-to-end.}, |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Client relationships and approval authority create organizational friction and customer preference for human interaction; however, no hard legal or licensing barrier explicitly prevents automation of discussion tasks, making these barriers moderate rather than prohibitive. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but client preference for human relationship-building and trust in creative decision-making creates moderate organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI copy generation has low per-unit cost, the discussion component requires either human facilitation or expensive AI systems with conversational sophistication; total integration cost to replace a skilled account executive remains high relative to human wage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Human account/creative staff still needed for judgment and relationship management, so AI reduces but doesn't eliminate labor cost, and integration/oversight costs offset savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI systems can draft advertising themes and copy suggestions, but no deployed product reliably conducts authentic, back-and-forth client discussions with the adaptive listening and contextual reasoning this task demands; such systems remain primarily assistive rather than autonomous in production. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI chatbots can hold structured conversations but no deployed product reliably manages nuanced client relationship discussions and creative direction negotiation in production today. |
Develop advertising campaigns for a wide range of clients, working with an advertising agency's creative director and art director to determine the best way to present advertising information.
34CI 30–38 · exposure 25 · augmentation 75 · importance 4.5/5 · click for rater detail
Develop advertising campaigns for a wide range of clients, working with an advertising agency's creative director and art director to determine the best way to present advertising information.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Advertising agencies are digitally sophisticated but adoption remains concentrated in supplementary copy generation and brainstorming aids; core campaign development responsibility still rests with human creatives in most production workflows. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Advertising and marketing sectors show fast AI tool adoption for content generation, though full campaign strategy work still involves significant human-led production processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools meaningfully augment writers and creative teams by rapidly generating multiple copy variations, brainstorming headlines and angles, and accelerating the ideation phase, allowing humans to focus on strategy, client relationships, and creative direction—a clear productivity boost while humans remain central to decision-making. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up brainstorming, drafting taglines, and generating creative options, giving writers a strong productivity boost while directors and writers retain creative control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate copy drafts and suggest campaign concepts, developing advertising campaigns requires iterative collaboration, strategic judgment about client positioning, and alignment with creative directors' vision—tasks where AI produces candidates but cannot reliably complete the full workflow end-to-end at quality parity without substantial human oversight and revision. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can draft copy and generate concept variations quickly, but the collaborative, client-specific strategic development involving creative/art directors requires human judgment, negotiation, and taste that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no legal licensing is strictly required, client relationships, creative accountability, and organizational norms strongly favor human creative professionals leading strategy and bearing responsibility for campaign success, creating meaningful adoption friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists, but client relationships, brand trust, and creative accountability create moderate organizational friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI systems reduce time on ideation and drafting but cannot eliminate the strategic and collaborative roles of creative directors and writers, so the marginal cost savings remain modest relative to full loaded compensation for experienced campaign developers. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI can cheaply generate draft copy, the overall task requires significant human oversight, revision, and interpersonal coordination, keeping total cost closer to human-comparable levels. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs end-to-end campaign development autonomously; AI copywriting tools exist but require heavy creative direction and produce output that typically needs significant human refinement before client presentation or media deployment. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI copywriting tools are used in production for drafting and ideation, but no deployed product reliably manages full campaign development including cross-functional creative direction collaboration. |
Collaborate with other writers on specific projects.
29CI 21–38 · exposure 25 · augmentation 75 · click for rater detail
Collaborate with other writers on specific projects.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While writing tools are widely adopted, AI-driven automation of inter-writer collaboration remains rare in production. Most adoption is assistive (drafting, editing suggestions) rather than replacing the collaborative relationship itself. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Publishing and media are moderately fast adopters of AI writing tools for drafting and editing, but collaborative authorship workflows still rely heavily on human coordination. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist collaboration through real-time suggestion, version tracking, research synthesis, and style consistency feedback, allowing human writers to focus on creative decisions and interpersonal coordination. These tools noticeably raise productivity when the human remains the decision-maker. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly assist collaborative writing by suggesting edits, reconciling versions, generating drafts, and summarizing feedback, boosting the productivity of co-writing teams. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with drafting and revision coordination but cannot fully replace the creative negotiation, ideation, and interpersonal problem-solving central to writer collaboration. The task requires ongoing back-and-forth dialogue, creative compromise, and human judgment that current AI cannot meaningfully execute end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Collaboration involves interpersonal negotiation, creative alignment, and real-time judgment calls that AI cannot autonomously replace; AI can only assist portions of the collaborative writing process like drafting or merging text. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Collaboration is inherently human-contact dependent; contractual agreements, creative control concerns, and industry norms strongly protect human authorship and collaborative decision-making. Replacing a writer's role in collaboration faces deep organizational and contractual friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but creative ownership, credit, and trust dynamics between collaborators create organizational and interpersonal friction against full AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI-assisted collaboration tools plus the human overhead of using them effectively remains comparable to or exceeds the cost of direct writer-to-writer collaboration. Humans still drive the essential creative and interpersonal work. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools reduce some drafting/editing costs but human writers still need to coordinate, negotiate voice and vision, so overall cost savings are modest compared to full human collaboration. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can help with task management, document sharing, and draft suggestions, no deployed product reliably performs active creative collaboration at the level writers demand. Products like collaborative writing platforms exist, but they do not automate the decision-making and creative negotiation between writers themselves. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products like Google Docs AI or shared drafting tools with AI assist exist, but no deployed system manages the interpersonal, negotiated aspects of co-authoring reliably. |
Consult with sales, media and marketing representatives to obtain information on product or service and discuss style and length of advertising written material.
24CI 13–35 · exposure 13 · augmentation 63 · click for rater detail
Consult with sales, media and marketing representatives to obtain information on product or service and discuss style and length of advertising written material.
24| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Marketing and creative sectors are experimenting with AI drafting and ideation tools, but actual adoption of AI-driven consultation (replacing the back-and-forth stakeholder engagement phase) remains minimal and pilots are rare. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | While marketing/media sectors adopt AI tools quickly for content generation, the specific consultative meeting function has seen little displacement by AI agents in practice. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at augmenting this task: it can pre-draft briefs, summarize previous campaigns, suggest messaging frameworks, and organize stakeholder input—significantly boosting a writer's productivity in preparation and synthesis while the human leads the actual consultation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help writers prepare talking points, summarize prior briefs, or draft follow-up notes after the consultation, offering moderate assistance around the core human interaction. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can draft meeting notes and summarize product information, but consulting requires negotiating style preferences, understanding nuanced business context, and building collaborative relationships—tasks that demand human judgment and real-time interpersonal adaptation that current systems cannot replicate end-to-end at 50% time savings. |
| Task automatability | claude-sonnet-5 | 1/5 | This task is fundamentally a live, interpersonal consultation to gather requirements and negotiate creative direction, which requires real-time human interaction and judgment AI cannot substitute for end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Consulting roles often have organizational expectations for human presence and trust-building, though no formal legal requirement mandates human involvement. Client preference for direct human contact and organizational norms create moderate friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No formal licensing barrier exists, but organizational and relational friction is high since colleagues expect to consult with a human writer who understands nuance and can build rapport. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | A consultation session with sales/marketing teams involves human facilitation, relationship-building, and strategic discussion that AI cannot fully replace. The overhead of human oversight and iteration outweighs any inference cost savings. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot perform the actual consultation, there is no viable cost comparison—a human must still conduct these meetings, so AI offers no cost substitution here. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate meeting summaries and extract product details from documents, no deployed product reliably conducts live consultations with stakeholders to elicit requirements and reach consensus on messaging. Systems struggle with the dynamic back-and-forth negotiation required. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product conducts autonomous consultations with sales/marketing staff to extract nuanced stylistic requirements; this remains outside current product capability. |
Present drafts and ideas to clients.
18CI 5–30 · exposure 13 · augmentation 75 · importance 4.2/5 · click for rater detail
Present drafts and ideas to clients.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI for client-facing presentations by writers remains negligible. The task involves high-stakes relationship management and human credibility signaling where sectors actively resist automation, favoring direct author-to-client interaction. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Creative and media services are adopting AI for drafting content, but the interpersonal client-facing presentation step sees little automation deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially augment presentation productivity by drafting talking points, organizing ideas, generating visual mockups, and summarizing complex content—allowing the writer to focus on delivery and client dialogue while staying in the loop. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can help writers prepare talking points, generate alternate drafts, summarize feedback, and create supporting materials to strengthen the presentation itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Presenting drafts requires live audience engagement, real-time adaptation, and relationship-building that AI cannot currently perform end-to-end. AI can generate draft content and slide decks, but the interpersonal presentation itself remains fundamentally human; no current system achieves 50% time savings on the full task. |
| Task automatability | claude-sonnet-5 | 1/5 | Presenting drafts and pitching ideas to clients requires live interpersonal interaction, persuasion, and real-time responsiveness that current AI cannot substitute for end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: clients expect direct contact with the human author, professional norms prioritize human judgment in presentations, and liability for idea presentation and client relationship continuity falls on the human. Organizational and reputational friction strongly favor human presenters. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement exists, but strong client preference for direct human relationship-building and trust in creative pitches creates real friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of an AI system capable of autonomous client presentations with appropriate credibility and accountability would likely exceed the loaded wage of a writer conducting the presentation themselves. Integration and oversight costs for replacing the human presenter are prohibitive. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot perform the presentation itself, there is no viable AI cost comparison for full task substitution; a human must still conduct these meetings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in creating presentation materials and summarizing drafts, no deployed product reliably performs the act of presenting to clients with persuasive delivery and responsive engagement. This task requires human presence, credibility, and live interaction in production contexts. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts client presentations of creative work on behalf of a writer today; this remains a human-led interaction. |
Related occupations — Arts, Design, Entertainment, Sports & Media
How to read this
A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.
What would change this score
New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.