America has lost 200k of jobs in media, film and the arts. How much is AI?

Who Should Read

A product or service developer who wants to gauge, using data, how AI adoption affects content and creative professions

Core Mechanics

• Creative jobs down 200k in four years
  • The US lost more than 200k jobs in creative industries over four years.
  • About 50k of those were lost in the last year alone.
  • Only the 2001 and 2008 recessions saw similar intensity and duration.
• Movies and TV took the biggest hit
  • Movie and TV lost over 100k jobs, nearly one-third of the sector.
  • Nearly half of all media job losses are in movie and sound recording.
  • Written publishing lost over 70k, the second-largest share.
• AI's share is hard to prove
  • The author says media firms hide AI use to protect their copyright basis.
  • Hollywood consolidation, offshoring and social media creators also play a role.
  • The clearest hint cited is that all digital arts subsectors are losing jobs while in-person entertainment grows.
• Music had live shows; video has no fallback
  • Music shifted to concerts and merch as recorded revenue fell.
  • The author says video has essentially no equivalent live fallback.
  • Movie ticket sales are down 25% from their 2019 peak and cable revenue is down 18% from its peak.
• The worst case is the newspaper path
  • Newspaper payrolls fell about 85% from roughly half a million jobs.
  • Magazines cut 66% of staff since the 90s and book publishers cut 40%.
  • Copyright dilution and forced unbundling are the worst case for media in the AI era.
• Hard to predict, may spread to other fields
  • About 32% of arts, entertainment and recreation workers use generative AI, versus a 62% average across all industries.
  • Book sales rose 70% over six years thanks to TikTok (BookTok).
  • Tech went from nearly 300k job gains in 2022 to yearly losses since.

Community Reactions

• Opinions that AI is not the main cause
  • Some commenters said interest rates and inflation matter far more than AI.
  • Another said ad budgets were already shrinking media jobs, so they want to see a trend break.
  • Others blamed content oversupply, cheaper overseas talent and remote work.
• Experiences of direct AI impact in commercial art and captioning
  • One commenter said ad illustration is almost certainly automated now and the decline lines up with Stable Diffusion.
  • A captioning professional said half the hit came from AI tools and half from clients lowering quality standards.
  • Another said SFX and animation job counts held steady but bids dropped from $500k to $50k.
• Pushback on data interpretation
  • A commenter said the movie chart shows employment only slightly below post-2008 levels, possibly a post-COVID hiring correction.
  • Others said the article ignores growth in influencer and creator earnings.
  • One said the music analogy is forced since demand did not fall, so job losses may be productivity gains.
• Opinion that streaming economics are the issue
  • One commenter said filmmakers get only a sliver of a $12/month subscription, so blockbuster returns are gone.
  • Another said media is oversupplied and so is worth less.

How to Apply

• Write a hypothesis on creative roles in your product
  • Spend 30 minutes today counting how many users are in content, design or captioning roles.
  • Make a table splitting tasks AI replaces directly from tasks it only assists.
• Add confounder checks to AI-impact claims
  • When reporting customer results, list other causes such as rates, downturns and outsourcing shifts.
  • Judge AI's contribution by whether the trend broke before and after adoption.
• Set a policy for disclosing AI use
  • Add an AI-use flag to features that handle generated content.
  • Keep generation history logs for customers who need to prove copyright.

Try It

# Paste into Claude or ChatGPT as is (simplified version of the article's analysis approach)

Our service is {{service description}}, and our main users are {{occupation}}.
Below is recent data on employment and work changes for this occupation: {{data}}

Answer in this order.
1. List candidate causes of this change: AI, interest rates/economy, industry structure, offshoring, demand change.
2. For each cause, find evidence for and against using only the data above.
3. Mark anything the data cannot confirm as 'cannot confirm'.

Next step: rerun the same prompt with data for another occupation and compare the results.

Terminology

Related Papers

• A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis

  A fine-tuned Inception v3 beat GPT-6 Astra and Claude Fable 5.1 on 200 skin lesions

• No signs of AI payoff outside the tech sector

  Revenue per worker and margins rise at the Mag 7 but stay flat across the S&P 493

• LLMs are making formal software verification feasible at a much larger scale

  Claude proved no-crash properties for 123 open-source libraries and found bugs along the way.

• PHP2Shell: AI uncovers critical RCE in PHP

  Wiz used AI to turn a decade-old PHP core flaw into a working RCE exploit

• Using LLM for confidence scoring is a bad idea

  An LLM's self-reported confidence is plausible text, not a measurement; build scores from outside rules.

• The Judge Can Debias Itself: Awareness and Self-Debiasing for Reliable LLM-as-a-Judge

  Warning an LLM judge about bias or letting it self-critique improved 20 of 21 model–bias combinations.

• 90% of US working hours are spent on tasks not covered well by AI benchmarks

  Mapping 29 benchmarks to job tasks shows 90% of US working hours lack good coverage

• Postgres 19 delayed after AI reviewers find more bugs than humans can fix

  LLM code review surfaced so many bugs that Postgres 19 slipped to November and 8-10 features were reverted

Related Resources

• https://www.apricitas.io/p/ai-and-the-fall-of-the-creative-class
• https://fred.stlouisfed.org/series/RPSGENAIUSAGESHAREWORK