Unique Research · Insight / Stats Digest · August 27, 2024 · Compiled from Josh Howarth “60+ Stats on AI Replacing Jobs” (CN digest)
Ingest note: ~22 CN charts. All headline stats from CN retained below; attach charts on ingest. Signup CTA omitted.
Experts still disagree on AI’s net labor impact: mass displacement vs. better work lives vs. hundreds of thousands of new roles — while millions have already lost jobs to robots. Whatever the mix, as adoption spreads the workforce effect only grows.
Headline board
- 300 million jobs may disappear due to AI (≈9.1% of global jobs) — Goldman Sachs via BBC
- 44% of companies using/planning AI think it will drive 2024 layoffs — Resume Builder
- By decade-end, 14% of workers forced to switch careers by AI — McKinsey (≈375M globally by 2030)
- 14% of workers already lost jobs to AI (CN headline board)
- Advanced economies: 60% of jobs exposed; low-income countries: 26% — IMF
- 24% of workers fear AI will soon make their job obsolete
- Ages 18–24 fear obsolescence 129% more than 65+
- 15% of U.S. workers would consider an AI boss (vs 8% UK) — Asana
Potential impact
- By 2025: 2M manufacturing workers may be replaced by automation tools (robots + some AI) — MIT / Boston University
- 1/4 of CEOs expect GenAI to cut ≥5% of workforce in 2024 — PwC (4,702 CEOs / 105 countries)
- 75% of CEOs: GenAI will materially change business within 3 years — PwC
- 80% of U.S. labor force may have ≥10% of tasks affected by LLMs — arXiv
- By 2027: >7.5M data-entry jobs gone (largest projected loss) — WEF; admin secretaries #2, accountants #3
- Of AI-using/planning firms: 21% say layoffs “certain” in 2024; 23% less sure — Resume Builder
- Next decade: 47% of U.S. workers at unemployment risk from automation (incl. non-AI) — CEPR
- ≥20 years to automate 50% of global work tasks — McKinsey
Impact already visible
- 23.5% of U.S. companies have replaced employees with ChatGPT — Resume Builder (49% adopted ChatGPT; of those, 48% say it replaced staff)
- May 2023: 3,900 U.S. jobs lost directly to AI (5% of that month’s losses; 7th-largest cause) — SEO.ai
- Since 2000: automation cut 1.7M manufacturing jobs — BuiltIn
- 13.7% of U.S. workers report job loss to robots — Socius (displaced workers overestimate peer rates)
Which jobs are exposed?
- US + Europe: 2/3 of jobs face some automation exposure — Goldman Sachs
- Global: 40% of jobs face AI threat — IMF
- Advanced economies ~60% exposed; of those, ~half benefit / half hurt (pay cuts or job loss) — IMF
- 19% of workers in “most AI-exposed” jobs (top quartile by automatable tasks) — Pew
- 27% of bachelor’s+ workers in most-exposed jobs vs 3% below HS / 12% HS / 19% some college — Pew
- Most-exposed earn $13.3/hr more than least-exposed ($33.3 vs $20) — Pew
- Wide adoption of existing automation could touch half the global economy — 1.2B workers / $14.6T wages — McKinsey
Tasks, not whole jobs
- 34% of business-related tasks already done by machines; 66% still human — WEF
- 25% of all work tasks doable by AI; admin 46% / legal 44% / construction 6% / maintenance 4% — Goldman Sachs
- 19% of U.S. workers may see >half of tasks affected by GPT-4-class systems — OpenAI / arXiv
- In 77% of jobs, highly AI-exposed tasks are important — Pew
Worker anxiety & demographics
- Asian workers most worried; white workers least; Black 32% / Hispanic 35% fear obsolescence — SurveyMonkey
- Income <$50k: 30% worry vs income >$150k: 16% — SurveyMonkey
- Fully remote workers 42% more likely than fully on-site to expect AI disruption (54% vs 38%) — SurveyMonkey
- Advertising & marketing: 51% expect AI to change their work — SurveyMonkey
Adoption
- Large firms ~2× as likely as small to adopt AI — IBM; 60% of large cos adopted AI in 2022
- 74.9% of orgs may adopt AI by 2027 — WEF (#8 on planned tech list after digital platforms / cloud / crypto)
- Firms >1,000 employees: 42% actively deployed AI; 40% explored but not deployed; only 18% never tried — IBM
New jobs & attitudes
- 81% of office workers broadly positive on AI; >4/5 say it lifted performance/experience — SnapLogic
- 51% of office workers: AI improved work–life balance — SnapLogic
- 25.6% of WEF-surveyed orgs expect AI to create jobs (widest disagreement of any change surveyed)
- Largest global net job growth 2023–2027: “AI and ML specialists” (~1M) — WEF; agriculture equipment ops / truck & bus drivers each >2M
- LLM access: tasks +15% faster with no quality drop — arXiv
- UK: only 17% often/always recognize when they are using AI — ONS (men, young, degree holders more confident)
- 26% fear using AI makes them look lazy; 18% feel like a fraud when AI does the work — Asana
- Past 25 years: 1/3 of U.S. jobs created “didn’t exist / barely existed” before — McKinsey
— END —