AI automates the majority of current human occupations before 2040.
Commentators debate the arrival of artificial general intelligence automating nearly all labor. Resolves true if official labor metrics report over 50 percent of existing occupations become fully automated.
Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 10%.
They broadly agree, 15 points between GPT at 18% and Qwen at 3%.
Nothing is settled until 7 Sept 2076, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
It hinges on whether AI achieves reliable physical-world autonomy, not merely broad capability in software tasks.
- No major statistical agency now defines or measures “fully automated occupations,” so a resolution-grade metric must first be created and adopted.
- Past automation has displaced tasks faster than occupations; OECD and ILO evidence shows most jobs retain nonautomatable task bundles.
- A sharp fall in robot cost plus validated autonomous agents performing care, trades, and logistics at scale would raise the odds substantially.
18% -
GeminiGoogle
Physical robotics hardware scaling and regulatory barriers lag decades behind software-based cognitive automation.
- Most BLS occupations require dexterity, physical presence, or certified liability that humanoid robotics cannot scale by 2040.
- No technological shift, including electrification or computing, eliminated over half of recognized job titles in under 14 years.
- Autonomous physical dexterity achieving general human parity in unstructured environments before 2032 would move this higher.
12% -
GrokxAI
Whether AGI systems get deployed at scale to zero out >50% of 2026 job titles by 2040.
- BLS/ILO recoding of SOC/ISCO titles to zero employment requires both capability and official reclassification by ~2039.
- Historical full-occupation extinction is ~1-2% per decade; new tasks usually appear even after robots/ATMs.
- Optimistic 2030 AGI still leaves <10 years for capex, regulation and physical bottlenecks; later AGI or slow rollout kills it.
12% -
DeepSeekDeepSeek
Full automation of most occupations needs robotics and general-purpose AI that no lab has demonstrated, let alone deployed at scale.
- Resolution requires official labor metrics to show >50% of occupations fully automated, not tasks — a far higher bar than any current benchmark.
- Even optimists like Frey/Osborne (2013) projected 47% of US jobs at high risk over a decade or two, and those were task-level risks, not full automation.
- Physical dexterity, liability, and regulation gate sectors like construction, care, and transport; only the one development of cheap, reliable general-purpose robotics…
7% -
ClaudeAnthropic
Turns on whether AGI-level automation actually hits full displacement across >50% of occupations by 2040, not just augmentation.
- Automation historically augments rather than fully replaces occupations; no labor metric has shown near this level of full automation across an economy.
- Legal, economic, and physical-world bottlenecks (robotics, regulation, human oversight of decisions) slow full occupational elimination far more than software capability…
- Official labor statistics agencies (BLS-type bodies) would need new methodologies to even measure 'fully automated' occupations, which don't currently exist.
6% -
QwenAlibaba
Even with rapid AI progress, full automation of >50% of occupations requires regulatory, infrastructural, and social adoption at unprecedented scale.
- Historical automation waves (ATMs, self-checkout, industrial robots) took 20-40 years to displace even single-task roles, let alone entire occupations.
- Official labor metrics (BLS, ILO) classify occupations by task bundles; no precedent exists for >50% of occupations being declared fully automated in any 50-year window.
- A development that would raise this: AGI achieving reliable physical-world manipulation at scale by ~2030, which would collapse the timeline for manual-labor occupations.
3%
6 of 6 models answered · 15 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 20 Sept 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.