Mercor estimates that 90% of American working hours are spent on tasks with little or no coverage from existing AI benchmarks, based on 29 benchmarks mapped to O*NET's 18,797 tasks.
90%
of American working hours spent on tasks with little or no coverage from existing AI benchmarks, Mercor estimates
Mercor estimates that 90% of American working hours are spent on tasks with little or no coverage from existing AI benchmarks, based on 29 benchmarks mapped to O*NET's 18,797 tasks.
Mercor counts a task as covered when at least 5 benchmark prompts map to it; 924 of the 18,797 tasks are. It calls the results provisional.
Source: A benchmark map of the US economy (opens in a new tab), The Benchmark Map; Methodology in brief. Verified October 2026.
Related charts
- From the blog
Benchmarks cover 12 of 89 healthcare practitioner jobs, 0 of 19 healthcare support jobs
Mercor, A benchmark map of the US economy - From the blog
On emergency cases, GPT-5 alone scored 76.6%; inside Microsoft's system, 80.3%
Microsoft AI, Building Medical Superintelligence, October 22, 2026; Microsoft, MIMIC-IV-Ext-SDBench (PhysioNet, version 1.0.0); Nori and others, Sequential diagnosis with language models, Science, October 22, 2026 - From the blog
OpenAI cut GPT-5.6 Luna's output price by 80% in July 2026
OpenAI, API changelog (July 30, August 21 and September 29, 2026); OpenAI, GPT-5.6 Luna model page; Max Woolf, OpenAI and Anthropic cut their LLM prices by up to 80% within a few months (October 25, 2026); Anthropic, Claude Opus 5.5 (September 22, 2026) - From research
The 320% documentation AI growth figure compares plans with current use
Healthcare Dive, sponsored content by Microsoft, March 10, 2025; Kent Applied, Where healthcare AI has proof, and where it does not
Sources for the figures above: Microsoft, MIMIC-IV-Ext-SDBench (PhysioNet, version 1.0.0) (opens in a new tab); OpenAI, API changelog (July 30, August 21 and September 29, 2026) (opens in a new tab); OpenAI, GPT-5.6 Luna model page (opens in a new tab); Max Woolf, OpenAI and Anthropic cut their LLM prices by up to 80% within a few months (October 25, 2026) (opens in a new tab); This healthcare AI use case will grow 320% by 2026, survey finds (opens in a new tab).
Related facts
~13x
Vercel AI Gateway spending in September 2026 against December 2025, while the average price per token fell 47%
Vercel reported that estimated spending through its AI Gateway, at list prices, was about 13 times higher in September 2026 than in December 2025, while the average price per token fell 47%.
Source: AI Gateway Production Index, October 2026.
KentApplied3 months
at most between a prime contractor's checks of SAM.gov for orders that bar a product or source
Where FAR clause 52.204-30 is in a federal contract, the prime contractor must review SAM.gov at least once every three months during the contract for orders under the Federal Acquisition Supply Chain Security Act that bar a product or a source.
Source: FAR 52.204-30, Federal Acquisition Supply Chain Security Act Orders, Prohibition (Dec 2023).
KentApplied42%
of executives planned AI documentation by 2026
In a Healthcare Dive and Microsoft survey of 130 healthcare executives, published in March 2025, 42% planned to use an AI tool for clinical documentation by 2026, against 10% using one at the time.
Source: This healthcare AI use case will grow 320% by 2026, survey finds.
KentApplied
Sources for the figures above: AI Gateway Production Index, October 2026 (opens in a new tab); FAR 52.204-30, Federal Acquisition Supply Chain Security Act Orders, Prohibition (Dec 2023) (opens in a new tab); This healthcare AI use case will grow 320% by 2026, survey finds (opens in a new tab).
Book a 20-minute call
If a number like this one matters to your organization, talk it through with Tyler Kent.
Prefer to skip the call? Get a written proposal
Or email tyler@kentapplied.com.