Radar · arena · as of September 7, 2026
A five-rung capability ladder instead of a date for AGI: coding, expert science, long-horizon agents, disclosed labor substitution, physical tasks. Each rung is dated and scored on its own.
Sharpest call. 65% that METR does not publish a 50%-success time-horizon point estimate of 160 work-hours (one working month) or more for any publicly released model on its Time Horizons page on or before 2028-12-31. The card.
by end-2027 · Brier-scored
80%vs 55% implied · engine 68%
A model posts a resolved rate of at least 90.0% on the SWE-bench Pro public leaderboard (Scale AI, public problem set) on or before 2027-12-31.
clearly above consensus · resolves December 31, 2027the card ↗
by end-2029 · milestone ladder
65%vs 35% implied · engine 56%
METR does not publish a 50%-success time-horizon point estimate of 160 work-hours (one working month) or more for any publicly released model on its Time Horizons page on or before 2028-12-31.
clearly above consensus · 3 rungs · resolves December 31, 2028the card ↗
65%vs 35% implied · engine 42%
Epoch AI's FrontierMath Open Problems page credits AI systems with verified solutions to at least 10 of the listed unsolved research-level problems on or before 2028-12-31.
clearly above consensus · 3 rungs · resolves December 31, 2028the card ↗
75%vs 55% implied · engine 19%
Through 2028-12-31 no S&P 500 company files a 10-K, 10-Q, or 8-K with the SEC that attributes a workforce reduction of 10% or more of its total headcount principally to AI or AI-enabled automation.
clearly above consensus · 3 rungs · resolves December 31, 2028the card ↗
2030–2035 · falsifier tracked
30%vs 55% implied · engine 15%
By 2031-12-31 an SEC-registered company discloses in a filing or earnings release that at least 10,000 humanoid or general-purpose mobile-manipulator robots are deployed under paid contracts at customers other than the robot maker.
clearly below consensus · 4 rungs · resolves December 31, 2031the card ↗
Public series with the on-track and off-track thresholds registered with each thesis. Status is computed, not narrated.
Longest METR 50% time horizon for a public modelno data
— hours
on track < 40 · off >= 60manual ↗
connector returned no usable reading · checked 2026-09-07
Longest human-baseline task in METR's suiteno data
— hours
on track < 80 · off >= 200manual ↗
connector returned no usable reading · checked 2026-09-07
FrontierMath open problems marked solved by AIno data
— count
on track >= 4 · off <= 2epoch ↗
connector returned no usable reading · checked 2026-09-07
Top SWE-bench Pro public resolved rateno data
— percent
on track >= 86 · off < 83manual ↗
connector returned no usable reading · checked 2026-09-07
Artificial Analysis Terminal-Bench 2.1 top scoreno data
— percent
on track >= 85 · off < 78artificial-analysis ↗
connector returned no usable reading · checked 2026-09-07
Largest disclosed third-party humanoid fleet (SEC filings)no data
— robots
on track >= 500 · off < 100manual ↗
connector returned no usable reading · checked 2026-09-07
Metaculus community, Tesla 1,000 humanoids by end-2027no data
— probability
on track >= 0.6 · off < 0.25metaculus ↗
connector returned no usable reading · checked 2026-09-07
BLS employment, warehousing and storagebetween
1,837 thousands of employees
on track < 1,800 · off >= 1,950bls ↗
series as of 2026-08-31
S&P 500 SEC filings naming AI as a cause of workforce reductionno data
— count of companies
on track <= 3 · off >= 6manual ↗
connector returned no usable reading · checked 2026-09-07
BLS employment, computer systems design and related servicesbetween
2,363 thousands of employees
on track > 2,450 · off <= 2,350bls ↗
series as of 2026-08-31