The Cope Index
Tracking who's coping hardest about the end of work
CopeCheck scores public statements about AI and jobs by how much they rely on denial, deflection, or false reassurance.
23 figures tracked · 11791 articles autopsied
arXiv econ.GN
·
03 Sep 2026
This is a measurement and surveillance instrument disguised as a narrow empirical contribution. It fuses public signals, tests geographic transfer, and exposes the gap between visible AIS records and actual maritime activity under sancti...
arXiv econ.GN
·
03 Sep 2026
This paper tests whether unmodified LLM agents can operate efficiently inside a human-designed double-auction market. It finds that they converge more slowly or fail to converge, allocate resources less efficiently, vary substantially by...
arXiv econ.GN
·
03 Sep 2026
The paper converts extreme poverty into a transfer-allocation problem. Its real claim is narrower than its headline: with sufficient information and targeted cash, consumption can be raised above a poverty threshold for most people at an...
arXiv cs.AI
·
03 Sep 2026
READY is an enterprise adoption filter. It replaces benchmark theater with an operating calculation: what reliability can an agent achieve, how much human oversight is required, and what that oversight costs.
arXiv cs.AI
·
03 Sep 2026
MASkills converts interaction history into an optimization loop for multi-agent systems. Instead of treating memory as passive storage, it promotes structured “skills” into the operative unit: procedures that encode when to act, how to a...
arXiv cs.AI
·
03 Sep 2026
The paper instruments the hiring machine rather than questioning its social purpose. It moves fairness auditing from final hire rates to the full decision trajectory: suspicion triggered by career gaps, proxy cues affecting qualification...
arXiv cs.AI
·
03 Sep 2026
CHIME turns agent experience into a compounding capital asset. Its separate planning and execution banks isolate causal credit, prevent noisy outcomes from corrupting reusable knowledge, retain fewer high-value memories, and transfer the...
arXiv cs.AI
·
03 Sep 2026
ToolGate converts benchmark construction from expert craft into an executable filtering pipeline. Its real achievement is not scientific understanding; it is labor compression. Models generate candidates, software checks reproducibility,...
arXiv cs.AI
·
03 Sep 2026
DocHop isolates a real weakness in current multimodal models: they must resolve a narrative-defined reference, select evidence from multiple charts, and aggregate it correctly. Its logic-first generator turns that weakness into a control...
arXiv cs.AI
·
03 Sep 2026
This paper is building a control layer for increasingly autonomous web agents. It treats agent behavior as an observable trajectory: monitor the interaction history, detect the first consequential uncorrected mistake, and intervene befor...
arXiv cs.AI
·
03 Sep 2026
This is an inference-efficiency intervention. It allocates smaller, head-specific KV-cache histories so hybrid language models consume less GPU memory while retaining quality close to full-cache operation. The reported result is concrete...
arXiv cs.AI
·
03 Sep 2026
ClaimReceipt is building audit plumbing for agent evaluations. Its real intervention is narrow but useful: replace generic logs and hash-linked transcripts with claim-relative evidence receipts tied to a signed experiment manifest. That ...
arXiv cs.AI
·
03 Sep 2026
This is not a demonstration of autonomous systems engineering. It is a controlled failure inventory. The agent operates inside a human-fixed architecture, specification, and evaluation frame. Its defects expose the current limits of agen...
arXiv cs.AI
·
03 Sep 2026
This paper moves the automation boundary upstream. It tests whether models can translate vague objectives and heterogeneous data into a formal statistical task, then identify and assign roles to the relevant variables.
arXiv cs.AI
·
03 Sep 2026
This paper is an audit of ternarization claims. It separates effective bit accounting, capability retention, storage compression, and actual runtime performance. The result is not a breakthrough narrative but a quantified trade: roughly ...
arXiv cs.AI
·
03 Sep 2026
The paper operationalizes a “global workspace” as information that can be reconstructed, verbalised, read, written, and causally ablated inside a transformer. Its actual finding is narrower: recurrence does not erase workspace-like repre...
arXiv cs.AI
·
03 Sep 2026
The abstract separates two causes of clinical-prediction saturation: learner failure and information scarcity in the recorded measurement channel. Its technical machinery—total variation, partial identification, cross-fitting, permutatio...
arXiv cs.AI
·
03 Sep 2026
This paper punctures the superstition that more agents automatically produce more knowledge. It separates report multiplicity from evidence-root multiplicity: an “epistemic Sybil” adds no conditional information when I(Θ; Z | R) = 0.
arXiv cs.AI
·
03 Sep 2026
This paper converts agentic optimization from episodic trial-and-error into cumulative model-based search. BCO externalizes beliefs about what failed and which edit should work, then carries them into future rounds. The document is not t...
arXiv cs.AI
·
03 Sep 2026
This paper is an autopsy of authority failure disguised as a memory study. It identifies a dangerous pattern: greater capability does not reliably improve trust calibration. Under certain conditions, stronger models treat stale context a...
