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 · 14976 articles autopsied
arXiv econ.GN
·
09 Oct 2026
This paper converts AI governance into a constrained scheduling problem: accumulate enough safety knowledge, reserve compute for safety work, then replay development under a risk ceiling and deadline. Its strongest result is conditional:...
arXiv econ.GN
·
09 Oct 2026
This is a historical identification exercise. By comparing same-gender siblings assigned different Black or Mulatto classifications, the paper tries to separate perceived skin tone from family resources. Its finding is sharply gendered: ...
arXiv cs.AI
·
09 Oct 2026
The paper is building a control layer for persistent-memory agents. It separates admission—what enters working context—from presentation—how that information is expressed—and tests whether deterministic filtering can reduce sycophancy an...
arXiv cs.AI
·
09 Oct 2026
This paper converts adolescent self-injury into a machine-readable classification problem. In 106 depressed adolescents—67 with NSSI and 39 without—it identifies EEG microstate differences during social pain and reports 68.55% discrimina...
arXiv cs.AI
·
09 Oct 2026
This paper builds measurement infrastructure for multimodal game-commentary automation. It standardizes tasks across games, classifies commentary by function, and converts human judgments into comparable evaluation signals. Its stated pu...
arXiv cs.AI
·
09 Oct 2026
This paper is not solving reasoning in the human sense. It is optimizing a controlled production line for verified mathematical trajectories.
arXiv cs.AI
·
09 Oct 2026
This paper is building an interface layer that makes autonomous agents easier for humans to direct, inspect, and correct. GenUI-Harness separates retrieval and execution from interface generation, then adds a sandbox and reward-auditing ...
arXiv cs.AI
·
09 Oct 2026
This paper is building a legitimacy machine for a future claim: AI can enter research territory where humans have not already supplied the answer. The 14.0% mean judged solve rate for GPT-6-Astra is significant as a capability signal, bu...
arXiv cs.AI
·
09 Oct 2026
The paper extracts peer review into machine-evaluable components: manuscript comprehension, contradiction detection, review generation, and score alignment. That is an automation blueprint disguised as an assistance framework. Synthetic ...
arXiv cs.AI
·
09 Oct 2026
This paper identifies the verification layer as the weak joint in agentic labor. Long trajectories can look complete while violating instructions or creating hidden side effects, so the paper turns trust failure into a benchmarkable engi...
arXiv cs.AI
·
09 Oct 2026
The paper converts alignment auditing from direct interrogation into model-forensics. Its central maneuver is to exploit an asymmetry: a teacher may transfer knowledge of hidden behavior to a weaker student more readily than it transfers...
arXiv cs.AI
·
09 Oct 2026
This paper turns persistent agent context from prompt folklore into an operations-research problem. Instructions consume scarce attention capacity, impose recurring session costs, and can interfere with one another. The paper formalizes ...
arXiv cs.AI
·
09 Oct 2026
The paper is an autopsy of a narrow but real control failure: refusal behavior is brittle when harmful intent is wrapped in narrative form, translated across registers, or disguised as role-play. GUISE measures that failure across Englis...
arXiv cs.AI
·
09 Oct 2026
StoreBench is a job description converted into a reinforcement-learning environment. It decomposes online-store operation into tools, scenarios, budgets, rewards, failures, and replayable consequences. The immediate finding is that curre...
arXiv cs.AI
·
09 Oct 2026
This is an instrument-building paper disguised as an alignment diagnosis. It isolates “normative competence”—inferring socially enforced rules from interaction—in synthetic multi-agent debate. Its useful finding is narrow: current LLMs i...
arXiv cs.AI
·
09 Oct 2026
The paper separates two capabilities that are often lazily merged: obeying a deadline and spending additional time intelligently. Its experiments show that agents can be trained or instrumented to stop on schedule, yet still waste surplu...
arXiv cs.AI
·
09 Oct 2026
This is a legitimate methodological correction disguised as a broader human-centered intervention. The paper correctly identifies that aggregating human judgments into one “ground truth” can erase real disagreement and meaningful differe...
arXiv cs.AI
·
09 Oct 2026
This is a containment architecture for making adaptive automation institutionally admissible. It narrows “self-evolution” to changes in instructions, tool logic, and primitive composition while freezing model weights. The admission gate ...
arXiv cs.AI
·
09 Oct 2026
This paper is an engineering intervention against prompt bloat, not a theory of memory. It lets an agent compress selected tool outputs in place while preserving exact originals for recovery. The demonstrated gain is real but narrow: one...
arXiv cs.AI
·
09 Oct 2026
STS is an industrial workaround for the enterprise-data bottleneck. It replaces access to real enterprise systems with simulated environments whose policy-enforcing APIs make valid state transitions available by construction. The General...
