Tag
The author asks practitioners how effective human approval gates on agent actions really are — what fraction of approvals are genuine catches versus rubber stamps, how trust and fatigue evolve over time, and whether approval queues expose supporting evidence. The post highlights that an unread approval gate may just add latency rather than safety.
The article discusses an incident where hikers were misled by Gemini's advice, contrasting it with OpenAI's GPT-6 Astra launch, and highlights the growing challenge of trust calibration as AI capabilities improve.
This research article reviews findings on structuring harnesses for LLM agents, emphasizing that deterministic guardrails, effective multi-agent hierarchies, and proper memory architecture are crucial for performance and safety, while highlighting gaps in current evaluations and certifications.
EduPanel is a rubric-grounded, learner-conditioned LLM judge that uses three specialized agents to evaluate teaching video quality, achieving reliability comparable to human experts and improving scoring accuracy while maintaining detectability of unreliable outputs.
This review synthesizes computational and interactional mechanisms of affective dynamics in human-AI agent collaboration, proposing that affective cues function as a coordination layer for trust, delegation, and oversight rather than internal emotions.
The paper proposes the Minimum Sufficient Oversight Principle (MSO) for governing delegated AI systems, deriving mathematical solutions for autonomy allocation and trust calibration, and introduces concepts like water-filling allocation and masking pathology.
This paper formalizes trust calibration for agentic tool use as a preference learning problem, using Gaussian processes and Bayesian optimization to decide when an AI agent's actions should be autonomous or require human approval.
This paper investigates the use of LLMs to generate multimodal behaviors (verbal, vocal, gestural, facial) for trust calibration in socially interactive agents. The study finds that while LLMs can produce coherent behaviors aligned with intended trustworthiness traits, they also reproduce societal gender stereotypes.