Tag
SALSA is an open-source human-in-the-loop platform that combines AI and user intervention to extract structured datasets from multimodal scientific literature, focusing on materials research.
An individual discusses building a tool for independent AI review in agentic workflows, which is being used as a gate in various AI-driven processes for merge/release reviews.
SAGE is a schema-guided system that uses large language models to automate grant review by structuring rubrics and linking evidence, with human-in-the-loop validation to improve accuracy.
ERPBench introduces a benchmark for evaluating screenshot-only computer-use agents on a live ERP system, showing that strong general GUI performance does not transfer to enterprise reliability and includes a production harness with human approval for safe deployment.
The article questions the common practice of limiting AI agent runs for human verification and explores structural alternatives when task volumes exceed human oversight capacity.
The article discusses trust boundaries in AI agent systems where an LLM plans an executable DAG of agents, highlighting challenges in validating plans and ensuring safety through human approval and per-agent permissions.
Shane Parrish discusses with Shopify's CEO how AI can inform better decision-making through human-in-the-loop systems and neutral ground truth.
The paper proposes ATAL, a decision assurance framework that assesses the reliability of AI-generated outputs in air traffic management to ensure safety and operational consistency.
The paper presents ShopEase, a generative AI-based multi-agent framework for enterprise customer support using hybrid retrieval-augmented generation. It evaluates retrieval configurations, finding dense retrieval with FAISS achieves the best accuracy.
The author questions whether Git diffs remain adequate for human review as coding agents accelerate, suggesting a shift toward reviewing agent consequences and decisions with enhanced harnesses for better supervision.
The article critiques human-in-the-loop systems for leading to disengaged approval and advocates for scoped authority in AI agent design, emphasizing evidence and enforcement over routine clicks.
This paper introduces a human-in-the-loop, multi-agent GeoAI system for Arctic eco-navigation that integrates operational, ecological, and community criteria to plan safer and more socially responsible maritime routes.
The Muse security architecture provides detailed control over when AI agents require human-in-the-loop approval, enhancing safety and usability in automated systems.
A user experiments with the AI tool Astra in Blender, showing that while a single prompt yields poor results, 200 guided prompts enable photorealistic modeling, highlighting current AI limitations and potential for future improvements.
The author discusses the ethical decision to avoid full AI autonomy in generating content for children, opting for mandatory human review to ensure safety and trustworthiness.
A company shares their pipeline for generating K-5 learning content using AI, with every output vetted by teachers, discussing the balance between automation and human oversight in high-stakes education.
The author shares real-world reasons why AI agent pilots fail in regulated industries, emphasizing integration issues, lack of defined criteria, misplacement of human oversight, and absence of post-launch ownership.
Legora used GPT-6 Astra to review 41 financial documents in minutes, achieving a 40% performance improvement on financial-tie-out tasks while maintaining human oversight.
Argus is a general agentic runtime for autonomous research systems, enabling continuous learning and self-evolution through persistent memory, verification, and a pluggable architecture for deployment.
Dr. Claw is an open-source AI workspace that wraps coding agents in a controllable, auditable human-in-the-loop workflow for research, improving completeness with a traceable process trail.