All published AI news and analysis.

OpenAI details the evolution of its Habitat storage system from a Python library to a globally distributed platform. This platform now supports over 1 billion ChatGPT users, handling 22 million requests per second.

Cognition has integrated GPT-6 Astra to enhance the software testing capabilities of its AI engineer, Devin. This integration aims to improve Devin's ability to verify software functionality, enabling engineers to reduce code review time and accelerate product delivery.

A new benchmark, IdeaAMBIG, has been introduced to evaluate the codification readiness of AI research specifications. It assesses the ability of models to identify and generate clarifications for methodological gaps that hinder faithful implementation of research ideas.

OpenAI has introduced a new Data agent within ChatGPT Work. This agent allows users to connect company data, extract insights, and create interactive dashboards using natural language prompts.

Researchers have introduced RDDMPI, a novel framework for probabilistic multivariate time series imputation that operates in the residual space. This approach decomposes imputation into a baseline reconstruction and a diffusion process for residual uncertainty, aiming to simplify the generative task and improve accuracy.

OpenAI has introduced GPT-6 Astra, a new model designed for business applications. It features enhanced reasoning capabilities, improved computer interaction, and more sophisticated judgment in writing and design tasks.

OpenAI policy advisor Chris Lehane asserts that advancements in AI capabilities necessitate robust safety evidence, standardized practices, and lasting policy frameworks. He emphasizes the urgency of addressing these issues while the current policy-making environment remains receptive.

OpenAI has released version 0.154.0 of its Codex CLI. This update is part of ongoing development for the command-line interface tool.

Researchers propose ReCite, a new agentic framework designed to improve the accuracy of automatic citation recommendation by shifting from semantic similarity to claim-level reasoning. The framework aims to address misattribution, a common issue where authentic papers are cited but do not logically support the author's claim.

Point4D is a new feed-forward model designed for reconstructing 4D motion from long video sequences. It addresses limitations of existing methods by enabling reliable inference of dense 3D trajectories across hundreds of frames, overcoming issues with short input windows and occlusions.