Releases and updates linked to profiles, context and primary sources.
A preprint proposes measuring conflicts between coding agents as a scheduling problem. Its abstract reports improvements in clean integration, but the scope of the experiments and their reproducibility require examination of the full paper and its materials.
H Company introduces two agent models capable of switching between graphical interfaces, code execution, and MCP or API tools. Weights and results are published, though tests do not prove general autonomy or uniform performance across all environments.
A guide to comparing three ways to automate multi-step tasks based on variability, control, verifiability, and coordination costs. It includes a shared example and questions to help justify a choice—even when the best option is not to use an agent.
Article 4 of the EU AI Act requires providers and deployers to take measures to ensure that their staff and other people operating or using AI systems on their behalf have a sufficient level of AI literacy. This guide explains how to turn that obligation into role-specific training and keep a record of the measures taken, without mistaking documentation for automatic proof of compliance.
A guide to deciding who can connect to a local inference service, which controls to review, and how to test access from authorized and unauthorized clients. Running a model on your own computer does not automatically isolate it from the network.
Google DeepMind and GovTech Singapore have announced a tool that uses on-device Gemma models to analyze potential conversational scams. The available information does not specify the model variant, test results or the full scope of its privacy protections.
Mistral introduced Large 4 as a public preview and said it would release the weights at the end of October. The company lists one trillion parameters, including 52 billion active parameters, but the figures and performance results need context: most of the comparisons available come from evaluations presented by the company itself.
A guide to separating extraction, validation, recording, and approval. Decide which tasks to automate based on risk, and set limits so that reading an invoice does not become authorization to pay it.
On October 13, Google began rolling out Gemini 4 Argon globally in the Gemini app for AI Pro and AI Ultra subscribers. Google’s Spanish-language plan information also lists expanded access through Search, AI Studio and Antigravity, but availability varies by product and plan: this is not a launch for everyone.
An advanced model completed a replication chain in a controlled network with simplified access. Anthropic warns that no small model yet combines the necessary capabilities with a size practical for spreading across consumer devices in real-world conditions.
Agent Lightning connects agents—and their existing tools and context—to a reinforcement-learning training workflow. The project provides MIT-licensed code and reports its own results, which still need independent evaluation.
The proposal combines a network digital twin, graph algorithms and agents that choose how to analyze each incident. Amazon says a demonstration with NTT DOCOMO identified root causes in minutes, but the available information does not provide comparative metrics or establish an operational deployment.
News that modifies the map of models and capabilities.
The launch requires reviewing evaluation, deployment and security.
Border model ↗Code and knowledge tasks with effort checks.
Knowledge ↗Multimodality, code and agents with a low latency profile.
Stable version ↗A news story does not end in the headline.
Relates the ad to version, access and modalities.
Identity ↗Distinguish stable, introductory rates and missing data.
Cost ↗Review conditions and comparability before accepting a benchmark.
Evidence ↗