kt cloud is pursuing an internal storage strategy using Ceph to reduce costs and secure technical control.
This piece details the physical and logical path a prompt takes through the various components of a data center during AI inference.
According to Vercel data, Open-Weight models accounted for 29% of AI Gateway token usage while showing high efficiency.
In response to a surge in demand, OpenAI has temporarily lifted the 5-hour usage limit for GPT-5.6 Sol and reset usage counts.
AI code editor Cursor is developing a general-purpose AI agent called 'Sand' that goes beyond coding to handle office tasks.
OpenAI's Fidji Simo is stepping back due to health issues, leaving President Greg Brockman to take over product and business operations across the board.
OpenWiki Brains is a proactive memory framework that autonomously gathers information from various sources to provide AI agents with persistent context.
Claude Code for Desktop now supports an in-app browser feature for checking documents and designs.
Due to computing resource constraints, Anthropic has extended the availability period of Claude Fable 5's paid plan to July 19.
Vercel's Eve project has added a feature that lets you view subagent activity and execution details.
You can now add, reorder, and inspect Vercel Flags targeting rules directly from the terminal via the Vercel CLI.
Vercel has launched a Deployment Policies feature that restricts deployment creation to specific mechanisms, organizations, and repositories.
Vercel has reduced the cost of Web Analytics and Speed Insights by about 10% through infrastructure and event processing improvements.
Anthropic quantified the differences in values shown by Claude across models and languages using 4 axes.
llama.cpp has fixed a bug where excessive context checkpoint creation during agent tasks degraded performance.
Anthropic is focusing on talent acquisition by successively hiring key talent including Google DeepMind's Nobel laureate John Jumper.
An open-source kernel has been released that enables efficient training of MiniMax Sparse Attention (MSA) on Hopper and Blackwell GPUs.
Ireland's data center electricity consumption has surged, now accounting for 23% of the country's total electricity use.
Moondream 3.1, a model that applies MoE architecture to boost efficiency and performance, has been released.
A GUI-based coding agent has been released that lets you precisely control and review LLM code modifications through a visual workbench.
A comparative analysis of token usage and cache efficiency between Claude Code and OpenCode found that Claude Code's overhead was far higher.
This covers the latest research trends and limitations in understanding the internal reasoning mechanisms of LLMs.
Anthropic has launched a feature called 'Reflect' that summarizes and displays users' Claude usage patterns and topics.
The proliferation of AI coding tools is increasing the maintenance burden on open source projects and threatening the ecosystem.
AI Briefing