AI Briefing
GeekNews news
The latest AI and developer news from GeekNews, with short summaries.
Visit website →Google Antigravity Terms Specify Possible Account Suspension for Third-Party Tool Use
The Google Antigravity terms of service include a clause allowing account suspension for using third-party tools, requiring developer caution.
IFM Releases 'K2 Horizon', a Suite of 6 AI Models Exposing the Entire Training Process
IFM has open-sourced K2 Horizon, a suite of 6 models that exposes the entire process from pre-training to agent post-processing.
Cerebras Serves Qwen 3.8 27B Model at 1,500 Tokens per Second
Cerebras has started serving the Qwen 3.8 27B model on its public endpoint at a speed of approximately 1,500 tokens per second.
Claude Fable 5.1 Agent Swarm Generates 3D Worlds Based on Real Terrain in San Francisco and Kyoto
The Claude Fable 5.1 agent swarm unveiled a project that generates precise 3D worlds of San Francisco and Kyoto using only open data.
3.85x Inference Speedup via Diffusion Draft Generation and AR Verification Based on GLM-OCR
A technique combining Diffusion draft generation and AR verification with the GLM-OCR model has been released, increasing decoding speed by 3.85x.
Pickzg(zvec-grep) Open-Sourced: Hybrid Local Search Infrastructure for AI Agents
zg, a hybrid local search infrastructure designed to enhance code and document search efficiency for AI agents, has been open-sourced.
Anthropic Releases Open-Source Blueprint for Building Commerce Agents
Anthropic has released implementation code for shopping and merchant agents designed for retail and travel platforms.
AWS Labs Releases 'CAO', a Multi-Agent Tool for Orchestrating Multiple AI Coding CLIs in Parallel
AWS Labs has open-sourced CAO, a multi-agent orchestrator that enables parallel or sequential control of various AI coding CLIs, including Claude Code and Codex.
Uber Reveals Strategy for Operating AI Agent-Based 'Software Factory'... Cutting Costs by 34%
Uber has disclosed a four-tier hierarchy and metrics to enhance the cost efficiency of AI agent-based software development lifecycles.
TrueForge Releases Open-Source Harness for Running LLMs as Real Agents
TrueForge has released an open-source agent execution environment that integrates LLM tool calling, sandboxing, and session management.
GPU World: SF Contest Imagines a Future Where Everyone Owns a GPU by 2040
A $100,000 prize SF/non-fiction contest imagines a future in 2040 where all humans use LLMs for free.
Codex Pro 20X Provides Exactly 20x Weekly Usage… No 5-Hour Limit Unlike Claude
Codex Pro 20X offers exactly 20 times the weekly usage compared to Plus, and unlike Claude, it does not apply a 5-hour limit.
The 'New Golden Age' of AI Chip Architecture: The Evolution of Domain-Specific Hardware like GPU, TPU, and LPU, and Strategies to Solve the Memory Wall
Following the end of Moore's Law, the surge in AI computing demand has led to an explosive increase in domain-specific architectures such as GPU, TPU, and LPU, with various technical approaches proposed to resolve memory bottlenecks.
OpenAI Releases ChatGPT Work... Supporting Internet-Accessible Code Execution and Headless Browsers
OpenAI has released ChatGPT Work for paid subscribers, introducing internet-accessible code execution and headless browser capabilities in a cloud environment.
Booking.com Reveals Why It Chose Weaviate Over OpenSearch as Its Vector DB
Booking.com disclosed the technical rationale behind migrating from OpenSearch to Weaviate for its RAG and semantic search infrastructure.
NVIDIA Releases 'NOOA', a Python Class-Based AI Agent Framework
NVIDIA Labs has released NOOA, a model-agnostic AI agent framework that integrates prompts and tools into Python classes.
Local Memory Service 'memnest' Released to Prevent AI Coding Agents from Losing Session Context
The open-source service memnest has been released, enabling local permanent storage and retrieval of conversation history and rules by integrating with MCP clients such as Claude Code and Codex.
Debian Holds General Resolution Vote on Allowing LLM Contributions
The Debian project is holding a General Resolution vote in August 2026 to decide whether to allow code and documentation contributions made using LLMs.
Linux Kernel Repository Overloaded by AI Crawlers' HTML Rendering Load
AI crawlers for training are depleting server resources by massively requesting HTML rendering on git.kernel.org.
Overcoming LLM Agent Memory Limitations: Datalog-Based 'Lemmalog' Released
A memory system named 'Lemmalog', which applies the logic programming language Datalog to address memory loss issues during LLM vulnerability analysis, has been released.