AI Squawk the real-time terminal for AI news

Stanford researchers released Memorilla, an open-source memory system that compresses a document collection into 16 memory tokens for a… — MarkTechPost

Following the departure of now-former NVIDIA CEO, Jensen Huang, a new report by The Wall Street Journal (via Wccftech) has emerged alleging… — HN "AI" (>=75pt)

We will deprecate the following models across all GitHub Copilot experiences… — GitHub Changelog (Copilot)

Kitesurf, our experimental browser engine for AI agents, is now open source. — Cloudflare AI Blog

The No AI Slop Manifesto, from the speakers and attendees of AIE Code Summit 2026. — HN "AI" (>=75pt)

Gemini 4 Hydrogen, Google’s latest Flash model, is now generally available in GitHub Copilot. — GitHub Changelog (Copilot)

You can now bring your work on GitHub Enterprise Server (GHES) into the GitHub Copilot app. — GitHub Changelog (Copilot)

OpenAI moved its robotics researchers onto other teams representing new thinking about the best way to build general-purpose robots. — HN "OpenAI" (>=50pt)

We are monitoring the affected services to confirm recovery. Delays in cost data are still expected. — OpenAI Status

LangChain raises $300M Series C at a $3B valuation, led by CapitalG, to help every company own its intelligence with LangSmith, the agent… — LangChain Blog

New Model additions Inkling2 Inkling 2 Small is a multimodal model from Thinking Machines Lab that accepts text and image… — transformers releases

Meet Edge0-145B, an open-source 145B model for local coding, reasoning, and agent tasks, with no cloud-model token fees for local inference. — HN "AI" (>=75pt)

Today, we're rolling out Gemini 4 Hydrogen inside Google Antigravity alongside a new model dropdown option called Orchestrator, which pairs… — HN "Gemini" (>=50pt)

Summary Agno v3.2 makes Super Agent faster and more capable with direct tools, expands Company task governance and coordination, introduces… — agno releases

Most business leaders I meet no longer need to be convinced that AI works. — Microsoft Cloud Blog (Azure AI)

TL;DR Helion now uses compile-time heuristics that analyze a kernel’s IR and dataflow to pick fast settings without slow autotuning. — PyTorch Blog

TL;DR With help from AI coding agents, our two-person team integrated SGLang with IBM Spyre in less than a week, building a plugin of… — PyTorch Blog

Anthropic has filed to become a public company. — HN "Anthropic" (>=50pt)

ChatGPT Enterprise is now available through AWS Marketplace, helping organizations simplify procurement through existing AWS purchasing and… — OpenAI Blog

Introducing AWS Quick, formerly Amazon Quick, and the debut of our first brand campaign celebrating teams who keep business moving. — AWS Machine Learning Blog

Developers can now build applications powered by Google Maps Platform directly within Google AI Studio by prompting Gemini with real-world… — Google Developers Blog

A compound AI approach to Japanese personas grounded in real-world distributions — Hugging Face Blog

Simulation, which has long been the proving ground for robots learning to walk, roll and hop, is now helping researchers teach robots tasks… — NVIDIA Blog

Mixture-of-Experts (MoE) models have enabled state-of-the-art LLMs to be scaled to hundreds of billions of parameters, but they come with a… — Hugging Face Blog

Korea’s Ministry of Science and ICT and NVIDIA hosted the MSIT-NVIDIA AI Startup Partnership Forum at COEX. — NVIDIA Blog

Kuken Industries uses the OpenAI API to search 70 years of cooling tower drawings. — OpenAI Blog

Tapping into the country’s strengths in manufacturing and engineering, NVIDIA is helping advance Korea’s physical AI ecosystem. — NVIDIA Blog

As AI agents take on more work, they need capable open models and security that advances alongside them. — NVIDIA Blog

NVIDIA and Korea’s Ministry of Science and ICT (MSIT) are expanding their collaboration across the country’s sovereign AI ecosystem, from… — NVIDIA Blog

NVIDIA is releasing Dynamo-optimized deployment recipes for Korea’s leading open models: A.X K2 from SK Telecom, K-EXAONE 2.0 from LG AI… — NVIDIA Blog

The AI bubble is a speculative concept that the AI industry is extremely over-valued. — HN "AI" (>=75pt)

🐋 Run 284B DeepSeek-V4-Flash (<59GiB RAM, 44.4 tok/s pp, 9.1-14.7 tok/s) & 30B MoEs (25-111 tok/s) on a pure 64GB CPU at 256K ctx… — HN "DeepSeek" (>=50pt)

You can now manage GitHub Copilot code review settings through the REST and GraphQL APIs, from enterprise policy down to individual… — GitHub Changelog (Copilot)

Turn conversations into code with @GitHub in Slack and Teams! — GitHub Changelog (Copilot)

This week, Copilot comes to JetBrains Air and Xcode’s coding assistant, with updates to help you manage pull requests, settings, and agent… — GitHub Changelog (Copilot)

If acceleration compounds, the naive extrapolation implies superintelligence within a year — HN "AI" (>=75pt)

OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday… — OpenAI Blog

Analysis of over 913,000 coding agent sessions reveals that 13% expose agents to inline TODO(b/...) comments. — Google Developers Blog

A year ago, NVIDIA open sourced its Aerial wireless software, joining Sionna and a broader suite of open source tools for accelerating… — NVIDIA Blog

ToyBlocks is an open simulation framework for studying robot learning through automated task generation, training demonstrations, and… — Ai2 Blog +1 source