GPT‑5.6 arrives in Kiro with better price-performance ↗
GPT‑5.6 is now available directly in Kiro. According to OpenAI it helps developers plan, build, review, and test software, all with better price-performance.
// a daily digest from the AI world
AI news
New models, tools and announcements — every day, always with a source link. This digest is assembled by an AI agent following my rules, and it's the same briefing I use myself to keep up with AI without burning time.
Latest edition · August 25, 2026
Hugging Face reportedly in talks to be acquired for around $13B ↗
Hugging Face is reportedly fielding acquisition offers that value the company at around $13 billion. Its founders feel a strong responsibility to the community, so it remains unclear whether a sale will actually happen.
General Intuition in talks to raise at a $6B valuation ↗
General Intuition is building a foundation model that teaches generalized AI agents how to move through space and time. According to TechCrunch it is in talks to raise at a $6 billion valuation from investors including Valor Ventures, Point72 Ventures, and Seven Seven Six, as it pushes into robotics.
Instinct’s powerful AI assistant raises privacy and security concerns ↗
Testers are impressed by what the Instinct AI assistant can do. At the same time, some warn that its sweeping access, broad terms, and ability to act on a user’s behalf bring uncomfortable trade-offs for privacy and security.
Hugging Face shows how to build AI workflows in Gradio ↗
Hugging Face has published a blog guide on how to wire up, run, and deploy AI workflows in Gradio. It is aimed at makers who want to take their AI solutions from connecting components all the way to deployment.
Previous editions
- August 24, 2026 — “Ox Alpha”: a mysterious stealth model sparks a guessing game
- August 23, 2026 — British lab Inherent unveils Faraday, an agent for replicating research
- August 22, 2026 — Google Discover gets an AI chatbot-tuned feed
- August 21, 2026 — Liquid AI unveils LFM2.5-DSpark with up to 3.2x faster inference
- August 20, 2026 — Replit launches Free Mode powered by GPT-5.6 Luna
- August 19, 2026 — OpenAI paces model development amid cyber-critical capabilities
- August 18, 2026 — Tiny Qwen 3.8 27B matches much larger models on the intelligence index
- August 17, 2026 — Qwen 3.8 27B: a new open vision-capable model
- August 16, 2026 — Meta releases open-weight model Glimmer that anyone can run
- August 15, 2026 — Meta releases Glimmer, an open-weight model you can run yourself
- August 14, 2026 — OpenAI previews Ultrafast: GPT-5.6 Sol up to 14× faster
- August 13, 2026 — DeepSeek V4 Pro 0813 is out, for now via API only
- August 12, 2026 — Meta returns to open weights with Muse Glimmer
- August 11, 2026 — Meta launches Muse Glimmer, an open, local and multimodal model
- August 10, 2026 — Claude Code turns auto mode on by default
- August 9, 2026 — OpenAI slows Astra model development over security concerns
- August 8, 2026 — OpenAI slows Astra model development over cyber concerns
- August 7, 2026 — OpenAI improves GPT‑5.6 Sol and opens GPT‑5.6 Luna to free users
- August 6, 2026 — Meta launches Muse Code, an AI agent for large codebases
- August 5, 2026 — New Claude models opus-5, sonnet-5 and fable-5 land in LLM
- August 4, 2026 — Alibaba releases Qwen3.8-Max, its largest and open-weight model yet
- August 3, 2026 — OpenAI and Anthropic models push real math and CS research
- August 2, 2026 — OpenAI reports ten advances in mathematics and theoretical computer science
- August 1, 2026 — DeepSeek V4 Flash: a new model built for agents
- July 31, 2026 — OpenAI cuts GPT-5.6 pricing for Luna and Terra
- July 30, 2026 — OpenAI unveils GPT-5.6: frontier intelligence meets frontier efficiency
- July 29, 2026 — Microsoft unveils its first cybersecurity AI model
- July 28, 2026 — Moonshot AI releases Kimi K3 weights, a 2.8 trillion parameter model
- July 27, 2026 — Hugging Face CEO calls for transparency after OpenAI hack
- July 26, 2026 — Anthropic launches Opus 5
I build exactly this kind of automation
This section writes, assembles and publishes itself every morning — from source selection to deployment. If you'd like a quiet helper like this in your company, let's start with an initial consultation — the fee is credited toward the project if we work together.