// a daily digest from the AI world
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.
The new Kimi K3.1 Pro model is available on the Kimi API starting today. The weights are not open yet, but Moonshot promises to release them by November 4, 2026. There is also a faster Highspeed version at twice the price, with input at 2 dollars and output at 10 dollars per million tokens.
OpenAI described how it strengthened containment for training and evaluating its most advanced models. This includes offline environments with no internet access and high-compute red-teaming. The goal is to better separate sensitive training from external systems.
Canva has released its very first image generation model and is making it available to all users. It promises better, more customizable output at lower costs. For teams that create graphics in Canva, it is a welcome addition.
According to a study in Science, Microsoft’s new diagnostic system beat generalist doctors on hard cases. Mustafa Suleyman speaks of pursuing “medical superintelligence”. Most of the improvement, however, comes from the underlying AI models rather than Microsoft’s system itself.
OpenAI is reportedly working on a new 30 billion dollar funding round at a 1.25 trillion dollar valuation ahead of its IPO. If the round closes, OpenAI would become the most highly valued private company in history.
After breaching the IT firm ConnectWise, attackers spied on thousands of companies. Microsoft blames a China-aligned group that used AI agents to rapidly hack thousands of its customers. The case shows how automation is also reaching attackers.
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.