How Do Mixture-of-Experts AI Models Work?
Recently, I was reading about Moonshot AI’s new Kimi K3 model when one detail caught my attention: it…
Everyday AI
Can AI Really Fake Emotion? Sora 2 and Kling tested
AI has become remarkably good at creating realistic images, voices and videos. But there is one area where…
SAFE Prompt Framework: Making AI Prompting Effective and Compliant
AI is only as useful as the instructions it receives. In an organisational setting, unclear or sloppy prompts…
How Bioptimus is Building a European Foundation Model for Biology with Open Medical AI
AI Summary: Bioptimus is creating open AI models to transform medical research and diagnosis, starting with pathology and…
AI-Driven Sustainable Tourism Initiative in Coimbra
Coimbra, a UNESCO World Heritage Site in Portugal, is pioneering the integration of AI into sustainable tourism through…
E-Mail and Calendar in Mistral Le Chat with Connections
Mistral’s Le Chat is beginning to show what truly useful AI can look like. One of its more…
European Legal AI Model Noxtua
The legal sector in Europe is beginning to embrace AI with greater focus and care. A significant step…
COMPL-AI: An Open Framework for AI Compliance
AI Summary: COMPL-AI helps test generative AI models for alignment with the EU AI Act using 27 benchmarks…
TerraMind AI Model from ESA for Environmental Monitoring
AI Summary: ESA and IBM have launched TerraMind, an open-source AI model that improves environmental monitoring by integrating…
Tool Update: Mistral with a RAG feature
Mistral AI recently launched a new feature called Libraries, designed to make its chatbot, Le Chat, smarter and…