A new language model called Bonsai 27B has been compressed by quantization techniques to the point where it can run on consumer hardware such as smartphones and laptops with less than 4 GB of memory – without a cloud connection.
More than half of German companies rely on artificial intelligence, yet only 11 percent let their AI agents work independently. Trust in autonomous systems has dropped considerably within a single year.
Two self-hosted AI assistants compared: features, smart home, cost, security, and digital self-determination Let’s imagine Jonas. It’s a Tuesday evening in March, it’s drizzling outside, and Jonas is sitting on his bike at a red light. His hands are cold, his head is full, and suddenly there’s this one thought: the garage door. Did he
AI agents are independently taking on tasks that seemed unthinkable not long ago – from penetration tests to drone control to autonomous purchasing. Yet alongside these breakthroughs, security incidents are piling up …
Artificial intelligence can support security departments through automated monitoring, the consolidation of SOC processes and the continuous assessment of the security situation. Machine learning helps to identify threats faster and to automate repetitive tasks.
Google is extending the Gemini app with the ability to present complex scientific topics as interactive 3D models and simulations directly in the chat window. The function is being rolled out step by step for Pro users.
Algorithms search through millions of molecules, identify target molecules and predict safety risks. Pharmaceutical research is increasingly relying on computer-assisted methods to shorten development times and raise success rates.
In the transition from classic robotic process automation to AI-supported automation, many companies are repeating old mistakes: missing process selection, inadequate data quality and insufficient change management put automation projects at risk.
OpenAI is working on GPT-6 as an autonomous agent for companies. At the same time, Moonshot AI is presenting Kimi K3, an open model that can be run locally and could shift the competitive landscape in the AI sector.
A field-tested step-by-step guide, including the stumbling blocks from a real-world build A single missing line break was all it took. Hermes could no longer read the configuration file and silently fell back to default values. The container kept running, the dashboard kept responding, nothing flashed red only every security rule that had been set
This week reveals a pattern that points beyond individual developments: the speed at which AI systems act autonomously is outstripping our ability to control them. While OpenAI is shifting the boundary between tool and actor with GPT-6 Astra …
With the Agents API, OpenAI provides developers with an interface for autonomous cloud agents. The API is in public beta and makes it possible to build agents that carry out tasks independently, process code, and use subagents.
Artificial intelligence is fundamentally changing the management of data centers. Concepts such as infrastructure as code and digital twins make it possible to automate routine tasks and fix errors proactively, while administrators increasingly become orchestrators of AI-supported checking processes.
Google Research has developed a framework that provides AI agents with a permanent knowledge base in wiki form. Failures and successes are stored there, so that the agents can keep learning.
An open-source server based on the Model Context Protocol is intended to give scientists access to more than 1,100 research data sets in natural language. The tool was presented at a UN summit.
With GPT-6 Astra, OpenAI has presented a model that is to take on complex tasks such as software development, data processing, and project management as an automated engineer. According to OpenAI, use costs less than six US dollars per hour.
It is a Sunday afternoon in July, and Felix is sitting in front of a grey box that hums softly beside his router in the living-room cupboard. The box is a NAS (Network Attached Storage, a small always-on storage server for the home), and on it sit three terabytes of his professional life. Proposals, contracts,
This week shows how the AI industry is shifting from the question of capability to the question of control. OpenAI and Anthropic are racing each other for the best models, yet at the same time both companies are losing …
According to the Fivetran Agentic AI Readiness Index 2026, only 15 percent of companies are actually ready for the productive use of autonomous AI systems, even though almost 60 percent are investing millions in the technology.
A survey by the Boston Consulting Group shows that employees in customer service save considerable time by using AI. Two thirds, however, receive no guidance on how to use the capacity gained, so strategic opportunities go unused.
Despite growing investment in artificial intelligence, many companies fail to reach their goals. The main reasons are a missing data foundation, poor collaboration between IT and business departments, and unrealistic expectations of the technology.
The AI system Worldclaw turns simple text input into complete, editable 3D environments. The technology could considerably speed up development processes in gaming and virtual worlds.
A developer has built a locally run AI system for vehicles that runs on a Raspberry Pi 5 and uses the Qwen model. The solution can read vehicle data, control functions, and also work offline.
Imagine that every night someone walks through your house once, checks whether a window is jammed somewhere, a battery is running low, or a smoke detector has gone silent, and in the morning leaves you a short note: “Everything’s fine” or “Take a look at the sensor in the bathroom.” This guide creates exactly that
This week reveals a pattern that points beyond individual product announcements: AI systems are becoming more independent – and with that, the question of control presents itself anew. While companies fight over infrastructure …
The Munich-based company Celonis is expanding its platform from pure process analysis into a comprehensive process intelligence solution. With digital twins of business processes and AI integration, companies should be able to optimize their value chains holistically.
Zoom has unveiled an AI agent meant to turn meeting results directly into concrete work steps. ZoomMate connects to enterprise systems and automates the implementation of decisions.
An AI add-in makes Claude available inside Word, Excel, and PowerPoint. The solution preserves context across several applications and is meant to reduce media breaks when working on documents.
Researchers have developed an ultra-thin patch that processes body data without external servers. The flexible electronics are said to detect critical conditions within milliseconds.
Moonshot AI has unveiled Kimi K3, an AI model said to keep pace with leading Western models in programming tests while being considerably cheaper. The model, with 2.8 trillion parameters, is to be released as open source.
A Friday Evening, an Automation, and a Small Deception It’s just past ten, the kids are asleep, the dishes are still in the sink, and Miriam really has no energy left for technology. Yet she sits on the sofa with her laptop, because she’s been annoyed by a small thing for weeks: when she walks
AI agents are leaving the experimental space and pushing into production systems – often faster than companies are prepared for. What holds this week together is the tension between technical possibility …
Windows 11 comes with several AI-powered functions, among them local real-time captioning, automated speech translation and the controversial Recall feature for searching past screen content. Many features, however, depend on the hardware.
A new beta version makes it possible to use AI directly in PowerPoint. The add-in is available to all account types and helps with creating and revising slides through natural language commands.
Research identifies four learnable routines – systematic use, creative experimentation, effective workflows, and regular reflection – that separate AI professionals from occasional users.
Accenture is deploying Microsoft 365 Copilot to its entire workforce of 743,000 after pilots with 200,000 users, in one of the largest company-wide AI implementations to date.
Around 85 percent of AI and analytics projects never reach production. The causes are rarely technical: unclear goals, poor data governance, missing user feedback, and overengineering.
There is a particular moment in almost every large corporation that always plays out the same way. Someone from Controlling asks: “Are we actually efficient? How do we compare to our own subsidiaries?” And then comes the reflexive follow-up: “Shouldn’t we bring in a consultancy for that?” Consulting firms specialize in answering exactly this question
Anthropic’s Claude Security analyzes GitHub repositories with Claude Opus 4.7, tracing data flows to find complex vulnerabilities and integrating with Claude Code, Slack, and Jira.
DeepL’s new Voice-to-Voice line delivers real-time speech translation across up to 40 languages for meetings, in-person conversations, and API integration into third-party software.
Executives using AI assistants like ChatGPT Work or Claude Cowork can raise productivity by up to 44 percent – provided they apply prompt frameworks such as RAFT or RACE systematically.
AI photo editors like Luminar Neo automate enhancement, object removal, sky replacement, and portrait retouching – delivering professional results without deep Photoshop knowledge.
Physical AI combines artificial intelligence with robots and machines acting in the real world – across manufacturing, logistics, buildings, and smart cities, in a market forecast to near $500 billion.
It is just past midnight, and Marlene can’t stop staring at her screen. She’s a copywriter — freelance, for twelve years — and she has just done something she would have considered impossible a year ago: she threw a vague idea at a machine, “write me an opening about the history of the Viennese coffeehouse
Google’s Gemini-powered Ask Maps lets users order meals and book hotels without leaving the map app, using natural-language queries and personalized preferences.
AI helps detect tumors earlier, tailor therapies to genetic profiles, and adapt treatment plans during therapy – potentially avoiding or better preparing invasive surgery.
Anthropic’s Claude now works as an Excel add-in, letting users clean data, build charts, and debug formulas through natural-language prompts directly in their workbooks.
As AI agents take over routine coding, data engineers shift toward strategy: defining precise specifications, safeguarding data quality, and bridging technology and business.
Notice: This article was created with AI. What’s this about? Microsoft’s AI assistant Copilot has made its way into PowerPoint and is changing the way users create presentations. The artificial intelligence takes over tasks such as slide design, structuring content, and text optimization. This is meant to help users save time while achieving more professional
A step-by-step guide for an AI agent that is allowed to capture notes and make suggestions, but must never secretly rewrite your own knowledge archive. How the retrieval behind it works, and why a RAG pipeline is more than full-text search, is explained in From Second Brain to Real AI Knowledge Management. Why a second