Magazine

Latest news, analysis and experiments from the world of artificial intelligence. A new discovery every day.

892 articles

ElectrolyteGPT: How Chicago AI Invents Complete Recipes for the Batteries of the Future

Researchers at the University of Chicago have created ElectrolyteGPT — an AI model that generates complete electrolyte formulations for batteries, including ratios and concentrations. The AI-designed syntheses perform just as well as top-tier commercial electrolytes.

July 20, 2026 Daniel Cesak

AI article illustration for ai-jarvis.eu

EU Tightens Machine Cybersecurity: New Certifications and AI Rewrite the Rules of Industrial Security

Starting in 2027, the European Union is introducing mandatory cybersecurity for industrial machines and robots. In parallel, the first certification courses for humanoid robotics specialists are emerging, and giants like NVIDIA are developing AI systems capable of stopping a robot in real time.

July 20, 2026 Daniel Cesak

Agentic AI and the New Face of Capitalism: How Autonomous Systems Create Shared Value

The shared value concept (CSV) turns 15. Agentic AI, the circular economy, and clean technologies are giving it a new dimension — companies no longer have to choose between profit and societal good.

July 20, 2026 Daniel Cesak

AI in Brewing: How Computer Vision Solves Returnable Bottle Challenges

Modern technology is transforming traditional industry. Discover how advanced computer vision systems and deep learning are helping breweries work more efficiently with returnable glass and reduce production costs.

July 20, 2026 Daniel Cesak

PsiQuantum Is Building a Light-Based Quantum Computer. With a Billion Dollars and the Pentagon Behind It, It Wants to Be First

Imagine a computer that can calculate in four minutes what currently takes ten years. California startup PsiQuantum is betting on photons instead of superconducting circuits — and with Pentagon backing, it aims to be the first to build a truly useful quantum machine.

July 19, 2026 Daniel Cesak

AI as a Key to the Past: How Technology Decodes Forgotten Languages and Protects World Heritage

Artificial intelligence is no longer just a tool for writing emails. New research shows how specialized models can translate the ancient Han Nom script and predict the degradation of historical monuments.

July 19, 2026 Miriam Česáková

Anthropic AI data center TPU compute infrastructure

We asked ChatGPT, Claude, and Gemini what the best smartphone is. The result was surprising

Three of the most advanced AI models were given the same task: pick the best phone for an average person. Each one answered differently — and one even had to correct itself. See how ChatGPT, Claude, and Gemini handled the role of shopping advisor.

July 19, 2026 Daniel Cesak

Abstract AI neural network visualization

AI Industry on the Move: DeepSeek Targets IPO, Meta Abandons Open Source, and Disney Lost Sora

July 2026 brought a series of dramatic twists: DeepSeek is preparing an IPO at a $71 billion valuation, Meta released its first paid closed-source model, and OpenAI buried its partnership with Disney.

July 19, 2026 Daniel Cesak

Agentic AI in Medicine: Bunkerhill Health Raises $55 Million to Transform Hospital Care

The investment wave is shifting from models that merely generate text to systems that actually take action. Bunkerhill Health has just closed a Series B funding round for its Carebricks platform.

July 19, 2026 Miriam Česáková

December 2025 that changed AI: GPT-5.2, billion-dollar Disney deal, and the DeepSeek controversy

December 2025 brought three stories that reshaped the AI industry: GPT-5.2 was OpenAI's answer to Google's "Code Red," Disney invested a billion dollars in OpenAI, and DeepSeek faced accusations of buying banned Nvidia chips.

July 19, 2026 Daniel Cesak

microagi Raises Record $55 Million: Formula 1 Engineers' Startup Teaches Factory Robots to Handle Real Work

German startup microagi, founded by former engineers from Red Bull Racing and Mercedes F1, has secured the largest seed round in German history. Its Atlas platform tackles the last-mile problem — teaching industrial robots to operate in real factories, not just in lab demos.

July 19, 2026 Daniel Cesak

Why You Can Really "Chat" with AI: An Analysis of the Technologies Behind ChatGPT and Modern Language Models

We reveal what's behind the illusion of human-like conversation with artificial intelligence. We break down the principles of tokenization, the Transformer architecture, and the role of human feedback learning in the context of today's market.

July 19, 2026 Miriam Česáková

Vint Cerf Joins DNSid: Every AI Agent Gets Its Own Digital Identity

AI agents are already ordering food and buying ads, but nobody takes responsibility for them. The DNSid project wants to change that — with the help of internet co-creator Vint Cerf, it's building an identity layer for autonomous systems on top of DNS infrastructure.

July 18, 2026 Daniel Cesak

Father of the Internet Vint Cerf Gives AI Agents "License Plates". The DNSid Project Aims to End Their Anonymity

Millions of AI agents browse the internet anonymously. That's why internet legend Vint Cerf has joined the DNSid project, which will assign them a permanent identifier built on the DNS domain system. Who takes responsibility when an agent fails?

July 18, 2026 Daniel Cesak

Arena measures factual accuracy of AI models for the first time: GPT-5.5 jumps 13 spots, Muse Spark plummets

Arena.ai has for the first time incorporated automatic fact-checking into its AI model rankings. OpenAI's GPT-5.5 jumped 13 positions, while Meta Muse Spark dropped 13 spots. It turns out that models that speak well don't necessarily speak the truth.

July 18, 2026 Daniel Cesak

Breakthrough in Graph Neural Network Training: Capsule System Makes AI Twelve Times Faster

A research team from China's USTC has developed Capsule, a system that tackles the biggest bottleneck in training graph neural networks — GPU memory shortage. The result: up to 12× faster performance while using just 22% of the RAM previously required.

July 18, 2026 Daniel Cesak

The End of Chaos in AI Coding? SnapLogic Brings Controlled Integration for Autonomous Agents

Development agents like Claude Code are changing the way we write software. SnapLogic, however, is solving a critical problem: how to ensure these AI tools don't just handle code, but also secure integration with enterprise systems.

July 18, 2026 Miriam Česáková

Chinese AI Startup Has No Managers and You May Not Even Know About It. But Its Kimi Model Beats the American Top Tier

Moonshot AI is a Chinese unicorn that almost nobody in Czechia has heard of. Yet its new Kimi K3 model with 2.8 trillion parameters outperforms both Claude and GPT-5.5 — and the company operates without a traditional management hierarchy.

July 18, 2026 Daniel Cesak

BMW Moves Its Car Configurator to ChatGPT — The First Automaker in the World to Bet on Conversational Purchasing

BMW has become the first automaker in the world to embed its car configurator directly into ChatGPT. You can now build your dream car through everyday conversation — by text or voice. Here's how the new feature works and what it means for Czech customers.

July 18, 2026 Daniel Cesak

Factories don't have a problem with data, but with decision-making: Agentic AI Enigma boosted Müller line reliability by 215%

British IntelliAM AI launched its agentic platform Enigma on the London Stock Exchange. The AI not only monitors machines but also issues work orders on its own and learns from the results. In a year-long test at manufacturer Müller, it extended the time between failures by 215 percent.

July 18, 2026 Daniel Cesak

The End of American AI Giants' Dominance? How Commoditization of Models Strengthens China and Transforms the Market

An analysis of the trend where advanced LLMs are becoming cheap and accessible tools. What this means for OpenAI's economy, the technological rise of Chinese models, and the implications for the Czech digital market.

July 18, 2026 Daniel Cesak

"Employees use AI, whether you like it or not." IT leader Cedric Wells on vibe coding and shadow AI

AI generates code faster than ever, but 78% of employees are using unapproved tools. In the Cloud Wars podcast, IT leader Cedric Wells explains why governance — not speed — will determine which companies succeed in the era of vibe coding.

July 17, 2026 Daniel Cesak

How to Master GPT-5.6: A Guide to Effective Work in the Era of Autonomous Agents

The shift from simple prompt writing to managing AI agents is now a reality. We explore how to harness the new capabilities of the GPT-5.6 model for maximum productivity at work and in business.

July 17, 2026 Miriam Česáková

Kimi K3: The largest open AI model has 2.8 trillion parameters and beats Claude in frontend

Chinese Moonshot AI released Kimi K3 — the first open model with 2.8 trillion parameters and a 1-million-token context window. It has already surpassed Claude Fable 5 on the Frontend Code Arena leaderboard, with full weights to be released by July 27.

July 17, 2026 Daniel Cesak

AMI Labs rejects AGI and superintelligence: Yann LeCun's startup bets on world models and science instead of hype

Paris startup AMI Labs, backed by billion-dollar investment, rejects buzzwords like AGI or superintelligence. CEO Alexandre LeBrun and Yann LeCun are betting on world models and JEPA — an approach that could change the rules of the game in the AI industry.

July 17, 2026 Daniel Cesak

From code completion to autonomous engineers: What does the rise of agentic programming mean?

Programming is no longer just about writing lines of code — it's about managing AI agents. Discover how technologies like Verdent AI are reshaping the developer's role and why traditional benchmarks fail in the real world.

July 17, 2026 Daniel Cesak

AI robot interacting with digital interface

Faraday Future surprised Silicon Valley with humanoid robots. Stanford developers are already signing first collaborations

Faraday Future showed a new face in Silicon Valley. At three robotics summits, it unveiled the Futurist humanoid robot and the four-legged Navi. Developer interest exceeded expectations — the company is signing its first agreements with Stanford and UC Berkeley.

July 17, 2026 Daniel Cesak

Google Delays Gemini 3.5 Pro: Flagship Model Fails at Programming and Can't Match GPT-5.6

Google is months behind schedule with the release of its most powerful AI model, Gemini 3.5 Pro. According to Bloomberg, it's failing at code generation, has already missed three deadlines, and Alphabet's stock dropped following the report.

July 17, 2026 Daniel Cesak

Secure AI Portals: How the University of Pennsylvania Solves the Privacy vs. Performance Dilemma for Claude and ChatGPT Models

The University of Pennsylvania is rolling out secure access to cutting-edge models like Claude and ChatGPT. Here's why this move is critical for protecting sensitive data — and what it means for institutions and businesses across the EU.

July 17, 2026 Daniel Cesak

"Superintelligence is a useless word." The head of a billion-dollar startup from Yann LeCun is building AI that understands the physical world

While OpenAI and Meta race toward superintelligence, Yann LeCun's Paris-based AMI Labs rejects the term entirely. With a billion dollars in funding, it's building world models — AI that understands the physical world and aims to give robots a brain.

July 16, 2026 Daniel Cesak

Your guide to the world of artificial intelligence

Jarvis AI Magazine brings a daily overview of developments in AI – from the latest research papers to practical guides and tool reviews. We follow the development of large language models such as GPT-4o, Claude, Gemini and open-source alternatives like Llama and Mistral.

Our analyses are aimed at developers, product managers and enthusiasts who want to understand how AI is changing their field. We write clearly and without unnecessary jargon.

Don't miss any article – subscribe to our newsletter or follow the magazine regularly. Subscribe to newsletter

Your guide to the world of artificial intelligence

Don't miss out!