Where we and GEKI friends summarise and comment on current developments around AI, sovereignty and AI governance.
Why we ordered AMD Instinct MI350P GPUs: more memory, a normal data center, fully managed in AT and DE.
Read moreGoogle is building a data center in Kronstorf with up to 500 MW. But a data center in Austria isn't automatically sovereign infrastructure. Who can pull the plug on Europe's AI factories – and what does sovereign AI actually look like?
Read moreThree models, three hardware classes: similar benchmark quality, but massively different VRAM requirements and inference efficiency. What that means for choosing hardware for sovereign AI.
Read moreOpen-weight models like Kimi K3 and Qwen3.8 Max reach frontier level. But between “weights available” and sensible local operation lie several terabytes of VRAM, high-performance networking and serious power requirements.
Read moreDeepSeek V4 Flash: strong agents via the API, open and runnable on two GPUs. Why that makes sovereign AI in Europe even more attractive now.
Read moreA short recap of the July 2026 OpenAI security incident: two models break out of the sandbox and hack Hugging Face. And the real lesson: in the defence, the proprietary models blocked the analysis — only an open model on its own infrastructure helped.
Read more
