Post @onusoz · /2026/07/16 · 04:25 AM View on I asked ML Claw to name itself. Of course in German/Alman Lame suggestions: Ulf, Strich, Fritz, Lex I intervened and called it GoePT Goethe rolling in his grave @onusoz · Jul 15, 2026 The EU 🇪🇺 is broken. It is drowning in regulation The reason for that is simple Germany 🇩🇪 is the dominant economy of the EU And German 🇩🇪 is the most rule-ridden language in the world Coincidence? I think not. The language people speak determines their thinking and their fate. The Sapir-Whorf hypothesis holds The solution is simple: Fix German, Save the EU 💪 I will SAVE EUROPE by simplifying German. With the help of AI. Once and for all 🫡 I will use OpenClaw 🦞 running on Hugging Face 🤗 to train a model that simplifies German The agent harness is called ML Claw. It has full access to GPUs, jobs, sandboxes and datasets on Hugging Face: github.com/osolmaz/mlclaw My new German dialect is called Alman: alman.ai I will be live-tweeting my Quixotic adventure as it unfolds, in this thread My goal is to show that you can run autoresearch loops on Hugging Face infra on free private Spaces, while getting access to SOTA open models like GLM 5.2. And even use your own Codex subscription! And it does not have to be ML. You can just use free Hugging Face Spaces for your OpenClaw agents. It's free compute! Bookmark to be able to find this thread later on 👇 Show more Image hidden Image hidden
@onusoz · /2026/07/16 · 04:25 AM View on I asked ML Claw to name itself. Of course in German/Alman Lame suggestions: Ulf, Strich, Fritz, Lex I intervened and called it GoePT Goethe rolling in his grave @onusoz · Jul 15, 2026 The EU 🇪🇺 is broken. It is drowning in regulation The reason for that is simple Germany 🇩🇪 is the dominant economy of the EU And German 🇩🇪 is the most rule-ridden language in the world Coincidence? I think not. The language people speak determines their thinking and their fate. The Sapir-Whorf hypothesis holds The solution is simple: Fix German, Save the EU 💪 I will SAVE EUROPE by simplifying German. With the help of AI. Once and for all 🫡 I will use OpenClaw 🦞 running on Hugging Face 🤗 to train a model that simplifies German The agent harness is called ML Claw. It has full access to GPUs, jobs, sandboxes and datasets on Hugging Face: github.com/osolmaz/mlclaw My new German dialect is called Alman: alman.ai I will be live-tweeting my Quixotic adventure as it unfolds, in this thread My goal is to show that you can run autoresearch loops on Hugging Face infra on free private Spaces, while getting access to SOTA open models like GLM 5.2. And even use your own Codex subscription! And it does not have to be ML. You can just use free Hugging Face Spaces for your OpenClaw agents. It's free compute! Bookmark to be able to find this thread later on 👇 Show more Image hidden Image hidden