To be featured in the OpenClaw section of Jensen Huang’s keynote at NVIDIA GTC was a surreal moment. What we’ve experienced afterwards has been even more surreal. Yesterday, Jensen called OpenClaw the “operating system” for personal AI agents and said organisations will need an OpenClaw strategy. And the coolest part: OpenClaw was built by Peter Steinberger, as open source. In that OpenClaw section, frontira and our Lobster Lager project were featured. Lobster Lager is our playful but very real proof point of agentic AI in the physical world, with human approval gates at the critical steps. If Lobster Lager can make it to GTC, you can build your thing too. youtube.com/live/jIviHI7fqyc…

Sometimes you need to keep your agents running when you close the lid of your mac. I build Agent Awake for this github.com/eranium21/agent-a…

Hello

LinkedIn is the real MoltBook - prove me wrong. 90% soulless AI slop by people mistaking outsourcing their PR to Chatbots for great leadership. I have zero respect for this, and it makes me question how they solve other tasks of their work...

If you are in Vienna this weekend, you should swing by luma.com/9dmqcdn0

codex is getting really good

anyone have a spare desk in their startup office in manhattan i can crash at for a month? would love to chill & work with some great peeps. obviously will provide anything you need in return. :)

Will do another test with the @_austrian free Flynet and how much work you can get down with codex and cmux from above the clouds

Current tech stack: Agentic Engineering: cmux claude code / codex Agentic Business Engineering: codex obsidian various APIs

Built my first mac menu bar app today, as always it was way easier than expected. should have done that earlier.

The Fable Moment: Europe’s AI Sovereignty Debate Is Too Slow open.substack.com/pub/eraniu…

I'm still homeless in SF. If anyone's renting or selling a nice place, slide into my DMs. The hotel room is getting old.

LinkedIn was already slop. All that's changed is that it's now AI slop.

best productivity hack of 2026

Europe 2031 just dropped. A five-year scenario of our continent's slide into AI irrelevance, by Judith Dada & the people who advised European leaders. It’s pretty dystopian. But it doesn't have to end where it ends. A few weeks ago I sat down with Judith in front of the Acropolis, birthplace of Western civilization, to find out what it takes to make Europe into a global AI powerhouse. Judith Dada is one of the sharpest voices on Europe, AI, and what needs to happen next. She is a GP at Visionaries Club, runs Relativity Collective, writes one of the best blogs on AI and Europe. The numbers she lays out are brutal. Europe has 5% of global compute. The US has 80%. Europe is 25% of global GDP. That math ain’t mathing. So we talked about the hard trade-offs: - Why Europe shouldn't try to rebuild every layer of the stack from day one, but instead own a few choke points where the world can't circumvent us, and use that leverage to build the rest over time. - Why we need the second-best AI labs here even if they never become number one, because those teams spawn the next generation of companies that do AI for medicine, energy, research. - And why the real problem isn't regulation or paper trails. It's that not enough people in Europe have had their wake-up moment yet. Her bet on fixing it: get the best young people across Europe to know each other from the youngest age. Exposure to ambition breeds ambition. It's the single most powerful force anyone can experience. Full conversation in the video. Filmed where it all started. 🦾🇪🇺

Integrating Plaud, voice messages and screenshots into Second Brain is a game changer

How good is AI? Never heard of of Microsoft having a problem with rolling out Co-Pilot

You're not supposed to prompt an LLM. You're supposed to build a system that prompts itself.

Turns out getting from San Francisco to Lake Charles is not that easy