Hands-on, open-minded. Software Dev & Product Engineer (since 1986), Entrepreneur by Heart, CPTO. Creator of #Gravity—the Social Net App for Nokia (1M users)
The new open source LLM from @OpenAI feels unreal, insanely fast for a locally running 20B model 🚀 and quality looks good, too. First response: Horizon Beta @ OpenRouter (#GPT5?), second: local #gptoss20b 🥰
One of the most amazing things of current coding agents:
You're working with whatever open source library and run into some issue. The agent will happily dive into the source code and quite often reveal a bug and even provide a "good-enough" fix or work-around ... 🥳
Still using my one-year-old VS Code plugin to auto-generate commit messages (configured to openai/gpt-4.1-mini 😂)
Codex/Claude/GLM-5.1 is creating some code, I'm selecting the code blocks or files I feel comfortable with and then use the commit message as a sanity check 🥇
The worst thing about this is that any measure taken in the EU could even turn out net negative if it leads to direct or indirect transfer into countries w/ far less regulation on [* fill-in] (* CO2 emissions, environmental protections, labor standards, compliance burdens, ...)
The EU is 6% of global emissions. Committing economic harakiri in response to climate concerns is pure vanity. Europe can't save the planet because the continent doesn't have any meaningful influence on the outcome.
It's time for interfaces to break out of flat land.
Lately, I have been experimenting with spatial experiences supported by foundational gestures.
Truly feels magical.
🚨 How the TanStack npm attack actually happened:
1. Attacker opened a normal-looking pull request (#7378) on the TanStack repo.
2. GitHub automatically ran CI tests on that PR.
3. Code inside the PR stole the workflow's GitHub Actions Cache write token during the test run.
4. The attacker used that token to plant poisoned files in the shared build cache. The PR could be closed afterwards. The poisoned cache stays.
5. The official release workflow later pulled from the cache, baked the malicious files into the build, and signed and published 84 malicious package versions to npm.
Today is a hard day. I shared this note with the @linear team today: We’ve made the difficult decision to increase our workforce. This is not a cost-cutting exercise or a reflection of anyone’s performance. We’re simply reimagining every role for the agentic AI era. We’re hiring. We’re sorry about that.
I've been auto-generating "memories" for each interaction with my coding agent for a while now. Just some summary, goal, resources involved, etc. ... amazing how easy it is to search for stuff or just let the LLM tell me what's "going on" 🥳
European tech's biggest builders —Airbus, ASML, Siemens, SAP, Ericsson, Nokia, Mistral — are now openly URGING Brussels to get off their backs. They are totally correct: the current situation is a “CRISIS”! The Draghi report said it. The CEOs are saying it. Politicians keep dressing it up. The answer is simple: deregulate. Cut taxes. LET BUILDERS BUILD!
https://t.co/T4uLK5KYF3
Some people use the Force… 🤨 I prefer a well-timed toast and a five-year mission. 🥂
May the Fourth be with you… if you must but let’s be honest; I’ve been boldly going since before it was cool. 😉🚀
Europe AI strategy should be to specialize on AI inference and improvement of large open weight models, while we try to recover the GPU / companies gap to have a viable internal path. A large Chinese open weight model that works is only better than an European-trained weak one.
Don't just reset Codex rate limits for fun, it costs money.
Don't just reset Codex rate limits for fun, it costs money.
... but the vibes are good ...
I have reset Codex rate limits for ALL paid plans to celebrate a good week and allow everyone to build more with GPT-5.5. Enjoy