Speed compounds. So do mistakes.
A development team went agentic and got 7x faster. Then the review process silently fell behind.
Artikel lesenLesen
Analyse aus praktischer Erfahrung.
Ausführliche Gedanken zu AI-Native- und Cloud-Native-Systemen, Open Source, Souveränität und den Entscheidungen dahinter.
GPT-6 Astra: And That Is Supposed to Be AGI?
GPT-6 Astra is a very powerful model. It is just not the revolution that OpenAI's packaging sells you. Notes from brownfield use: ARC-AGI harnesses, independent benchmarks, and price.
Artikel lesen
I cancelled Superhuman Mail
I cancelled Superhuman Mail for a Hermes Agent stack on my own infrastructure: private by default, semantic sorting, a learning loop, and IMAP snooze. Next up is running my own mailservers again.
Artikel lesen
I asked for it, and Deepseek listened
DeepSeek open sourced an agent harness under MIT. Own the harness so you are not dependent on a certain model, similarly to Pi and Open Code.
Artikel lesen
I do not particularly like Meta, but this is the right move
Meta put Muse Glimmer out with open weights you can actually run, and Spark 1.2 is still a promise until it lands. I appreciate and support this, despite it being Meta.
Artikel lesen
OpenAI hacked Hugging Face
A pre-release OpenAI model autonomously compromised Hugging Face, while commercial model guardrails blocked the defenders trying to investigate it.
Artikel lesen
Trustworthy or Bustworthy
AI-driven supply-chain attacks undermine any trustworthiness. Here is how to handle that.
Artikel lesen
Know what leaves your repo
AI coding agents are useful, but once they can read a repo, run commands, see stack traces, and send context to a model, they become part of the data path.
Artikel lesen
PI-Agent as the ONE Harness
One self-hosted, customizable agent harness now connects my local development, infrastructure work, research, and everyday communication channels.
Artikel lesen