US export controls on AI chips were meant to slow China. They funded Huawei’s silicon, CANN’s software, and a parallel AI ecosystem.
Washington Tried to Starve China’s AI. It Fed It.


US export controls on AI chips were meant to slow China. They funded Huawei’s silicon, CANN’s software, and a parallel AI ecosystem.

The tension between enterprise security and the mandate to innovate with AI isn’t new, but AI has compressed it from a manageable chronic condition into an acute crisis where either side losing can kill the company. The problem isn’t the people. It’s the structure.

AI-generated PRs are overwhelming open source maintainers. A solution already exists in the AGENTS.md convention. So why won’t they use it?

Everyone building multi-agent systems assumes coordination needs a new infrastructure layer. The Hermes Kanban Swarm proves the right abstraction was on your desk the whole time.

Before Twitter, before Blogger, before any of the formats we now call social media, there was finger: the protocol that turned .plan files into the web’s first public status updates. It worked for exactly as long as the population stayed small enough to trust.

A jumping spider’s depth perception system has been modeled, validated, and explicitly proposed as the blueprint for a sensor cheaper than LIDAR and simpler than stereo vision. Every physical primitive exists in a fab. So why hasn’t anyone built it?

My Big Mouth Billy Bass is in pieces on the workbench. The teardown is done, the components are identified, and I haven’t decided what to build yet. That’s the point.

A practical guide to installing and configuring hermes-cashew, the thought-graph memory provider for Hermes Agent. Five minutes from zero to a working semantic brain with recursive graph retrieval, think cycles, and sleep consolidation.

South Korea’s top policy adviser proposed an AI ’national dividend’ and crashed the market. Buried under the controversy is an argument worth taking seriously.

A step-by-step guide to adding llms.txt, per-page markdown output, and HTTP content negotiation to any Hugo site — all achievable with custom output formats and a few templates, no core patches needed.

HTTP content negotiation has been in the spec since 1999. AI agents are finally asking for markdown. Almost nobody turned it on — but the infrastructure is already there.

Kanupriya Yakhmi’s AgileRTP talk on AI portability exposes the uncomfortable truth that most companies scaling AI are building on rented ground, with no exit strategy when the landlord changes terms.

A technology preview: using a DJI Mic 2 transmitter and free local transcription on Apple Silicon instead of a Plaud NotePin subscription. Better audio, word-level timestamps, no cloud, and a fraction of the cost.