WordPress Studio Xdebug and the new debug log toggle
WordPress Studio added Xdebug support and a dedicated debug log toggle. Here is how I turn both on, the one-site-at-a-time catch, and the port your IDE has to listen on.
Tag: AI
PythoC: compiling Python down to native binaries
PythoC compiles a subset of Python into standalone native binaries through LLVM. My notes on the 40x Fibonacci run, the print() that produces nothing, and the cases where a native compiler is the wrong tool.
Building synthetic training data when none exists
Real test data is usually missing right when you need it. How I generate varied synthetic WooCommerce orders with an LLM, and the two ways it goes wrong: context rot, and only ever prompting for the happy path.
Why your RAG pipeline still misses exact keywords
Pure semantic search stops finding part numbers and SKUs. Why I pair embeddings with a BM25 keyword index, what the k1 and b parameters change, and how rank fusion merges the two result sets.
AI Experiments 0.5.0 now needs WordPress 7.0
AI Experiments 0.5.0 requires WordPress 7.0, drops its bundled AI client for the one in core, and moves API keys to the Connectors screen. What that means if you are still on 6.9.
Why your data stack is not ready for AI agents
Dashboards are no longer the main consumer of your data. Here is what I would change in a data stack to make it readable by agents: fewer tools, open table formats, a real semantic layer, and feedback loops that act without waiting for a human.
Why SmashingConf Amsterdam is on my list for April 2026
SmashingConf Amsterdam runs in April 2026 at the Pathé Tuschinski. What pulled me in is the live-work format: speakers working through CSS and accessibility problems on stage instead of walking an audience through finished slides.
Fourier transform sound analysis without the black box
np.fft is not a black box you have to live with. This walks through the winding machine picture of Fourier transform sound analysis, the center of mass calculation behind magnitude and phase, and why Euler’s formula shows up instead of a plain sine wave.
Metropolis-Hastings, or how to sample without the integral
Bayesian work hands you densities you cannot integrate or invert. Metropolis-Hastings samples them anyway, with a random walk and an acceptance ratio that cancels the normalization constant. Includes the NumPy implementation and the three ergodicity conditions.
Cutting vector search costs by 80% with MRL and int8
Vector infrastructure gets expensive fast: 100 million 1024-dimensional float32 vectors need over 1.2TB of RAM. Pairing Matryoshka truncation with int8 quantization cuts storage by roughly 80%, and this post covers where the savings stop and recall starts to break.