Blog
Notes and news from Santiago Vazquez on full-stack engineering, SvelteKit, serverless architecture, AI, and building products that scale.
-
The Productivity Gap You Can't Feel
Developers using LLMs feel 24% faster but measure 19% slower — why the gap between perceived and real productivity should reshape AI tooling decisions.
-
AI Engineering Just Entered the Cleanup Phase
Cost ceilings, security gates, and ROI scrutiny arrived in the same two weeks — the AI-coding conversation has shifted from adoption to governance.
-
OpenAI's Twelve-Day Governance Sprint Is an IPO Move
OpenAI published six governance documents in twelve days after clearing the Musk verdict — the cadence reads like the front matter of an IPO prospectus.
-
AI's Hardware Story Is Splitting in Two Directions
Two real hardware stories landed the same week — a datacenter memory shortage and capable open models collapsing onto ten-year-old servers.
-
AI's Hardware Story Is Splitting in Two Directions
Two real AI hardware stories landed the same day — a datacenter capacity crunch and capable open models collapsing onto decade-old CPUs. Budget for both.
-
The Expertise Tax Is Going Unpaid
The senior engineer's AI edge was built through years of manual debugging — work AI now makes optional, which threatens to stop producing new seniors.