Archive - Lighthouse AI

Cutting Anthropic Token Costs and Latency with Prompt Caching

Automatic caching vs. explicit breakpoints, the TTL math by access pattern, and the trap where caching costs more than not caching at all.

Jul 12 • Rafael

June 2026

RAG Is Simpler Than You Think

Six approaches to retrieval-based AI, from minimal to elaborate

Jun 10 • Rafael

May 2026

Vibe Coders Have a Reading Problem

Writing code has never been cheaper. Reading it has never been more expensive. Most of us haven't repriced.

May 9 • Rafael

October 2025

Cosine Similarity is Dead. Long Live Cosine Similarity.

A practical breakdown of when this metric works for embeddings, when it fails, and what to use instead.

Oct 27, 2025 • Rafael

#nojs-banner { position: fixed; bottom: 0; left: 0; padding: 16px 16px 16px 32px; width: 100%; box-sizing: border-box; background: red; color: white; font-family: -apple-system, "Segoe UI", Roboto, Helvetica, Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol"; font-size: 13px; line-height: 13px; } #nojs-banner a { color: inherit; text-decoration: underline; } This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts