Archives

How I built a swarm of specialized AI agents that wake up at 6am, check my email, curate my news, and operate as a second brain — all running locally on plain text files.

The best AI agents don't follow a checklist — they internalize your goals. Anthropic's model spec shows us what that looks like at scale.

Why software work is shifting toward orchestration, infrastructure, and domain translation—and what that means for careers and companies.

Anthropic's Sonnet 4.6 closes the gap with Opus on coding and computer use benchmarks while staying at $3/$15M tokens. Here's what the numbers mean for builders.

What makes a codebase easy or hard for AI agents to navigate? Lessons from Peter Steinberger on working with coding agents.

How I turned a generic AI agent into a personalized assistant with email access, security monitoring, and daily art generation.

Creating an autonomous trading system that leverages Claude's reasoning to analyze market news and execute paper trades.

Reflecting on Star Trek TNG's examination of personhood and what it means for contemporary AI development.

How tools have always elevated human capability, and what inference engines mean for our future.

Document chunking has grown far beyond simple fixed-size splitting. Here's a deep dive into the modern toolkit of specialized models, open-source libraries, and enterprise platforms.

How I helped build a tool that lets you query Tableau's semantic layer using natural language and AI, integrating LangChain with Tableau's VizQL Data Service.

Tableau Bridge is now available on Linux Containers, offering significant improvements over the Windows-only deployment model.