AI and hardware
Learning Rust with Claude Code Through Local Hardware Projects
A power strip and a thermal printer gave me a reason to learn Rust. Claude Code kept both projects moving.
KahWee Teng
I build software with AI coding agents.
I use AI agents at work and on projects around my house. I write about what broke, what it cost, and what I kept using.
Start here
Start with these
Web systems
Cloudflare Redirects for Static Blog URL Migrations
I keep rebuilding this blog. One small Worker keeps old links alive while the rest stays static.
AI economics
GLM Is Cheap. Steering It Isn't.
The cheaper model saved money per token and charged it back in latency, attention, and lost cache.
Latest writing
TanStack Table v9 Trades Monoliths for Explicit Feature Trees
Upgrading TanStack Table v8 to v9 cuts rendering time by 28% on 5,000-row tables. The real migration effort lives in TypeScript generic boundaries.
Read article
React Router v8 Learned to Make Upgrades Boring
React Router v8 keeps the framework-mode API familiar and moves v7's future flags into defaults. That restraint matters after an expensive v6-to-v7 migration.
Read article
Claude’s Guardrails End at Claude
Closed-model guardrails remain tied to the provider, even as comparable open weights spread. Public safety needs controls that survive that reality.
Read article
AI-Generated Code Review: What Reviewers Should Still Block
Models make awkward code cheaper to replace. Review should block future risk and leave private ideas of neat to automated conventions.
Read article
The Best Coding Agent Is the One You Can Keep Using
Grok Build produces useful patches, but its weekly limit changes how I use it. The harness and quota now matter more than the model.
Read article
Learning Rust with Claude Code Through Local Hardware Projects
Claude Code lowers the cost of being a Rust beginner. It kept two local-hardware projects moving while I learned the language inside them.
Read article
Cloudflare Redirects for Static Blog URL Migrations
I keep changing this blog. One small Cloudflare Worker lets the URLs survive the changes while the site itself stays static.
Read article
GLM Is Cheap. Steering It Isn't.
GLM costs a fraction of Opus per token. OpenCode's Plan/Build toggle doesn't cut that cost back down — it just moves it.
Read article
I Can't Avoid the Price Gap Anymore
My company merged separate AI tool budgets into one shared pool. That's when Opus versus GLM stopped being an abstract question.
Read article
Thread Claims Interoperability. Your Network Has Three Meshes.
dns-sd exposes what smart home apps hide: Apple, Amazon, and eero each run separate Thread meshes on the same network, and devices on different meshes route through Wi-Fi instead of communicating directly.
Read article
Popular Tags
technology (56)
ai (36)
web-development (35)
performance (15)
software-engineering (15)
tutorial (14)
developer-tool (14)
claude (13)
ai-tool (13)
coding (11)
react (9)
personal (9)
cloudflare (8)
typescript (8)
build-tool (8)
anthropic (7)
openai (7)
corporate-strategy (6)
cursor (6)
mobile-development (5)
google (4)
ai-assistant (4)
browser (4)
swiftui (4)
vite (3)
open-source (3)
web-server (3)
ios (3)
san-francisco (3)
business (3)
layoffs (3)
chatgpt (3)
large-language-model (3)
privacy (3)
pricing (3)
windsurf (3)
news (3)
smart-home (2)
finance (2)
workforce (2)
perplexity (2)
search (2)
web-traffic (2)
user-experience (2)
legal (2)
github-copilot (2)
aws (2)
static-site-generator (1)
thread (1)
matter (1)
networking (1)
home-assistant (1)
wearable (1)
animation (1)
flash (1)
web-history (1)
adobe (1)
tools (1)
javascript (1)
jquery (1)
claude-ai (1)
productivity (1)
artificial-intelligence (1)
cli (1)
meta (1)
product-development (1)
sora (1)
social-media (1)
video-generation (1)
disinformation (1)
liquid-glass (1)
fly-io (1)
copyright (1)
fair-use (1)
trademark (1)
figma (1)
lovable (1)
microsoft (1)
benchmark (1)
webpack (1)
x (1)
cryptocurrency (1)
testing (1)