latest
4 Claude Code token-saving tricks nobody talks about
Before blaming Claude Code's limits, try fixing the way you're using it.
Claude now lets you manually trigger your own weekly reset, but you only get one
It also won't last forever.
My local LLM reconstructed deleted project documentation from just the code, and it was nearly perfect
Making the most of local processing to automatically document projects.
I gave my 7-year-old laptop a second life with local AI, and it's surprisingly useful
It's now the most interesting machine in my setup
I stopped paying for these AI features once I realized my laptop could run them locally
Why pay for what my laptop already does?
Google’s new experimental Gemini Intelligence feature makes the phone calls you’d rather avoid
"Call the Pizza place and ask them to add extra pepperoni"
Gemma 4 runs fine on my laptop, but putting it on my NAS is what changed how I use it
Running Gemma 4 on my NAS taught me to stop expecting instant AI responses
I tested Unsloth's new desktop app, and now I don't want to go back to LM Studio, Ollama, or AnythingLLM
The team behind the "UD" quants now has a new home for them
This Anthropic course's prompt formula fixed the one mistake I was making with every AI tool
Now I know how to get proper responses from AI with a structured prompt.
I ditched VS Code's Copilot for a local LLM, and I'm getting more done offline
Happy to sacrifice performance for offline use.
OpenAI returns fire against Anthropic's Opus 5.5 with better, cheaper GPT-6 models
It comes hot on the heels of Opus 5.5.
Anthropic releases Claude Opus 5.5, and it's not just more powerful: it's cheaper, too
The best of both worlds.
I put my local LLM on Tailscale, and now I can use it from the other side of the world
Tailscale makes accessing my local LLM easy no matter where I am.
If I had to start using Claude Code again, here's the one thing I would do differently
It's basically a shortcut that I took way too long to bother with
My local LLM's real problem wasn't hallucination, it was a two-year blind spot
I also learned an important lesson about the modern web along the way
Disabling Claude Code's memories is the best upgrade I ever made to my AI coding setup
Auto memory does more harm than good.
My mini PC can fit larger models than my old gaming card, but it still loses on one key metric
With MoE models, VRAM is no longer a hard cap, but that doesn't make mini PCs the better buy
I replaced every AI tool in my workflow with a dumb alternative for a week, and I feel smarter for it
Dumb tools are making me a smarter worker
Claude Code wastes tokens on tasks that don't need them, but this one setting fixes it
Some jobs really aren't that deep
Claude Code's output got dramatically better when I stopped letting it decide when it was done
Claude needed a much shorter leash
