Long-form posts: AI-assisted development, homelab infrastructure, projects, and works in progress.
A light week where the AI-money beat turned loud — DRAM rationing for data centers and a tech-stock sell-off dragging data-centre loans down with it — alongside MIT's gradient-free model editing and Gamers Nexus on tech money in elections.
Local AI keeps shrinking — a 74B MoE on a 24GB GPU and a 763B model claimed on 12GB VRAM — while identical-harness bake-offs tighten verification and Pivot to AI dissects Anthropic's numbers-free IPO and the Chinese AI correction.
AI's money meeting reality — Anthropic's underwhelming quarter, CoreWeave cash-outs, and warehoused GPUs — alongside a fair-use reversal for AI training and a growing verification bottleneck in agentic coding.
The exporter backfilled six years of Practical AI and Pivot to AI archives, while the fresh items — Micron's memory windfall, the Linux kernel's AI-slop problem, and Google's TabFM — replayed the archive's recurring themes of bubble economics and agentic coding.
A tiny Windows app that watches the clipboard and saves every URL it finds - tests what a fully local coding stack can build - and presents the results for your review.
Ed Zitron's Better Offline back catalog rolled through the exporter in bulk, framing the week's fresh AI news — Nvidia's 'cycular' financing, the White House superintelligence accord, and a wobbly AI IPO — while the local-model channel kept score on MoE models and context management for agents.
The AI bubble's finances dominated the week — Anthropic's losses, IPO numbers, and hyperscaler concentration risk — alongside a steady stream of hands-on local-AI tooling: spec-driven coding with local models, heavyweight voice cloning, and Claude Code mods.
How an AI agent took philwheat.net from an invisible Blazor WebAssembly app to a static Hugo site in three days — and why re-implementing your own existing site is a much smaller project than it feels.
A quiet week, joined by two media essays on how attention and value flow: Napster's remix of the music business, and Cory Doctorow's split between real LLMs and fake AI.
I’ve reviewed a lot of packages to find what worked best.
We’ve covered the hardware, the OS options, and the Servers.
The Server Software Stack — What I Run Locally
Well, at least at my home...
I’ve been doing neural networks for more than 30 years.
The start of a series of posts about using LLMs in the development process.
Informational Post
A quick introduction to the site.