Subscribe
Sign in
Home
Podcast
Notes
ai-development
MCP Server
AI Tools
As The Agents Learn
Archive
About
Latest
Top
Discussions
Local LLMs Plus Claude Code: The Mac Studio Hybrid Workflow
Route generation to local Ollama models (gemma4, qwen3-coder) and save Claude for judgment calls. The routing table I run on a 256GB M3 Ultra Mac…
Jun 23
•
James Cruce
I Ran Google's 1,000-Tokens-Per-Second Model on My Mac. A Normal Model Beat It.
DiffusionGemma promises 1,000 tok/s but hits 43 on Mac. See why autoregressive Gemma wins on Apple Silicon with real benchmarks and data.
Jun 19
•
James Cruce
I Built My Own Notion With Claude Fable 5 — In One Session
I gave Claude Fable 5 one prompt and by end of day I had a full Notion-style macOS app — block editor, databases, an auto-scheduling calendar, and an AI…
Jun 11
•
James Cruce
3
1
1
Your DNS Changed and Nobody Told You. Here's the Nightly-Diff Pattern That Catches It.
A silent DNS change broke production at 2 PM on a Tuesday. Here's the baseline-and-diff bash pattern that would have caught it the night before.
Jun 10
•
James Cruce
I Secured My AI Agent With a 7-Layer Threat Model
Using the MAESTRO framework to harden an autonomous agent—seven layers of things that can go wrong, translated from security-paper-speak into your day.
Jun 8
•
James Cruce
5 Questions to Ask Before You Build the AI Project Your CEO Just Pitched
A one-page checklist that turns a vague AI proposal into a decision you can defend in writing.
Jun 4
•
James Cruce
1
Anthropic Shipped an AI Security Scanner. Here's the Per-PR Cost Math.
Before you add anything to CI, know exactly what it costs per pull request and how to triage what it finds.
Jun 2
•
James Cruce
Stop Paying for Cloud APIs: Building a Local AI Stack on Mac Studio
How to leverage Apple Silicon's unified memory for production-grade LLMs and replace your cloud billing entirely.
Jun 1
•
James Cruce
May 2026
The Pope Wrote a Memo to AI Developers. Most of You Missed It.
A builder's read of Magnifica Humanitas—what 'disarm AI' actually means, why 'alignment' alone isn't enough, and what to do about it in your config.
May 29
•
James Cruce
I Built a Self-Improving AI Swarm. After 100 Runs It Was No Better Than Run One.
What a flat leaderboard taught me about feedback loops, reward hacking, and why your judge matters more than your model.
May 25
•
James Cruce
Managing Anthropic Agent SDK Costs: A Post-June 15 Billing Playbook
Anthropic moves Agent SDK calls into a $100/mo credit pool on June 15. Here's the two-phase mitigation I shipped: a billing cap plus a provider router.
May 16
•
James Cruce
ChatGPT Just Invented an Entirely Fake Version of My MCP Server
When AI engines don't have you indexed, they don't say 'I don't know.' They confidently make something up. Here's the receipt, and the weekly test I…
May 8
•
James Cruce
1
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
