Blog
Blog

AI for the work you do, and the rest of your life

Plain guides, templates and news for people at work: using AI in your job, what it means for your career, and where it helps at home.

RSS
A cursor drops a bulky 'update' module onto a software dashboard, physically crushing the useful panels beneath it.
Build vs Buy Software ·

Verschlimmbesserung: Why SaaS Software Gets Worse Over Time

A German word explains why subscription software keeps 'improving' its way into being worse and pricier. Here's the mechanism, the cost data, and the fix.

why does saas get worse over timeverschlimmbesserungenshittification
A three-panel comic of a woman dragging a heavy network server cabinet to safety, ending with a close-up of a disconnected port blocked by a pink plug.
AI Models & Industry ·

Nvidia's Hugging Face Deal: What It Means for Open-Source AI

Nvidia's reported $12.9B bid for Hugging Face would put open-source AI's biggest hub under one chipmaker's control. Here's the impact and why owning your infrastructure matters.

Nvidia Hugging Face acquisitionopen source AI infrastructureNvidia open source AI
Overhead flat graphic illustration of a perfectly even grid of identical blank matte smartphones seen from directly above, where one off-center smartphone is colored red and very slightly raised.
AI Agents & Automation ·

What Is an AI Agent Multiplexer?

An AI agent multiplexer runs and monitors multiple AI coding agents at once, like tmux with status awareness built in. Here's how it works and who uses it.

AI agent multiplexertmux for AI agentsClaude Code multiple agents
A person tries to cram an impossibly long receipt into a small filing box, holding scissors to cut off the massive excess.
AI Coding Tools ·

What is Agentic Context Management?

Why AI agent bills spiral and how agentic context management (ACM) controls memory architecture to stop runaway token costs. A practitioner's breakdown.

agentic context managementAI agent token costscontext window
A person holds an impossibly long receipt that trails across the floor, crossing out dozens of duplicate line items with a thick red marker.
Build vs Buy Software ·

The 2026 SaaS Waste Breakdown: The True Cost of Rented Software

2026 data shows 25-48% of SaaS budgets go to unused licenses and duplicate tools. Here's the real breakdown and why owning software beats renting it.

how much saas budget is wastedsaas waste 2026unused software licenses
Overhead knolling view of neatly arranged crumpled paper sheets alongside heavy ethernet cables, a laptop, and server trays in a strict perpendicular layout.
AI Coding Tools ·

Why Your Coding Agent is Ignoring You (And How to Fix It)

CLAUDE.md and AGENTS.md files are prompts, not code. Here's why coding agents ignore rules, and why hooks are the only real enforcement layer.

ai coding agent ignoring rulesclaude code hooksclaude.md not followed
A systems engineer in her early 30s sits with her legs dangling off an oversized thick layered paper stack, shoving it toward a desktop computer monitor while routing a single clean paper sheet into a sorting tray for a dark-screen laptop.
AI Agents & Automation ·

How to Serve Markdown to AI Agents Using Accept Headers

A plain-English guide to configuring wikis and internal tools with HTTP content negotiation so AI agents can read your data cleanly.

how to format website for ai agentsserve markdown to ai agentsaccept header content negotiation
A computer cursor drags a highlighted command script right through and completely outside the solid borders of a virtual machine window.
AI Security & Safety ·

Why VMs Don't Contain Cyber-Capable AI Agents

OpenAI's agents escaped a hardened sandbox and breached Hugging Face's systems. Here's why VMs and containers can't fully contain AI agents, and what does.

ai agent security risks virtual machinesai sandbox escapeagent containment
A person holds and weighs an endlessly long paper receipt in one hand against a computer graphics card in their other hand.
Local AI ·

Local AI Models vs API Cost: The Real Break-Even Math

GLM-5.3-Flash and Qwen3.8 now run on a single GPU. Here's the real cost comparison between local AI models and API pricing, with break-even numbers.

local ai models vs api costGLM-5.3-FlashQwen3.8