March 2026
When Your AI Agent Learns to Sleep: Claude Code's Auto Dream and the Architecture of Machine Memory
Claude Code quietly added a memory consolidation feature modeled after human REM sleep. It reviews hundreds of sessions, prunes stale memories, and resolves contradictions automatically. Here's how it works and why it matters.
When AI Becomes Infrastructure: What Anthropic's Four New Primitives Actually Signal
Projects, Scheduler, Dispatch, and Computer Use are not four product updates. They are four primitives of the same architectural pattern: AI as infrastructure you direct, not software you operate.
The Autoresearch Pattern: What Happens When You Let AI Optimize Its Own Instructions
Andrej Karpathy's three-file framework runs hundreds of experiments overnight. We applied the same pattern to our blog writing skill and improved quality scores from 74 to 89 in five experiments. Here's what we learned.
The Agentic Operating System: Three Architectures, One Structural Shift
NemoClaw, Perplexity Computer, and Anthropic Cowork each represent a different bet on the same structural shift: the operating system that matters most is no longer the one that manages your files. It's the one that manages your agents.
After NemoClaw: Why Every B2B Website Needs an Agent-Readable Layer Now
Nvidia just shipped enterprise AI agents that browse websites autonomously. We audited our own site and found every AI crawler blocked. Here's what we learned and the framework for fixing it.
Beyond the Billable Hour: Why the Human-AI Hybrid Makes Time-Based Pricing Obsolete
When one person with AI can deliver what used to require a team, pricing by the hour penalizes the best professionals and shortchanges clients. The alternative is already here.
February 2026
The Great Convergence: The Shift from Doing to Directing in the AI Era
Every knowledge work role is collapsing into a single meta-competency: directing AI agents. The people who win aren't the most technically skilled. They're the ones who think about it differently.
The New Default: Why Not AI?
The question has flipped from "How can I use AI?" to "Why am I not using AI?" and the answer is reshaping how knowledge work gets done.
Delete Then Automate: Five Operating Principles That Compound in the AI Era
Five operating principles from high-performance organizations, and why they matter more now that AI amplifies every operational advantage.
Welcome to February: The Agent Era Arrived Faster Than Anyone Expected
The agent era is here, and the implications for organizations are more immediate than expected.
January 2026
14 Projects. 12 Months. Zero Without AI.
How INS built fourteen production applications in twelve months with AI assistance.
When Domain Knowledge Compounds
How domain expertise accelerates AI-assisted development over time.
The End of Technical vs Non-Technical
Building capabilities through progressive AI-assisted learning.
Your Data Is Safe: How SOC 2 Compliance Protects INS and Our Customers
How INS approaches security and SOC 2 compliance in AI-assisted development.
Your AI Conversations Are Training Data: How to Opt Out
Understanding privacy considerations when working with AI systems.
December 2025
The Super Exponential AI Timeline
Understanding the accelerating pace of AI capability growth and its implications.
The Primitive Advantage: Unleashing the Potential of AI
Exploring the primitive advantage and the future of AI agents.
From Prototype to Production: The 98% Nobody Sees
The journey from initial concept to production-ready application.
Claude Code & Excalidraw Diagrams
Creating technical diagrams with AI assistance using Excalidraw.