Writing
All Articles
What If AI Could See the World the Way We Never Could?
What happens when AI systems gain real-time access to infrared, ultrasonic, satellite and microscopic sensors? They begin perceiving the world in ways humans never could — while still facing real technical challenges.
Give Coding Agents the Right Context (Not More of It)
A good prompt is not enough. Recent practical guidance shows how teams give agents the right files, structure, lockfiles, tests, and intent—without drowning them in the whole repo.
The Architecture of Exhaustion: Why Technical Leadership Kills Flow (And Why This Is Not the Way)
Technical leadership often kills flow through constant context-switching, meetings, and cognitive overload. Drawing on Steven Kotler's insights, this post explores why the current way of working is broken—and what needs to change.
Agents While You Sleep: the Runtime Is Trivial, the Pipeline Is Hard
Keeping an AI agent running around the clock is a weekend project. The work pipeline — task source, definition of done, landing gate, allocator — is the real problem. What ~300 multi-agent runs taught me.
The Software Factory Was Built for Humans. AI Needs Something Else.
I tried the agentic software factory. It felt productive at first. Then it became obvious: this whole model was designed for people, not for AI.
Build the Software Factory: Rules and Harnesses for Coding Agents
Coding is no longer the constraint. Recent talks show how senior engineers build repeatable agent loops, scaffolds, and deterministic harnesses so teams stay fast without drowning in slop.
Does Code Quality Still Matter in the Age of AI? It Depends on One Thing
Max Kanat-Alexander explains when code quality still matters in the AI era—and when you can safely ignore it. The key is distinguishing durable software from disposable software.
Disposable vs Durable: When Code Quality Still Matters with AI Agents
AI makes disposable software common. Max Kanat-Alexander’s frame plus software-factory practices show when to invest in quality and when to throw the output away.
The Bottleneck Moved: How Tech Teams Stay Fast When Agents Write the Code
Coding is no longer the slow part. Recent talks from Beyond Coding, the Agentic AI Foundation, David Ondrej, and IBM Technology point to the same bottlenecks: code quality over time, review, trust, and human verification.
The Rise of the 20x Company: How AI is Reshaping Startup Success
Forget the traditional startup playbook. A new breed of '20x companies' is emerging, using AI to automate nearly everything and outperform giants.
The Verification Loop: The $1 Billion Secret Behind Claude Code’s Rapid Success
Discover the "verification loop," the architectural secret behind Claude Code’s record-breaking $1 billion ARR and its unprecedented reliability. By transitioning from simple generation to autonomous testing and validation, AI agents are finally solving the industry’s biggest hurdle: trust. Learn how giving AI the tools to verify its own work can improve output quality by 3x and transform the future of software development.