Freedom Dumlao (aka the API guy) is a CTO and technology leader focused on building systems that scale. He writes about AI, agentic systems, engineering leadership, and the future of software.

Writing ยท Podcasts
The Age of the Agentic Loop (2026)
The most important pattern in AI isn't the chatbot. It's the loop that never stops running.
Vibe Coding Is a Gift. Vibe Security Is a Disaster. (2026)
I love vibe coding. I use Claude Code every day. But the security story around AI-generated code is genuinely alarming.
The CTO's Dilemma: When Your AI Tools Move Faster Than Your Governance (2026)
The gap between AI velocity and governance velocity is the defining challenge for every CTO right now.
2026 Is Not the Year of AI. It's the Year of AI Plumbing. (2026)
Prediction season is full of grand pronouncements. Here's the boring truth about what will actually matter.
Skills, Not Models, Are the Unit of AI Value (2025)
Everyone obsesses over which model to use. The real differentiator is the skill layer.
The Hype Correction Is the Best Thing That Could Happen to AI (2025)
The people panicking about an AI bubble are the same people who never built anything with it.
Disney Just Told You Where AI Is Going. Are You Listening? (2025)
When the most IP-protective company on Earth embraces generative AI, the debate about enterprise adoption is over.
95% of AI Projects Fail Because They Solve the Wrong Problem (2025)
That MIT headline is misleading but directionally useful. Most AI projects fail because teams start with the technology instead of the problem.
Stop Calling It Agentic (2025)
The word 'agentic' has become meaningless. Here's a framework for what it should actually mean.
MCP Won. Now the Hard Part Starts. (2025)
The protocol war never happened. The integration war is just beginning.
AI Layoffs Are a Strategy Tax, Not a Strategy (2025)
Companies cutting headcount and calling it AI transformation are confusing cost reduction with capability building.
The Reasoning Era Has a Reasoning Problem (2025)
Everyone's shipping reasoning models, but almost nobody is reasoning about when to use them.
PhD-Level AI Can't Spell: What GPT-5's Launch Tells Us About Enterprise Readiness (2025)
Benchmarks reward ceiling performance. Enterprise deployment rewards floor performance. Why worst-case behavior matters more than best-case benchmarks.

Find me on GitHub, LinkedIn, or X.