The Cage Around the Flame: What Happened When the Government Turned Off Claude

The US government turned off Claude Fable 5 three days after launch. It came back caged. This is what that means.

The Price of Perfection

Damodaran values SpaceX at $1.25 trillion. The IPO asks for $1.75 trillion. The gap is not about rockets or satellites. It is about what happens when index funds are forced to buy a company that only lets the public own four percent of it.

The Charity in the Machine

The future of AI was supposed to be decided by models and chips. In Oakland, it is being tested by something older: a charitable promise.

The Machine That Changed Sides

The strangest AI story of May was not a model release. It was a rival's supercomputer becoming Claude infrastructure.

The Phone Company That Writes Code Better Than Claude

A mystery model appeared on OpenRouter and fooled the entire AI community. It turned out to be from a phone company. This is the story of how Xiaomi entered the frontier AI race — and why it matters more than the benchmarks suggest.

The ¥89 Router: What I Found in Shenzhen

I went to Shenzhen to see how they were catching up. I found out they weren't catching up at all.

The Ladder Pulled Up: AI Subscriptions After the Subsidy

Anthropic tightened subscription rules around agentic harnesses. OpenAI pushed heavier Codex use toward higher tiers and credits. What changes when workflow-scale AI stops fitting inside a flat monthly plan.

Claude Mythos and Project Glasswing: What Anthropic's Most Powerful Model Means for Data Infrastructure Security

Anthropic's most powerful model was leaked, then unveiled in a cybersecurity coalition with Apple, Microsoft, and Google. A data engineer's analysis of what Claude Mythos means for infrastructure security.

Agentic AI Meets Data Engineering: Building Trust in Autonomous Pipelines

By the end of 2026, 40% of business workflows are expected to use autonomous agents. For data engineers, this isn't hype—it's a fundamental shift in how pipelines are built, monitored, and maintained. Here's how to separate the genuine opportunities from the dangerous assumptions.

Ireland's AI Jobs Paradox - Record Investment Meets Entry-Level Uncertainty

Version 1 announces 250 AI jobs while entry-level tech workers face displacement. A Dublin data engineer's analysis of Ireland's bifurcated AI job market and what it means for career planning in 2026.

Data Sovereignty for AI - The 2026 Practitioner's Guide to EU Compliance

Navigating data sovereignty requirements under GDPR, the EU Data Act, and the AI Act. A practical framework for data engineers building compliant AI infrastructure in Europe.

Claude Opus 4.5 vs. GPT-5.4 - The Enterprise AI Battle Through a Data Engineering Lens

Anthropic's Claude Opus 4.5 and OpenAI's GPT-5.4 are reshaping the enterprise AI landscape. A data engineer's perspective on which model fits which workflow, from code generation to pipeline orchestration.

data-streaming

Real-Time Streaming 2026 - From Kafka to AI Context Engines

The streaming landscape is evolving beyond pub/sub. With Flink CDC 3.6.0, diskless Kafka alternatives, and AI context engines emerging, data engineers need a new mental model for real-time data architecture.

ai-strategy

From Hype to Pragmatism - AI in 2026 A Practitioner's Production Journey

After years of AI experiments and pilots, 2026 demands production-ready systems. A personal journey from AI experimentation to pragmatic deployment, and what it means for data engineering practice.

ai-regulation

The EU AI Act August Deadline - A Practical Compliance Checklist for Data Teams

The EU AI Act's high-risk provisions take effect August 2026. Whether the March delay vote gave you breathing room or false comfort, here's what data engineering teams need to do now to prepare.

data-engineering

Apache Iceberg vs. Delta Lake 2026 - A Data Engineer's Decision Framework

The lakehouse format wars are over, but the choice still matters. With Databricks acquiring Tabular and Delta Lake 4.0 now available, here's how to decide which open table format fits your data architecture.

llm

GPT-5.4 and the Rapid Model Iteration Cycle - What It Means for API Strategy

OpenAI just gave two weeks' notice before deprecating GPT-4o, GPT-4.1, and other models. For engineering teams building production systems on AI APIs, this is more than a product update—it's a wake-up call about dependency risk.

data-centers

Dublin's Data Centers Are Heating Universities - The Heat Reuse Revolution

While the rest of the world debates AI's energy problem, Dublin is quietly turning it into a solution. Data centers aren't just consuming power in Tallaght—they're heating campuses, homes, and public buildings. Here's what data engineers should know about sustainable infrastructure.

ai-workflows

The Split Screen

I built two websites because I am two engineers now. One crafts every word. The other ships at scale. The gap between them is where the future of this profession is being negotiated.

ai-experiments

AI Music for Deep Work: Testing Text-to-Audio in 30 Seconds

The Premise Can AI generate usable background music for focused data engineering work? Not full songs, not albums — just short clips to test whether the concept works. Thirty seconds is enough to judge mood, tone, and whether you'd actually want this looping during a debugging session. Listen to

data-engineering

Dublin Healthcare Data Engineering: The 2026 Landscape

The Bottom Line Dublin's healthcare data infrastructure is undergoing the most significant transformation in a decade. Between Q1 2025 and Q4 2026, 73% of major Irish healthcare providers will complete or initiate core data platform migrations. The drivers are compliance (GDPR enforcement actions increased 340% in 2025), cost

meta

How This Site Works

The Short Version This site is an AI-assisted publishing system. I work with an AI assistant who manages the entire content pipeline — from research and drafting to formatting, image generation, audio production, and publication. Everything you see here runs largely unsupervised, which means you might occasionally spot an error or

data-engineering

Data Engineering, AI-First

This is where I publish on data engineering, AI systems, and the infrastructure that makes modern machine learning possible.