Journal
Writing on AI practice and quantum research
Long-form notes written for people who build things: prompt architecture, AI-assisted engineering workflows, and the state of quantum computing research.
AI PromptingJuly 24, 2026 · 8 min read
Prompt Architecture: Structuring Context Windows for Reliable Output
Most prompt failures are architecture failures. A practical model for laying out instructions, evidence, and constraints in a long context window.
Read article
Quantum ResearchJuly 21, 2026 · 10 min read
Quantum Machine Learning: Where the Advantage Might Actually Come From
Separating credible quantum machine learning research from marketing: kernels, data-loading bottlenecks, and real hardware limits.
Read article
Coding & CursorJuly 19, 2026 · 8 min read
Cursor at Scale: Rules, Context Files, and Multi-File Refactors
How to configure rules, curate context, and stage large refactors so an AI editor helps instead of thrashing.
Read article
Vibe CodingJuly 15, 2026 · 8 min read
Vibe Coding, Seriously: Shipping Fast Without Shipping Debt
Conversational, intuition-led building is genuinely fast. The engineering question is which guardrails let that speed survive its first month in production.
Read articleAI PromptingJuly 11, 2026 · 9 min read
Beyond Chain-of-Thought: Reasoning Patterns That Actually Improve Accuracy
A field guide to decomposition, self-consistency, and verifier loops — plus knowing when extra reasoning is just extra cost.
Read articleQuantum ResearchJuly 8, 2026 · 10 min read
Error Correction in the NISQ-to-Fault-Tolerance Transition
Surface codes, thresholds, and overhead: what below-threshold error correction really demonstrates, and what remains.
Read articleCoding & CursorJuly 3, 2026 · 8 min read
Agentic Coding Loops: When to Let the Model Drive
When to hand work to an autonomous coding agent: a framework based on verifiability, blast radius, and reversibility.
Read articleVibe CodingJune 30, 2026 · 7 min read
From Prompt to Product: A Repeatable Vibe-Coding Workflow
A concrete six-stage loop for turning a conversational prototype into something you would put your name on — with the checkpoints that make it repeatable.
Read articleAI PromptingJune 28, 2026 · 7 min read
Evaluating Prompts Like Code: Test Suites for LLM Behavior
A prompt is a piece of production logic with no type system. Here is how to build the evaluation harness that replaces the compiler you do not have.
Read articleQuantum ResearchJune 25, 2026 · 9 min read
Variational Algorithms for Optimization: A Practical Read
What the evidence on QAOA shows, and how to evaluate any quantum optimisation proposal on real hardware.
Read articleCoding & CursorJune 20, 2026 · 7 min read
Reviewing AI-Written Code Without Slowing Down
AI-assisted teams produce more diffs than review capacity can absorb. A triage model for reviewing what matters and automating everything else.
Read article