Archive

2024

What the Heck is Superposition and Entanglement?

This post explains superposition, interference and entanglement in plain words. It covers what a qubit really holds, why quantum computers only win on some problems, why entanglement can't send messages faster than light, and what Google's Willow chip can and can't do today. If you've ever heard the words superposition or entanglement in a conversation about quantum physics, you probably nodded politely and filed them away in the "too confusing" folder. They're worth a second look, because they're the basis of chips like Google's Willow (December 2024), which has 105 qubits. What's in it for us? In short, nothing yet. 105 qubits sounds awesome, but Willow won't crack encryption or make AI better in the next few years. Those 105 are physical qubits. In the Nature paper Google used 101 of them to build a single error corrected logical qubit, and that qubit got better as the code grew. That's the first real sign that error correcti...

Can AI Really Code? What I Got Wrong in 2024 and What Holds Up in 2026

In September 2024 I wrote that AI coding tools were a bit of a hoax and only good for simple tasks. In 2026 that view does not hold up. Coding agents like Claude Code, GitHub Copilot, OpenAI Codex and Cursor are now daily tools for most professional developers, and my own XRPayroll prototype in December 2024 was already about 70% AI-assisted. This update compares my 2024 take with what I see in 2026: what I got wrong, what still holds (architecture, review and security stay with humans), why benchmark scores like SWE-bench need care, and what that means for working with AI agents today. Updated October 2026. I first published this post in September 2024 and added my XRPayroll notes in December 2024. This is a full rewrite. I kept a few of the original lines below so you can see what I said back then. Why I am revisiting this Two years ago I was skeptical about AI coding tools. I tried them, they failed on a task I thought was easy, and I wrote a fairly sharp post about it. Toda...

How to scale MySQL

When MySQL reaches its limits, scaling cannot rely on hardware alone. This article explains how strategic techniques such as caching, sharding and operational optimisation can drastically reduce load and improve application responsiveness. It outlines how in-memory systems like Redis or Memcached offload repeated reads, how horizontal sharding mechanisms distribute data for massive scale, and how tools such as Vitess, ProxySQL and HAProxy support routing, failover and cluster management. The summary also highlights essential practices including query tuning, indexing, replication and connection management. Together these approaches form a modern DevOps strategy that transforms MySQL from a single bottleneck into a resilient, scalable data layer able to grow with your application. When your MySQL database reaches its performance limits, vertical scaling through hardware upgrades provides a temporary solution. Long-term growth, though, requires a more comprehensive approach. This invo...

Beyond Ctrl+F - Use LLM's For PDF Analysis

PDFs are everywhere, but traditional search tools barely go beyond glorified Ctrl+F. This article explores how Large Language Models and Retrieval Augmented Generation can turn static PDF archives into an intelligent, contextual knowledge base that answers real questions instead of just returning files. It walks through a DIY setup built with langchain, transformers and FAISS that loads PDFs, chunks their content, embeds them into a vector store and then uses an LLM to answer questions grounded in the original documents. The result is a practical, self-hostable way to search and reason over your existing PDFs with far more nuance, less hallucination and a clear focus on useful, organisation-specific answers instead of abstract AI hype. PDFs are everywhere, seemingly indestructible, and present in our daily lives at all thinkable and unthinkable positions. We've all got mountains of them, and even companies shouting about "digital transformation" haven't managed to ...

Run Llama3 (or any LLM / SLM) on Your MacBook

Running Llama 3 locally on your MacBook gives you privacy, reliability and speed that the cloud often can’t. Instead of trusting flaky networks and remote servers with your ideas, you keep everything on your own machine, work offline without interruptions and get snappy responses from smaller, well-tuned models like Phi-3. With Ollama handling local model management and easy integration of Hugging Face GGUF models, you can spin up Llama 3 or custom models, tweak templates and experiment freely without cloud limits or surprise bills. For product ideation, this setup means you can brainstorm, prototype and refine deeply specific prompts and workflows in your own sandbox, on your own hardware, with full control over data and experience. I'm gonna be real with you: the Cloud and SaaS / PaaS is great... until it isn't. When you're elbow-deep in doing something with the likes of ChatGPT or Gemini or whatever, the last thing you need is your AI assistant starts choking (...

Key Principles for Building the Best Products and Companies

This article argues that modern product work is not about headcount, rituals or corporate theater, but about small, empowered teams building things customers actually care about. It uses a failed “big team equals importance” sustainability project as a cautionary tale and then lays out core principles for healthy product organisations: cross-functional teams, end-to-end ownership, flexible methodology, a crystal-clear vision, real empowerment instead of micromanagement, tight customer feedback loops, outcome-focused metrics and protecting team health. It then gets practical: why shipping features alone does not win, how a hybrid of Kanban flow and agile iteration keeps teams responsive, the mindset shifts that keep a team customer-obsessed, prioritizing with MoSCoW, and Amazon’s working backwards method with an internal press release template, the "Family talk" test and the press release as the red line against overbuilding. Drawing on Nike’s classic “our business is change”...

OSX improved (Update)

Updated May 17, 2024 to fit M* architecture My favorite development environment on my MacBook includes an improved Zsh shell and an extended .vimrc configuration file with syntax highlighting, error checking, TextMate snippets, and the Solarized color scheme.  Here's a guide for setting up similar features:  The features include directional key navigation for directories and files, developer-friendly colors, command highlighting, improved history search, auto-complete for options and SSH connections (if keys are known), and many more useful enhancements.   Get Xcode:  AppStore => Xcode => Install Xcode From now we use a terminal window. Install Brew /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)" Install git and wget:   brew install git   brew install wget Install oh-my-zsh:   sh -c "$(curl -fsSL https://raw.githubusercontent.com/ohmyzsh/ohmyzsh/master/tools/install.sh)" The script wa...

AI in Product Development?

This piece explains how generative AI has become a genuine accelerator for product ideation after years of underwhelming brainstorming tools. Instead of chaotic workshops dominated by loud voices or stale thinking, modern multimodal models like GPT-4o can surface unconventional ideas, emulate customer reactions, analyse markets, and help prototype concepts rapidly. Used well, AI becomes an “unfair advantage” for founders and product leaders by expanding creative range, grounding ideas in data, and sharpening strategic decisions. The article outlines a practical workflow, defining the problem, generating solutions, prototyping, and running devil’s-advocate analyses, plus a role-play method used successfully with Apache Wayang. The takeaway: AI doesn’t replace human judgment, but in a mature XOps culture it becomes a force multiplier that helps teams think clearer, move faster, and build better products. I like product ideation brainstorming, done right and focused,  it opens my m...

What The Heck is XOps in Product Development?

This article reframes XOps as a practical, collaborative mindset rather than another corporate buzzword. Drawing on experience as a CPO, it argues that breaking down silos between product, design, data and sales is essential for building user-centric products, especially in greenfield environments where culture forms early. XOps provides the glue: shared outcomes, cross-functional learning, data-driven decisions and fast experimentation. The piece outlines how each role contributes to a unified product engine, why collaboration reduces wasted effort and accelerates time-to-market, and how flexible deployment models, from embedded ProductOps to external consultants, help teams adopt the approach. The takeaway is clear: XOps succeeds only when leaders drive cultural change, champion honest feedback and reward collective impact. Done right, it becomes a strategic advantage that speeds delivery, reduces waste and creates happier, high-performing teams. First: XOps is not a new Marvel mo...