# Jdhwilkins
> Field notes on math, technology and everything in between.
Public Ghost content for AI and LLM tooling. Use `/llms-full.txt` for consolidated page and post context.
Append `.md` to any post or page URL to get the content in Markdown (for example, `/example-post.md`).

## Pages
- [About](https://www.jdhwilkins.com/about.md)
- [Coffee](https://www.jdhwilkins.com/coffee.md)
- [Contact](https://www.jdhwilkins.com/contact.md)
- [Newsletter](https://www.jdhwilkins.com/newsletter.md)
- [Now](https://www.jdhwilkins.com/now.md)
- [Portfolio](https://www.jdhwilkins.com/portfolio.md)
- [Privacy Policy](https://www.jdhwilkins.com/privacy-policy.md)
- [Work With Me](https://www.jdhwilkins.com/work-with-me.md)

## Posts
- [Oh, your AI hallucinates? Yeah, I used to have that problem too](https://www.jdhwilkins.com/oh-your-ai-hallucinates-yeah-i-used-to-have-that-problem-too.md) - I build AI content workflows for a living, so reducing hallucinations matters. Here are the four techniques I actually use, from grounding to a critic-actor review step.
- [Lake District Trip June 2026](https://www.jdhwilkins.com/lake-district-trip-june-2026.md) - A lovely hike, great company, and incredible views over Crummock Water and Mellbreak Peak. Who needs GPS when you can navigate with a piece of paper as long as you are tall. A massive thanks to the guys who stopped to help with the car after parts started falling off. Absolute lifesavers! 0:00 /0:0…
- [AI-generated writing has its place—it's just not here.](https://www.jdhwilkins.com/ai-generated-writing-has-its-place-its-just-not-here.md) - AI is genuinely good at some writing tasks—just not all of them. Telling the difference isn't easy, and getting it wrong costs you more than you might expect.
- [How (and why) I Built My New Home on the Internet](https://www.jdhwilkins.com/how-and-why-i-built-my-new-home-on-the-internet.md) - From WordPress to Wix to coding my own — the story of why I rebuilt my website from scratch and what I learnt about building with Claude Code along the way.
- [What's That? Media and Text in the Same Feed!?](https://www.jdhwilkins.com/whats-that-images-and-text-in-the-same-feed.md) - Absolutely, it is. And what better way to begin than with a coffee? This is one of my new post content types. I can share media, add captions, and have it presented in a more natural format in the feed. Here's my current Obsidian vault graph. I posted about this setup recently, which you can read m…
- [6 Honest Lessons From 3+ Years of AI](https://www.jdhwilkins.com/6-honest-lessons-from-3-years-of-ai.md) - Using AI well is less about the tools and more about how you think with them. Six lessons from three years of experimentation, including where I got it wrong.
- [I Wish I Knew This Before Building an AI Second Brain](https://www.jdhwilkins.com/i-wish-i-knew-this-before-building-an-ai-second-brain.md) - Most AI second brain setups fail because the vault isn't built for it. This covers the structure, linking scheme, and agent protocol that actually make it work.
- [Everything You Need to Know Before You Start Using Obsidian](https://www.jdhwilkins.com/everything-you-need-to-know-before-you-start-using-obsidian.md) - Discover how to master Obsidian with our comprehensive guide. Learn why Obsidian is unique and how to navigate its features effectively today!
- [How I Use Claude as a Writing Assistant (to actually improve my writing)](https://www.jdhwilkins.com/how-i-use-claude-as-a-writing-assistant-to-actually-improve-my-writing.md) - AI-generated articles and AI-assisted writing aren't the same thing. Four practical techniques for using Claude as feedback, without losing your own voice.
- [Seamless Content Ingestion for Claude-Obsidian Second Brain](https://www.jdhwilkins.com/seamless-content-ingestion-for-claude-obsidian-second-brain.md) - Explore seamless content ingestion for the Claude-Obsidian second brain. Discover tips on efficiently managing your Claude-Obsidian second brain.
- [Few-shot prompting Vs. Zero-shot prompting. Which approach and when?](https://www.jdhwilkins.com/few-shot-vs-zero-shot-prompting-which-approach-and-when.md) - Most prompts don't need examples. This explains what few-shot prompting does, why examples help models, and when to actually reach for them.
- [How to Rot Your Brain with AI](https://www.jdhwilkins.com/rot-your-brain-with-ai.md) - Studies show heavy AI users score lower on critical thinking tests. The difference isn't which tools you use - it's whether you stay engaged while using them.
- [How I Built an AI-Powered Task Management System with Obsidian and Claude Code](https://www.jdhwilkins.com/how-i-built-an-ai-powered-task-system-with-obsidian-and-claude-code.md) - I built an AI-powered task system in Obsidian using custom Claude Code skills. This covers context engineering, note types, and the morning briefing automation.
- [Role prompting - Giving AI a Persona](https://www.jdhwilkins.com/role-prompting-giving-ai-a-persona.md) - Role prompting is one of AI's most repeated tips. The research tells a different story. Here's what expert personas actually do to your outputs, and when to use them.
- [Jargon Buster i. AI, Machine Learning, Deep Learning, Generative AI](https://www.jdhwilkins.com/jargon-buster-i-ai-machine-learning-deep-learning-generative-ai.md) - Every tech pitch uses AI, ML, and deep learning interchangeably. They're not the same. Here's what each term actually means and why the differences matter.
- [AI's Blindspot: The 'Lost in the Middle' Effect](https://www.jdhwilkins.com/ai-s-blindspot-the-lost-in-the-middle-effect.md) - AI models struggle to recall information in the middle of long documents. It's a positional bias called the Lost in the Middle Effect. Here's how to fix it.
- [Tokens and Context Windows: What They Are and Why They Matter](https://www.jdhwilkins.com/tokens-and-context-windows-what-they-are-and-why-they-matter.md) - Tokens are how AI measures text. Context windows are working memory. When they fill, earlier content disappears. Here's what that means and how to manage it.
- [Claude Code: Everything you need to know in 8 minutes or less](https://www.jdhwilkins.com/claude-code-everything-you-need-to-know.md) - Claude Code works directly on your files, runs code, and manages tasks without you needing to be a developer. Here's what you need to know to use it well.
- [MCP Servers: Giving Your AI Agent New Capabilities](https://www.jdhwilkins.com/mcp-servers-giving-your-ai-agent-new-capabilities.md) - MCP servers let your AI agent read from and act on external services without manual copy-pasting. Covers setup, authentication, and when to build your own.
- [Level-up your AI Agent with Skills Engineering](https://www.jdhwilkins.com/level-up-your-ai-agent-with-skills-engineering.md) - Skills are reusable instruction sets for AI agents - write once, the agent uses them on demand. Covers structure, loading logic, security, and authoring tips.
- [Getting Started with Agent Skills](https://www.jdhwilkins.com/getting-started-with-agent-skills.md) - Agent skills let you brief your AI once and have it work your way every time. Learn what they are, how to write them, and where to find ready-made ones.
- [Exploring AI in Animation: My Journey with VibeManim](https://www.jdhwilkins.com/vibemanim-ing-spatial-reasoning-and-geminis-secret-superpower.md) - Testing LLMs on Manim animation code reveals where spatial reasoning breaks down. Four models, the same prompts, progressively harder challenges.
- [Getting Started with AI: The Fundamentals of Prompting](https://www.jdhwilkins.com/the-fundamentals-of-prompting-getting-started-with-ai.md) - Prompting is a skill. This covers what an effective prompt includes - context, task, constraints, format, and examples - plus five techniques to get started.
- [Forget "Think step by step", Here's How to Actually Improve LLM Accuracy](https://www.jdhwilkins.com/why-think-step-by-step-no-longer-works-for-modern-ai-models.md) - "Think step by step" used to reliably improve AI outputs. Modern reasoning models handle this internally. Here's what that means for how you should prompt now.
- [Using AI to Map Legacy and Inherited Code: A Practical Guide](https://www.jdhwilkins.com/using-ai-to-map-legacy-and-inherited-code-a-practical-guide.md) - Master AI-driven code mapping with this 4-step guide to navigating legacy projects. Turn complex code into clear diagrams and actionable documentation.
- [The pitfalls of over-reliance on AI for self-directed learning](https://www.jdhwilkins.com/the-pitfalls-of-over-reliance-on-ai-for-self-directed-learning.md) - Do you over-rely on AI? Understand the risks of using AI for self-directed learning. Struggle productively. Reclaim your critical thinking.
- [How Far Can You Push Excel with Regex?](https://www.jdhwilkins.com/how-far-can-you-push-excel-with-regex.md) - Learn how to use regex pattern matching in Excel for advanced data cleaning, text extraction, and analysis.
- [Convolution, Kernels & Filters: A Beginner’s Guide to Edge Detection](https://www.jdhwilkins.com/convolution-kernels-filters-a-beginner-s-guide-to-edge-detection.md) - Edge detection is how machines find structure in images. This beginner's guide starts with one-dimensional examples and builds up to convolution kernels and filters.
- [Still copy-pasting into ChatGPT? Here’s how to turn your ideas into AI-powered apps](https://www.jdhwilkins.com/still-copy-pasting-into-chatgpt-here-s-how-to-turn-your-ideas-into-ai-powered-apps.md) - Learn to build AI-powered apps using LLM APIs. This beginner's guide covers prompt engineering, function calling, and creating custom AI tools.
- [You’re using ChatGPT wrong. Here’s how to prompt like a pro](https://www.jdhwilkins.com/you-re-using-chatgpt-wrong-here-s-how-to-prompt-like-a-pro.md) - Six advanced prompting techniques that improve ChatGPT outputs: roleplay, decomposition, chain-of-thought, tree-of-thought, shared context, and agreeableness.
- [22/7 and the Approximation of Irrational Numbers](https://www.jdhwilkins.com/22-7-and-the-approximation-of-irrational-numbers.md) - Unlock the math behind 22/7! Learn how continued fractions generate increasingly accurate rational approximations for irrational numbers like pi
- [Formal Grammars – The Hidden Mathematics of Language](https://www.jdhwilkins.com/formal-grammars-the-hidden-mathematics-of-language.md) - Explore the math behind language. Learn how formal grammars shape speech, code, NLP, and the rules machines use to understand us.
- [Higher Mathematics: Sets and Notation](https://www.jdhwilkins.com/higher-mathematics-sets-and-notation.md) - Master set theory: definitions, notation, set operations (union, intersection), and common sets. Essential concepts for higher mathematics.
- [Model Evaluation: Why Accuracy Isn’t Enough](https://www.jdhwilkins.com/model-evaluation-why-accuracy-isn-t-enough.md) - Learn why accuracy fails in imbalanced data. Master essential model evaluation metrics: Precision, Recall, and the F1 score for robust ML classification.
- [Mountains, Cliffs, and Caves: A Comprehensive Guide to Using Perlin Noise for Procedural Generation](https://www.jdhwilkins.com/mountains-cliffs-and-caves-a-comprehensive-guide-to-using-perlin-noise-for-procedural-generation.md) - Minecraft generates unique worlds with Perlin noise. Covers the algorithm, octaves, and Fractal Brownian Motion - with a terrain implementation walkthrough.
- [From Love to Logic: How Algorithms Decide Our Matches](https://www.jdhwilkins.com/from-love-to-logic-how-algorithms-decide-our-matches.md) - Explore the Gale-Shapley Algorithm and the Stable Matching Problem. Learn how algorithms create stable pairings, not just perfect matches.
- [The Impossible Bridges of Königsberg](https://www.jdhwilkins.com/the-impossible-bridges-of-konigsberg.md) - Learn how to find an Eulerian Path using Graph Theory and the degree of a vertex. Discover the solution to the Bridges of Königsberg puzzle.
- [Surviving the First Year of Your Math Degree](https://www.jdhwilkins.com/surviving-the-first-year-of-your-math-degree.md) - Essential tips for math degree survival at university. Master notation, proofs, and the shift from high school to degree-level mathematics.
- [Information at a Glance: Do Your Charts Suck?](https://www.jdhwilkins.com/information-at-a-glance-do-your-charts-suck.md) - Master data visualization: use pre-attentive features, Gestalt theory, and proper visual data encoding to create clear, effective charts
- [Turing’s Turochamp: The Birth of AI in Games](https://www.jdhwilkins.com/turing-s-turochamp-the-birth-of-ai-in-games.md) - Discover Turing's Turochamp, Learn about Alan Turing's foundational paper-and-pencil chess algorithm and the birth of AI in games.
- [The Hunt For Prime Numbers](https://www.jdhwilkins.com/the-hunt-for-prime-numbers.md) - Explore prime number math: shortcuts, primality tests, Mersenne numbers, and the Lucas-Lehmer Test for finding the largest primes.
- [Flowers, Staircases, and The Golden Ratio](https://www.jdhwilkins.com/flowers-staircases-and-the-golden-ratio.md) - Explore the Fibonacci sequence, the Golden Ratio (φ), and their appearance in nature, from flowers to the staircase problem in math.
- [What if an Infinite Number of Spaceships Arrive at Hilbert’s Hotel?](https://www.jdhwilkins.com/what-if-an-infinite-number-of-spaceships-arrive-at-hilbert-s-hotel.md) - Explore Hilbert's Hotel Paradox and transfinite numbers. Learn how to accommodate infinite guests, buses, and spaceships using infinity math.
- [Proof by Induction (Ft. The Tower of Hanoi)](https://www.jdhwilkins.com/proof-by-induction-ft-the-tower-of-hanoi.md) - Learn the Tower of Hanoi Proof using Proof by Induction. Discover the minimum moves required for this classic combinatorics puzzle.
- [Python Chess: Game Simulation and Illegal Moves](https://www.jdhwilkins.com/python-chess-game-simulation-and-illegal-moves.md) - Part 3 of building a Python Chess Engine. Implement the move function for game simulation, encode move types, and filter out illegal moves.
- [Who Needs Strava Premium When You Have Python?](https://www.jdhwilkins.com/who-needs-strava-premium-when-you-have-python.md) - Build a Strava heatmap using the Python API for custom GPS route visualization. Filter by type and date range for advanced data analysis.
- [Python Chess: Efficient Move Generation Using Bitwise Operations](https://www.jdhwilkins.com/python-chess-efficient-move-generation-using-bitwise-operations.md) - Python chess move generation using bitwise operations and lookup tables. Covers all piece types, occupancy masks, and special moves including en passant.
- [How to Use the Strava API with Python – Beginner Tutorial](https://www.jdhwilkins.com/how-to-use-the-strava-api-with-python-beginner-tutorial.md) - Beginner's guide to the Strava API with Python. Use OAuth2 for authorization and get activity data. Handle access tokens and usage limits.
- [Image Segmentation With K-Means Clustering](https://www.jdhwilkins.com/image-segmentation-with-k-means-clustering.md) - Learn K-Means Clustering for Image Segmentation in Python. Achieve colour reduction and optimized performance using NumPy
- [1-Hot Encoding: The Chess Programmer’s Secret Weapon](https://www.jdhwilkins.com/1-hot-encoding-the-chess-programmer-s-secret-weapon.md) - Maximize chess engine speed with Python Bitboards. Use 1-Hot Encoding and bitwise operations to create an efficient board data structure.
- [Solving the Travelling Salesman Problem Using a Genetic Algorithm](https://www.jdhwilkins.com/solving-the-travelling-salesman-problem-using-a-genetic-algorithm.md) - Solve the Travelling Salesman Problem with a Genetic Algorithm. Learn selection, crossover, and mutation steps to find a near-optimal route.
- [Why We Can’t (Completely) Solve The Travelling Salesman Problem](https://www.jdhwilkins.com/why-we-can-t-completely-solve-the-travelling-salesman-problem.md) - The TSP is NP-Hard due to exponential complexity. Understand the time-accuracy trade-off in finding optimal vs. approximate routes.
- [Genetic Algorithms: Machine Learning with Darwinian Evolution](https://www.jdhwilkins.com/genetic-algorithms-machine-learning-with-darwinian-evolution.md) - The Genetic Algorithm uses Darwinian Evolution for machine learning. Learn selection, crossover, and mutation for efficient optimisation.

## Optional
- [RSS Feed](https://www.jdhwilkins.com/blog/rss/)
- [Sitemap](https://www.jdhwilkins.com/sitemap.xml)
- [Full content of pages and posts](https://www.jdhwilkins.com/llms-full.txt)