Why I Started Pattern Engine
And what you’ll get if you stick around.
I’ve spent the last few years building AI systems at a healthcare technology company. RAG pipelines, MCP integrations, agentic automation. Real systems that touch real patients through real clinicians who are already stretched thin.
I’ve also spent that time watching the AI conversation split into two camps that don’t talk to each other.
Camp one is purely technical. Architecture diagrams, benchmark tables, model comparisons. Valuable, but narrow. It answers how without ever asking why or for whom.
Camp two is purely reactive. Hype cycles, doom loops, hot takes, tool-of-the-week listicles. Lots of noise. Not much signal.
What’s missing is the layer underneath both.
The layer where you ask: what pattern is actually at play here? Not the surface feature. Not the trending topic. The real shape of what’s happening to people, teams, and judgment when machines start recognizing and scaling patterns faster than we can.
That’s what Pattern Engine is.
The premise
Every powerful technology is a pattern machine.
Search engines recognize patterns in information. Social platforms recognize patterns in attention. LLMs recognize patterns in language, then generate new outputs based on everything they’ve absorbed.
These tools don’t create meaning. They compress, reflect, and amplify patterns that already exist in us.
That makes them incredibly useful. It also makes them risky in a specific way: the easier they are to use, the easier it is to stop thinking about what they’re doing to our thinking. We start treating outputs like answers instead of reflections. We start mistaking confidence for truth, speed for progress, coherence for wisdom.
The builders I respect most already know this. They use these tools with skill and with suspicion. They keep humans in the loop not as a formality, but as a real checkpoint on judgment, values, and direction.
Pattern Engine is for those people. And for anyone who wants to become one of them.
What this will look like
Every week, I’ll publish a letter built around a single pattern.
Not a tool review or a prompt trick. A pattern: something showing up in both human life and machine systems that deserves more careful attention than it usually gets.
Things like:
Compression. What happens when nuance gets flattened into a summary, and how that shapes decisions downstream.
Attention. Why what we reward with focus is what grows, whether we’re training a model or training ourselves.
Authority. How “the model said so” is quietly replacing “I thought about it,” and what that costs.
Memory. What AI context windows and human storytelling have in common, and where the comparison breaks down.
Trust. The difference between trust that’s earned and trust that’s assumed by default.
Each letter will move through three layers. This is the spine of everything I’ll publish here:
Layer 1: Wisdom
What have thoughtful people already said about this pattern? Philosophy, psychology, theology, craft traditions. Not as window dressing, but as hard-won insight that showed up long before the current hype cycle.
Layer 2: Framework
How does this pattern show up in teams, organizations, and decisions? Where does it create leverage? Where does it quietly damage judgment or trust? What can a leader actually do about it?
Layer 3: Practice
What does this look like in a real system? Retrieval pipelines, evaluations, automation, agent design, prompt architecture, workflow. Grounded in implementation. No hand-waving.
The goal is a letter that’s useful to someone thinking about AI strategy over coffee and to someone debugging a retrieval pipeline at 2pm. If I can make both of those readers feel like the same conversation matters to them, Pattern Engine is doing its job.
The growing library
Over time, these letters become a library. Each one will include a Pattern Card: a clean, reusable reference that names the pattern, maps how it shows up in people and in machines, flags what it enables and what it distorts, and gives you a question or practice to carry into the week.
The archive is built to grow in value. Week 12 is more useful because weeks 1 through 11 exist. That’s by design. I want this to feel less like a feed and more like a field guide that gets better the longer you stay.
I’m also building this in seasons. Four per year, thirteen weeks each:
Signal - patterns of attention, meaning, perception, and formation
Systems - patterns of workflow, teams, incentives, and feedback loops
Machines - patterns of AI: context, retrieval, evaluation, automation, and agency
Stewardship - patterns of ethics, power, responsibility, and what it means to govern tools well
The seasons rotate every year. Same themes, but the examples evolve with the world. Think of it less like a curriculum and more like a rhythm. A way of paying attention on purpose.
Where I’m coming from
I’m not a futurist. I’m not a thought leader. I’m a working engineer who builds these systems during the day and thinks about what they mean at night.
My background is in software, infrastructure, and applied intelligence. I care about things working. I also care about what they’re working toward, and whether the people involved are being helped or quietly pushed aside.
I believe in God. That shapes how I think about humility, formation, and what intelligence is actually for. You’ll see it in how I write, but this isn’t a theology newsletter. It’s a place where old wisdom and modern systems meet, and neither one gets to look away.
I also believe in craft. In building things well because they should be built well, not because a dashboard told you to. If you’ve ever felt the pull between “move fast” and “get this right,” you’ll feel at home here.
What comes next
The first real letter drops next week. Season 1 is Signal, and we’re starting with a pattern I keep running into everywhere: in my own work, in the broader AI conversation, and in the quiet exhaustion of people who have more tools than ever and less clarity than they’d like.
If you want to be here for it, subscribe.
This isn’t going to be loud. It’s going to be steady, honest, and built to last.
Let’s get to work.
Connor