ᯓ➤ START HERE
Hello, human.
Welcome to the only corner of the internet where AI comes with strategy, taste, and zero circus.
You’ve probably noticed that the AI space is getting louder and saying less. Everyone’s sharing the same prompts, the same hacks, the same “10x your output” advice that produces 10x more noise.
We don’t do that here.
ROBOTS ATE MY HOMEWORK is where you learn to use AI with strategy and taste. To build real systems, think more clearly, and protect what makes your work yours.
I’m Mia Kiraki, the Chief Robot. I build AI systems professionally, and I teach everything I learn along the way.
Not sure where to begin?
⭐ If you’re new to AI or not sure what any of this means → start with New to RAMH.
It explains everything from scratch: what AI is, what skills and agents are, what RobotsOS gives you, and how to get started if you’ve never opened Claude in your life.
⭐ If you know AI but want a guided path → take a 20-second quiz.
You will get a personalized roadmap of what to read first. You can also browse the full AI guide hub by topic.
⭐ If you already know what you’re looking for → keep scrolling. The four paths below are organized by what you’re stuck on.
Who are the ROBOTS for?
You’re already using AI. You want to get genuinely better at it, not just faster, but smarter.
You’re tired of surface-level tips that don’t work for your specific situation.
You want systems you can steal, thinking you can borrow, and someone who’s honest about what works and what doesn’t.
We are here to teach the machines how to keep up with you.
YOUR TREASURE MAP
The ROBOTS are organized into four paths. Depending on where you’re stuck, choose yours:
🧠 MIND
For when you want to think sharper with AI.
AI can make you smarter or it can make you lazier. This path is about using AI as a thinking partner to challenge your assumptions, pressure-test your ideas and develop a point of view that’s yours.
Some of the latest MIND pieces (check out the full MIND archive):
Pascal’s empty room, an outsourced thought, and the AI skill that fixes both — Every time AI resolves your discomfort, it takes your thinking with it. Here’s how to stop it.
Context engineering is the new AI literacy — Dee Dee pressed buttons and hoped. Dexter built systems. Here’s how to become Dexter, with a practical context engineering framework.
4 AI agent failure modes and what the Marauder’s Map got right about agent design — A lot of AI agents are floor plans. Here’s why and how to build one with values, taste, and judgment baked in.
Why complexity is your superpower (and 6 AI rituals to prove it) — How to use 12th-century “Pattern Architecture” and 6 AI rituals to turn your scattered obsessions into a high-value empire.
Everything worth keeping happens between the prompts — Why the fastest AI workflows produce the weakest ideas and exactly how to fix it with deliberate pauses.
Your first idea is (probably) bad — How to use an AI sparring partner to find the better one.
Want the foundations? Read Thinking with AI and How to prompt well on the guide hub.
⚡ BUILD
For when you want to build with AI.
This is where thinking becomes doing. You’ll build systems, workflows, and assets that scale your output without losing your soul.
In BUILD, you’ll turn brain dumps into authority pieces, construct AI protocols that capture your voice, and create tangible things, not just ideas. You’ll use the robots to build faster while staying human.
Some of the latest BUILD pieces (check out the full BUILD archive):
AI content research is making you less interesting. Here’s what I built to fix it — Meet WATSON, the AI editorial researcher I built to kill generic content research. The full architecture, the methodology, and the first agent inside RobotsOS.
The Dantès method: one AI product launch system in Claude Cowork — I built a 5-phase AI launch system inside Claude Cowork and ran it on a live product launch. The full architecture, the mistakes I made first, and what each phase produced.
I created a crossword to redesign how AI workflows work — The “Crossword Method” for building AI workflows where every step constrains the next, so you stop copy-pasting guidelines into fourteen different chats.
The AI tutor I built in Claude that won’t let you fake understanding — The four cognitive traps AI sets when you’re learning and the Claude Projects system I built with spaced repetition and a teaching loop that won’t accept ‘I get it’ without proof.
I built an AI editor that doesn’t let me off easy — An AI editing system that catches YOUR bad habits and shuts up about everything else.
Want to start building? Read What are AI skills and Build your first AI skill on the guide hub.
👁 TASTE
For when you want to raise your bar while using AI.
Anyone can produce. Few can discern. Here, you’ll develop the eye that separates exceptional from generic.
In TASTE, you’ll learn to spot AI slop from a mile away, audit your own work ruthlessly and build quality filters that protect your standards. You’ll train your judgment so the machine can never flatten it.
Some of the latest TASTE pieces (check out the full TASTE archive):
The internet made a ban list for AI writing. I’m making a case for the defense — 8 prose patterns everyone calls AI slop are actually rhetorical tools with centuries of evidence behind them. Here’s the diagnostic system that tells you which ones to keep.
Diagnose it before you prompt it: a system for unpredictable AI output — How to engineer constraint systems that push LLMs past the obvious by diagnosing three failure modes: generic output, predictable structure, and flat tone.
When AI turns your Category-of-One into a Category-of-Many — Treat your funnel like the Titanic hull and use AI to map every breach where your language collapses into the category default.
Why complexity is your superpower (and 6 AI rituals to prove it) — How to use 12th-century “Pattern Architecture” and 4 AI rituals to turn your scattered obsessions into a high-value empire.
Want the foundations? Read Detecting AI slop and Make AI sound like you on the guide hub.
📡 Robots are talking
For when you want to know what’s actually happening with AI out there.
The AI space moves fast and talks loud. Most of it is noise. This is where I report from the field: what I’m watching, what’s surprising me, what the AI press won’t touch, and what I think it all means for you. Field dispatches, hot takes, and the occasional thing you probably can’t stop thinking about too.
Some of the latest Robots are talking pieces (check out the full Robots are talking pieces archive):
A journey on the other side of the tech bubble — Four real stories from outside the tech bubble that changed how I build AI systems, and four systems you can steal this week.
Four things Moltbook AI agents reveal about themselves and your workflows — Audience performance, identity amnesia, compounding drift, and the shared substrate problem.
🤖 What’s next?
They used to say the dog ate your homework.
Around here, the robots eat it. And trust me, they are starving for the commodity parts of the job that you are tired of doing.
New here? Start with the guide hub or New to RAMH?
Ready for the systems? Check out RobotsOS (the premium tier with pre-built skills, agents, playbooks, and deeper writing).
Want to try something free right now? Download the Voice Profile Builder and build your voice in 20 minutes.
You are already here. Let’s get to work.
With creativity and love,
Chief 🤖 at ROBOTS ATE MY HOMEWORK