Creative Intelligence

• One-week intensive

You didn't choose what showed up in your feed this morning. Something else did — and it knew exactly what it was doing. A lot of what it showed you? Made up.

Two systems are quietly shaping what you think without asking permission: the recommendation algorithms that have been profiling you since middle school, and the AI tools that confidently fabricate things and never apologize for it. They're built to be irresistible. This week, you learn how to run the game instead of being played by it.

We start by reverse-engineering your own data profile — what does the algorithm think you want, and why is it trying to make you angry? Then we turn to the content itself: real photo or AI-generated? Breaking news or deepfake? You have blind spots. We find them — and we find them using examples where AI fabrication caused real-world consequences, not just inconvenience.

From there, you'll build your own convincing piece of fabricated content and pitch it to the room while they try to catch you. You'll briefly try to completely derail your actual feed — just to prove you can run the machine when you choose to.

You leave with

A Personal BS Detector Field Guide and a Terms of Service for Your Own Brain — exactly when AI gets to help you, when the phone goes away, and what you're not handing over to anyone.

Led by

Kyle Homstead →

Every Minecraft world you've ever played in lived on someone else's server. This class puts the entire world in the palm of your hand.

Students install and configure a live Minecraft server from scratch — the runtime underneath it, the server software that manages the world, the configuration files that decide who gets in and what they're allowed to do. You'll see what's actually happening when a player connects: the handshake, the authentication, the chunk data loading for their coordinates specifically. Then you become the operator — adjusting permissions, managing the whitelist, running the server console like someone who knows what those commands actually do.

By the end of the week, your server is live. You picked the world seed, you control the whitelist, and your name is in the ops list. It's yours to take with you on a USB stick, so you can install it at home and keep your world alive.

You leave with

A live Minecraft server that's yours to keep running — your seed, your whitelist, your name in the ops list.

Signature

• Stand up a live Minecraft server from scratch — and run it as the operator

Led by

Aiden Chappuis →

• One-week intensive

Ventriloquists knew something computer scientists would spend decades debating: we're hardwired to see minds where there aren't any. From ancient shadow plays to Japanese Bunraku, from Jim Henson building characters out of felt and foam to the AI chatbots you use every day — humans keep trying to put something inside a thing and make it feel inhabited. This course is about how that actually works.

Students design a character — someone with a name, a history, a point of view — and give it two bodies. The first is physical, built by hand from materials in the studio (no experience required; the object can be as simple or as strange as the character demands). The second lives inside an AI agent you write and design from scratch, translating your character's personality into the choices and language that make AI respond the way it does. Neither version has anything going on inside it until you put it there — and working through both is what makes that visible.

The week closes with a live presentation filmed by the LightHouse video team. Performing is one option; walking through your process is another. What it produces isn't a theater piece — it's a point of view: a real understanding of how personality gets built into technology, and what it means to be the person doing the building.

You leave with

A character with two bodies — a hand-built physical object and an AI agent you designed — and a real grasp of how personality gets built into technology.

Signature

• Build one character in two forms — a physical object and an AI agent you write
• A live presentation filmed by the LightHouse video team

Led by

Kyle Homstead →

You've proven you're not a robot a thousand times — clicking crosswalks, traffic lights, blurry buses. This week, prove it for real.

Nobody on earth has your exact collection of words: the phrases your family repeats, the vocabulary of the things you love, the sentence stuck in your head since fourth grade. Monday through Wednesday, we'll write together — gathering the raw material of your one-of-a-kind voice, trading it with other writers, and doing something most schools won't let you do: collaborating directly with an AI to bend, break, and build poems and stories that didn't exist when you walked in. You'll discover firsthand what AI can do for a writer with a strong voice — and what it can never do without you. No formulas, no five-paragraph anything, no experience required.

Then Thursday is yours. Perform your piece out loud. Publish it — submit to a journal, print a zine, post it to the world. Share it with friends and family. Or leave with a working draft you can't wait to keep building. You choose what your writing is for, and you'll shape your project around that goal all week.

You leave with

A finished piece you set the purpose for — performed, published, or ready to keep building — and a real sense of what AI can and can't do for your writing.

Signature

• Collaborate directly with AI — and find what it can never do without you

Led by

David Lane →

A hundred years ago, an artist bought an ordinary object from a hardware store, signed it, and called it art — and the art world has never recovered. This week, you join that troublemaking tradition.

You'll learn the art of finding: pulling small sculptures out of junk drawers, thrift bins, and hardware scraps with basic hand tools, and pulling poems out of pages that were never meant to contain them — newspapers, manuals, old letters — by blacking out everything except the hidden message. No artistic skill required. The material is everywhere. The skill is noticing.

Your collaborator is an AI with a peculiar condition: it has read nearly everything and touched absolutely nothing. It will riff on your objects brilliantly, flood you with titles, and hand you endless raw text to mine — but it has never wandered a sidewalk, never felt the weight of a rusted key, never picked one thing over another because it gave them a feeling. Finding, it turns out, is the one art it can't commit. You'll spend the week discovering exactly where the machine helps you and exactly where it goes blind — and making art out of the difference.

On the final day, we open our gallery: sculptures and poems built entirely from things the world threw away, made by the only artists in the room with hands.

You leave with

A gallery of sculptures and found poems made from discarded things — and a clear sense of where AI helps a maker and where it goes blind.

Signature

• Finding is the one art the machine can't commit

Led by

David Lane →