CC Fest · Creative AI / ML camp · free & virtual

What is the machine actually doing?

A field manual for taking generative systems apart — how they write, imagine, and move. Serious enough for educators, artists, and researchers to bring real questions; accessible enough that beginners can enter and skeptics can opt out of direct use.

Not a prompt-engineering webinar — an investigation into what the machine is actually doing. Browser tools, session plans, and worksheets — no coding, no accounts, no live AI needed.

Three Saturdays · July 11 / 18 / 25 · 9–11 am PT · dates & logistics →

Recorded · async-friendly · global time zones welcome — can't make 9am PT? the async route is a full path →

Tool principle — every tool makes something invisible visible.

campaign board · live previewgenerating · temp 0.9
Learnp=.41 ingp=.87 Machp=.34 inesp=.92

a free creative AI camp — text, images, video — for educators, artists, students & curious learners.

Fig. 01 — Textone token at a time

Text is sequential.

Fig. 02 — Imagesrelationships across space

Images are spatial.

Fig. 03 — Videocoherence over time

Video must stay coherent.

3+Core sessions + studio
25Launch-ready tools
3Media taken apart
5Participation pathways

§ 00 · Start here

three quick starts — full route guide · five ways to take part →

§ 01 · Sessions

predict → change one thing → document what moved

Three core sessions plus optional studio. One investigation arc.

campaign signal · modality arc

Text becomes pixels becomes frames becomes evidence.

The same investigation loop moves through each medium: predict, change one thing, compare what moved, then name the human decision.

§ 02 · Tool index

browser-based · inspectable · no required accounts
tool index · machine in motion

Small machines, each exposing one mechanism.

Filter by modality, then inspect the visible thing: a token distribution, a denoising path, a drifting frame, or a documented claim.

Start with the session essentials.

The small set of tools featured live in the Zoom sessions. Mechanics curious? The row below names the under-the-hood tools for tokenization, diffusion, CFG scale, and latent space.

Open full tool index →

§ 03 · Method

the habit students carry out of the camp

Treat every tool like an experiment.

Not “play with the AI.” Investigate it. The loop is the same in every session, across every modality.

Naming rule — not “it was weird.” Name the specific behaviour: a default, a failure, a pattern.
1

Predict

Before running anything, write down what you expect to happen.

2

Change one variable

Adjust one setting, prompt, or input. Hold everything else still.

3

Compare the evidence

What actually appeared? How does it match your prediction?

4

Name what the machine did

Default, failure, or pattern — say it precisely.

5

Decide what the human does next

Revise, reject, document — or turn it into a project or lesson, and loop back.

§ 04 · Pathways

every activity, five ways in

Five ways to take part

The same five verbs used everywhere in the camp — the quick starts above are three of them. Opting out of direct AI use never means opting out: see the No-AI pathway for the full route.

01

Use

Try a tool directly and document what happens.

02

Observe / Critique

Analyse pre-generated examples without logging into a tool.

03

Teach / Design

Create a lesson, worksheet, or facilitation plan.

04

Build / Code

Make an explainer or a small interactive tool.

05

Critical / No-AI

Write a critique, consent checklist, model card, or unplugged activity.

The 2026 cohort is formed.

Three Saturdays · July 11 / 18 / 25 · 9–11 am PT · virtual · recorded and async-friendly. Join the waitlist / mailing list for recordings and the next cycle.

§ 05 · Materials

for facilitators and participants

AI use in development

Learning Machines was developed with AI assistance for planning, coding, copy drafting, research, and critique. AI-assisted workflows also helped organize session notes, transcripts, chat exports, and aggregate program data. Public recaps, quotations, attributions, and findings are reviewed by the facilitator against source records. Most interactive instruments on this site are authored browser simulations, not live machine-learning models. Genuine model outputs are labeled with the source information that was preserved; incomplete provenance is named rather than guessed. External services process submitted material under their own terms and account settings. The colophon explains the full labeling standard.

What is not AI. Most Learning Machines instruments do not run machine learning. They use authored data, deterministic JavaScript, rules, illustrations, and human input.

Some parts of the site use ordinary web services: Google Forms collects submissions, GoatCounter provides basic traffic counts, and a few activities retrieve Wikipedia content. Notes and preferences may also be stored locally in the browser. None of these features generates content with AI; each has its own data and privacy considerations.

Classroom origins

Learning Machines grew out of Generative AI as a Creative Collaborator, a course developed by Saber Khan and Danny Gámez at Campbell Hall School.

Consent & context

Public sharing should follow the Consent Protocol. For the full frame, read the Project Brief or visit the original CC Fest Coding Camp.