AI AdventureAICreate with AIDay 11
🗜️ Create with AI · Day 11

The Secret Small Canvas

Here's a secret that sounds fake but isn't: modern AI art doesn't paint on a big picture. It squishes everything down to a tiny canvas, does all its thinking there, then blows it back up. That squished world is called the latent — and walking through it makes pictures morph. 🌈

↓ start squishing
🗜️ Lab · The Compress-o-Tron

Squish it small. Blow it back up.

Pick a picture. The machine encodes it into a tiny N×N grid (the middle box — that's the whole picture, stored in just a few color blocks!), then a decoder blows it back to full size on the right. Drag the N slider and watch the trade-off: fewer blocks = smaller & blurrier, more blocks = detailed but heavier.

① original
② the latent (8×8)
③ decoded back up
8

grid size (N×N blocks)

✨ smooth decoder — OFF
64blocks stored
25,600original pixels
400×smaller!

The AI doesn't store every pixel — it stores the tiny grid and imagines the details back. That's why it's fast enough to run at all. 💾

🌈 Lab · The Latent Morph

Melt one picture into another

Here's the magic trick. Squish two pictures into latents, then take a step-by-step walk from one to the other. Every in-between step decodes into a real picture — so A smoothly morphs into B. Slide t, or hit ▶ Auto-morph and watch.

🅰️ start
🅱️ end
t = 0.00 (all 🅰️)
0.00
Notice the in-between frames aren't a fade (two pictures ghosting through each other) — the colors themselves slide from one to the other. That's because we're mixing in the latent, the AI's compressed idea-space, not on top of the pixels. 🌀
🏷️ The pro wordsThat tiny squished canvas is the latent space. The squisher is an encoder and the blow-it-back-up is a decoder — together they're a VAE (variational autoencoder). Running diffusion inside the latent instead of on full pixels is called latent diffusion — it's what makes Stable Diffusion fast. Walking between two latents is interpolation.
🕵️ Boss · Morph Detective

How far along is this morph?

I froze a morph at a secret spot between 🅰️ and 🅱️ (shown as thumbnails). Study the mix and set your slider to where you think it stopped — 0 = pure A, 1.00 = pure B. Get within 0.12 and it counts. Crack it 3 times to win the badge!

🅰️
❓ frozen here
🅱️
0.50

your guess (0 = A, 1 = B)

Cracked: 0 / 3
🚀 Beyond the Basics

Why the latent is a superpower

Squishing to a tiny canvas isn't just a speed hack — that compressed idea-space is where almost every cool AI trick happens:

Fast generation

Diffusion on a 64×64 latent instead of a 512×512 image is ~64× less work. This is the whole reason AI art runs on a laptop.

🌈

Morphing & blending

Walk between two latents (like you just did) and you get smooth morphs — the trick behind "AI transition" videos.

✏️

Editing

Nudge a latent in the right direction and you can add a smile, change hair, or swap a season — without repainting a single pixel by hand.

🏷️ Pro names to look upThe compressor is a VAE; the small world is the latent space; the fast method is latent diffusion (the "LD" in Stable Diffusion). Smooth walks use spherical interpolation (slerp). Real latents are 4-channel and ~48× smaller per side.

🎉 Day 11 Complete!

You learned the trick that makes AI art possible on a normal computer: it works on a tiny compressed canvas — the latent — then a decoder blows it back up. And walking between two latents gives you smooth morphs. That little squished space is where all the magic lives.

Day 12 →
Next up: more ways to bend a generator to your will. You've got the dial, the negative prompt, and now the secret small canvas. 🚀
AI Adventure · Create with AI · Day 11 — made for young inventors 🚀