hr-studioAI · 03
A computational design primer · Diffusion

Order, conjured from noise.

Diffusion is the engine behind today's image and video models, and its idea is almost cheeky: take any design and slowly dissolve it into pure static — then learn to run that dissolve backwards. Start from a screenful of noise, nudge it toward where real designs live, again and again, and a form precipitates out of nothing. Below it runs for real, with the exact denoising direction a perfectly trained model would learn.

Scroll to begin
i.

The forward dissolve

Start with a shape made of points and keep adding Gaussian noise. A little blurs it; a lot erases it. Push far enough and every design — any design — melts into the same featureless cloud of static. This direction is trivial: no learning, no choices, just noise piling on noise.

Drag the noise up and watch the form dissolve. The trick will be coming back.

Add noise · any data → static
ii.

Which way is the data?

To reverse the dissolve you only need one thing: at any noisy point, which way is the nearest real data? That direction — the score — forms a field that everywhere points back toward where designs live. A diffusion model spends its whole training learning this field; here we can draw it exactly.

Move your cursor through the noise and read the arrow. It always points home.

The score field · arrows toward the data
iii.

One point walks home

Generation is just following that arrow, patiently. Drop a point anywhere in the static, step a little along the score, lower the noise a notch, and repeat. Step by step it threads its way down out of the chaos and lands on the form — a single denoising trajectory.

Click anywhere to release a point and watch it find its way to the shape.

Click to drop a point · it denoises home
iv.

Noise becomes form

Now do it to a whole cloud at once. Begin as pure static, take every point one denoising step, lower the noise, and repeat — the swarm flows inward, warm chaos cooling into crisp form. This is sampling: hundreds of trajectories descending together until a design stands where there was nothing.

The same loop, scaled up, is how an image appears from a prompt.

Reverse diffusion · the swarm condenses
v.

The diffusion studio

Hand it any target and it will generate it. Pick a form and watch noise resolve into it — or switch to draw mode, scatter your own constellation of points, and set diffusion loose on a distribution that didn't exist a second ago. Because the denoising direction is computed exactly from whatever target you give, there's no training wait: sketch, and the static reassembles into your shape. This is the whole modern playbook in miniature — define a distribution, learn its score, run the noise backward.

Pick a target or draw your own · then generate
Dissolving a design into noise is free; the magic is the arrow that points back. Learn that arrow at every noise level and you can summon form from static on demand — which is, almost exactly, how the kings of generative AI paint.