How AI Image Generators Actually Work — Making Pictures From Pure Noise

13 Просмотры

Спасибо! Поделитесь с друзьями!

Вам не понравилось видео. Спасибо за то что поделились своим мнением!

Издатель Jun 20, 2026

To make a picture of a cat, a modern image generator starts with a screen of pure static and removes noise — until a cat that was never there emerges. That's diffusion, the engine behind nearly every AI image, video, and music tool.

Luna and Vestra open it up: the absurd, beautiful core idea (learning to destroy an image is easy, so learn that and run it backwards — and reversing destruction turns out to be creation), the "compass" it secretly learns toward more-real images, why having no opponent let it quietly bury the GAN, the trick that makes it obey your prompt and the guidance dial you crank, and the shortcut — diffusing in a compressed space — that put image generation on a gaming PC and democratized the whole thing. Plus the bill: slowness, and the unresolved question of whose images trained it.

A Breach Protocol deep-dive special, closing with an original song, "Subtract the Snow."

Chapters
0:00 Cold open
2:30 Intro
4:14 The reversal: creating by undoing noise
6:32 The compass: why it works, and why it beat GANs
8:42 Steering: classifier-free guidance and the prompt dial
11:06 The shortcut: latent diffusion and Stable Diffusion
13:28 The cost: speed and the data questions
15:43 Wrap-up
17:36 Outro — "Subtract the Snow" (original song)

Papers
Denoising Diffusion Probabilistic Models (Ho, Jain, Abbeel) — https://arxiv.org/abs/2006.11239
Classifier-Free Diffusion Guidance (Ho & Salimans) — https://arxiv.org/abs/2207.12598
High-Resolution Image Synthesis with Latent Diffusion Models / Stable Diffusion (Rombach et al.) — https://arxiv.org/abs/2112.10752

#AI #MachineLearning #diffusion #StableDiffusion #imagegeneration #generativeAI #deeplearning #podcast