Reviews from businesses like yours — matched by size and type. Available once you create a profile.
Complete profileRelated Tools
Community Reviews
Reviews from Reddit, Product Hunt, Hacker News & AlignAI members
people have built businesses around... Stable Diffusion
Hacker News • HackerNews • Use case: Running model weights locally to build businesses
With stable diffusion it's a lot of fiddling and putting manual emphasis on certain keywords until just getting it right.
Hacker News • HackerNews • Use case: Imaginary concept project for character designs (TMNT remake), using inpainting and img2img
Every other player:... Stability.ai s Stable Diffusion... are all on the path to extinction.
Hacker News • HackerNews • Use case: Image generation
Stable Diffusion and local models are still way too hard for 99.99% of people and will never see the same growth as a Midjourney or OpenAI that have zero sharp edges.
Hacker News • HackerNews • Use case: Local AI image generation for general users
You can do much more controllable inpaints with the latter.
Hacker News • HackerNews • Use case: Controllable image inpainting/generation
Stable diffusion does this quite handily.
Hacker News • HackerNews • Use case: Generating an image of a pelican on a bicycle from a text prompt
It will, however, definitely not affect the more-common use case of anime women with very large breasts. And people will be able to finetune SD 2.0 on NSFW images anyways.
Hacker News • HackerNews • Use case: Generating images while attempting to filter out NSFW content
that can be used to trained well performing models (eg openclip and stable diffusion).
Hacker News • HackerNews • Use case: Training a well-performing model using the dataset.
Since SD is trained by gradient updating against several different images at the same time, it of course never copies any image bits straight into it.
Hacker News • HackerNews • Use case: Analyzing its architecture and training process in the context of a copyright lawsuit.
For SD 2.X, OpenCLIP passes a (77, 1024) matrix.
Hacker News • HackerNews • Use case: Understanding the updated text encoding mechanism for image generation
For SD 1.X, CLIPText encodes a prompt and passes a (77, 768) matrix to the core UNet.
Hacker News • HackerNews • Use case: Understanding how text prompts are encoded for image generation
Stable Diffusion for example used a pre-trained CLIP transformer network from OpenAI (and subsequently OpenCLIP). CLIP can have internal associations between words that in turn steer the diffusion image generation.
Hacker News • HackerNews • Use case: Image generation, understanding how its training on images and text (via CLIP) allows it to generate concepts not explicitly in the image training set.
Stable Diffusion does not take or copy her characters - it mimics her drawing style in colors, hues (it reads what pixels are proximate to what other pixels and tries to apply that proximity to similar colors, line patterns in the future).
Hacker News • HackerNews • Use case: Generating art by mimicking an artist's style (colors, hues, line patterns) from training data
as with Stable Diffusion before it, it quite often produces stuff you just don't want. Junk dealers don't care about that and use the first result, an artist can afford to do 10 attempts and pick the right one.
Hacker News • HackerNews • Use case: Image generation (implied from context)
I'd love to use able to use stable Diffusion, but the learning curve, to be honest, intimidates me a bit.
Reddit • Entrepreneur
Stable diffusion could do this with imgtoimg or you could build a Lora model of the logo/product.
Reddit • Entrepreneur • Use case: Image-to-image generation or building Lora models for logos/products
Have You Used This Tool?
Sign in to share your experience with this tool.
Sign in