Stable Diffusion
An open-source generative AI model that creates images from text prompts; widely fine-tuned by the community.
As defined in Pieter Levels
Other definitions
A type of generative AI model that produces images from text prompts.
Grant Sanderson (@3blue1brown) — Past, present, & future of mathematics
Where it comes up · 8
- Pieter Levels4 claims
- Moment of Zen w/ Marc Andreessen on SBF/FTX, Twitter, and Open AI/Chat GPT2 claims
- The Future of Knowledge Graphs in a World of Large Language Models2 claims
- Generative AI Is About To Reset Everything, And, Yes It Will Change Your Life | Forbes2 claims
- You need to learn AI in 2024! (And here is your roadmap)2 claims
- Grant Sanderson (@3blue1brown) — Past, present, & future of mathematics1 claim
- Pat Grady - Relentless Application of Force - [Invest Like the Best, EP.378]1 claim
- “Deep Thoughts”: Tyler Cowen on AI and the Future of Work1 claim
Most important claims · 10 of 15Search all claims for “Stable Diffusion” →
- Research on how knowledge is stored in large language model parameters is beginning to understand why models must be so large; Stable Diffusion 1 has only 890 million parameters (200x smaller than GPT-3's 175 billion) because it doesn't need to internalize knowledge of individual entities.The Future of Knowledge Graphs in a World of Large Language Models
- Copyright and intellectual property issues remain unresolved in generative AI: models are trained on massive internet-sourced datasets without compensating creators, as exemplified by Getty Images suing Stability AI for training on millions of Getty photos without payment.Generative AI Is About To Reset Everything, And, Yes It Will Change Your Life | Forbes
- The current generative AI wave has been building for several months, reaching an inflection point around summer 2022, with companies like OpenAI (DALL-E 2), Stability AI (Stable Diffusion), and others releasing text-to-image models that achieved viral adoption and mainstream breakthrough.Generative AI Is About To Reset Everything, And, Yes It Will Change Your Life | Forbes
- Detecting and understanding things in images (image-to-object task) is the inverse of image generation (image production); Segment Anything inverts the generative direction by taking an image and extracting structured information from it.You need to learn AI in 2024! (And here is your roadmap)
- Stable diffusion and related image-generating programs can produce 'quite interesting, quite vivid visual images' from prompts and have become 'democratized much more quickly than people expected,' alongside rapid GPT-3 democratization.“Deep Thoughts”: Tyler Cowen on AI and the Future of Work
- Rather than manually guessing which model parameters produce better AI photos, you can randomly A/B test parameters on a small percentage of real users, measure which outputs they favorite or download, test for statistical significance, and roll out the winning parameters to everyone—letting users do the search through parameter space.Pieter Levels
- Open-source and decentralized AI development will naturally emerge from the 'bazaar' model (opensource community building) rather than the 'cathedral' model (corporate controlled), and if that happens, proprietary AI companies will benefit for free from open-source advances.Moment of Zen w/ Marc Andreessen on SBF/FTX, Twitter, and Open AI/Chat GPT
- In AI fine-tuning of faces, how good your training data is matters more in many ways than how many steps you train for; the photos should be diverse in everything (lighting, clothing, location) except the face, because whatever stays constant across photos becomes the trained concept.Pieter Levels
- Stable Diffusion represents the 'best case scenario' for open-source AI—democratized training, tractable economics, and open access to large datasets—and should be replicated for language models to prevent monopolizationMoment of Zen w/ Marc Andreessen on SBF/FTX, Twitter, and Open AI/Chat GPT
- Photorealistic fine-tuned Stable Diffusion models became realistic largely because they were trained on porn, which gave them accurate human anatomy, so AI imaging startups must actively prompt out nakedness and add NSFW detection to make them safe for users.Pieter Levels