lexica walkthrough + tutorial // creating ai-generated art with stable diffusion

kelly kim
16 Dec 202208:55

TLDRJoin the tutorial to explore Lexica, an AI-powered search engine for art creation. Think of it as Google for AI art, where you input a prompt and receive images based on your description. The platform offers a gallery of AI-generated creations, where users can like, explore, and even use seeds to recreate similar works. The guidance scale allows for varying levels of accuracy to the prompt, offering diversity in output. Learn how to generate your own artwork with tips from the community, starting with a medium like oil painting, followed by a subject, and adding details like style and setting for a unique piece of AI art. The tutorial also covers reverse image search and provides tricks for creating stunning AI-generated art that can be used for various purposes, such as Christmas cards or wall decorations.


  • 🎨 Lexica is an AI tool that functions like a search engine for AI-generated art, allowing users to input prompts and receive images based on those prompts.
  • 📈 Users can subscribe, like, and share their favorite AI-generated images, similar to social media platforms like Pinterest.
  • 🌟 The 'seed' is a unique number generated each time an image is created, which can be used to recreate similar images with variations.
  • 📊 The 'guidance scale' determines how closely the generated image adheres to the input prompt, with lower values offering more creative freedom.
  • 🖼️ The interface allows for the creation of art by specifying a medium, subject, and additional details, such as style or setting.
  • 🔍 Reverse image search is possible, where users can upload an image to find similar AI-generated content.
  • 🍞 Examples given include generating images of mundane objects like bread, as well as more specific prompts like characters from the show 'Peaky Blinders'.
  • 🎭 The process of creating AI art involves experimenting with different prompts and settings to achieve desired results.
  • 📚 Learning from community resources, such as blog articles, can provide valuable insights and techniques for generating better AI art.
  • 🧠 AI is not fully developed, so starting with broader prompts can lead to more diverse and inspiring outcomes.
  • 🐕 A personal example demonstrates creating an oil painting of a Boston Terrier in a psychedelic garden, inspired by the artist Fernando Botero.
  • 💌 The presenter encourages viewers to create their own AI art and share it on social media or via email for feedback and appreciation.

Q & A

  • What is Lexica and how does it function?

    -Lexica is referred to as a stable diffusion search engine, which can be thought of as a Google for AI art. It is an engine that takes a prompt from the user and generates an image based on that prompt.

  • How can users interact with the creations on Lexica's homepage?

    -Users can interact with the creations by liking them, similar to how one would on social media platforms like Pinterest. They can also explore the style and deconstruct the prompts used to generate the images.

  • What is the significance of the 'seed' in the context of Lexica?

    -The 'seed' is a uniquely generated number that is created when an image is made. It can be used to generate a similar image with alterations, providing a way to recreate a style or theme with variations.

  • What does the 'guidance scale' control in the image generation process?

    -The 'guidance scale' determines how closely the generated image adheres to the input prompt. A higher guidance scale makes the image more accurate to the prompt, while a lower scale allows for more diversity and creative freedom.

  • How does the 'reverse search' feature work on Lexica?

    -The 'reverse search' feature allows users to upload an image and search for similar images or styles. It's a way to find more of the same kind of artwork or to explore variations on a theme.

  • What are some tips for creating AI-generated art on Lexica?

    -When creating AI-generated art, starting with a medium (like oil painting), then specifying the subject (like a Boston Terrier), adding details (like the style of a favorite artist), and finally, adding more context (like a psychedelic garden) can help generate more personalized and interesting artwork.

  • How does the speaker suggest users approach the guidance scale when first starting out with Lexica?

    -The speaker suggests that beginners start with a lower guidance scale to allow for more diversity and inspiration, as AI is still developing and being more open to broader interpretations can lead to more creative results.

  • What is the purpose of the 'negative prompts' feature?

    -Negative prompts allow users to exclude certain elements or themes from the generated image. For example, if a user doesn't want a croissant in a bread-themed image, they can include 'croissant' as a negative prompt.

  • How does the speaker describe the process of generating AI art?

    -The speaker describes it as a fun and creative process, where users can input a description, use negative prompts to exclude unwanted elements, set the dimensions, adjust the guidance scale for creative control, and experiment with different styles and mediums to create unique pieces of art.

  • What does the speaker suggest as a potential use for the AI-generated art pieces?

    -The speaker suggests that the AI-generated art pieces could be used as Christmas cards, e-cards, or even enlarged and displayed on walls at home due to their high quality and aesthetic appeal.

  • What is the speaker's invitation to the audience at the end of the tutorial?

    -The speaker invites the audience to create their own AI art, share their creations on social media platforms like Twitter and Instagram, and tag the speaker, or even email their artwork for feedback and recognition.

  • How does the speaker describe the variety of AI-generated images on Lexica's homepage?

    -The speaker describes the variety of AI-generated images on Lexica's homepage as 'fabulous creations' made by people using AI, highlighting the platform's ability to produce a wide range of artistic styles and themes.



🎨 Introduction to Lexicon: The AI Art Search Engine

The video begins with the host expressing gratitude for joining the tutorial on an AI tool called Lexicon, described as a 'Google for AI art.' The host outlines the plan to explore Lexicon's search and generation functions and encourages viewers to subscribe, like, comment, or share if they are interested in AI technology. The homepage of Lexicon is displayed, showcasing AI-generated art creations by users. The host discusses the features of the platform, such as the ability to like creations, view prompts that led to the generation of images, and use seeds to recreate similar images. The concept of a guidance scale for creating artwork is mentioned, and the host expresses excitement to demonstrate the platform's capabilities.


🔍 Searching and Creating with Lexicon

The host demonstrates how to use Lexicon's search function by looking for AI-generated images of sourdough bread and the TV show 'Peaky Blinders.' They also show how to perform a reverse image search using a photo of the actor Cillian Murphy. The process of generating artwork is then covered, explaining the interface and the importance of the guidance scale for determining the accuracy of the generated image to the prompt. The host shares a formula for creating AI art, which includes selecting a medium, subject, and additional details like style and setting. They also mention learning from the build space community and experimenting with different prompts to create unique AI-generated art. The video concludes with the host's excitement about the generated Boston Terrier painting and an invitation for viewers to create their own AI art and share it with the host on social media.



💡Stable Diffusion

Stable Diffusion is an AI tool that functions as a search engine for AI-generated art. It allows users to input a prompt, which can be a description or an idea, and the engine generates images based on that prompt. In the video, it is likened to Google for AI art, showcasing its ability to create a wide range of artistic outputs.


A prompt in the context of the video is a text input that guides the AI in generating an image. It can include specific details, styles, or themes that the user wants to be reflected in the AI-generated artwork. For example, the script mentions 'cinematic lighting' as a prompt to influence the style of the generated image.


The seed in the video refers to a unique number generated each time an image is created with Stable Diffusion. It serves as a starting point for the AI's image generation process. Users can use this seed to recreate similar images with minor alterations, providing a level of control over the creative output.

💡Guidance Scale

The guidance scale is a parameter within the AI tool that determines the level of accuracy with which the AI adheres to the user's prompt. A higher guidance scale means the generated image will closely match the prompt, while a lower scale allows for more creative freedom and diversity in the output. In the video, the creator adjusts this scale to find a balance between control and creativity.


Dimensions in the context of the video refers to the size and proportions of the generated image. Users can specify the dimensions to suit their needs, such as choosing a more horizontal or vertical orientation. This feature is important for users who have specific display or printing requirements for their AI art.

💡Cinematic Lighting

Cinematic lighting is a style of lighting used in the film industry to create a certain mood or atmosphere. In the video, it is used as a prompt to guide the AI in generating images with a similar lighting style, enhancing the visual appeal and emotional impact of the generated art.

💡Reverse Search

Reverse search is a feature that allows users to upload an image and find similar images or styles within the AI tool's database. This can be useful for exploring a particular theme or aesthetic further or for finding inspiration based on existing artwork or photographs.

💡Generate Function

The generate function is the process within the AI tool where the user inputs a description or prompt and the AI creates new artwork based on that input. It is a core feature of the tool that enables users to produce unique pieces of AI-generated art without traditional artistic skills.

💡Arts Nouveau

Arts Nouveau is an international design movement that emerged in the late 19th and early 20th centuries, known for its curvilinear forms and organic patterns. In the video, it is mentioned as a style that can be incorporated into the AI-generated art, demonstrating the tool's ability to emulate various artistic styles.

💡AI Art

AI Art refers to artwork that is created or influenced by artificial intelligence. In the video, the host uses the AI tool to generate various pieces of art, showcasing the potential of AI to assist in the creative process. The generated art can be used for personal enjoyment, as digital decorations, or even as physical prints.

💡Social Media Integration

The video mentions the ability to 'like' creations within the AI tool, similar to social media platforms like Pinterest. This feature allows users to interact with and appreciate each other's AI-generated art, fostering a community around the shared interest in AI and art.


Introduction to Lexica, a stable diffusion search engine for AI-generated art.

Lexica allows users to input prompts to generate AI art.

The platform features a search function to explore AI art created by others.

Users can like and explore the creation process of AI art on the platform.

Each AI art piece has a unique 'seed' for generating similar images.

The 'guidance scale' determines how closely the generated art adheres to the input prompt.

Different dimensions and iterations of AI art can be explored.

The ability to search for AI art using text prompts or reverse image search.

Creating AI art with a personal touch by generating custom images.

The interface for generating AI art includes options for description, negative prompts, and dimensions.

Adjusting the guidance scale for more diversity in generated art.

Experimenting with different settings to create unique AI art pieces.

Combining mediums, subjects, and details to generate complex AI art.

The tutorial demonstrates creating an oil painting of a Boston Terrier in a psychedelic garden.

Incorporating styles like 'cinematic lighting' and 'Arts nouveau' into AI art generation.

The potential for using AI-generated art for various applications like Christmas cards or wall art.

Invitation for viewers to create their own AI art and share it with the presenter.

The presenter's enthusiasm for exploring and utilizing AI art tools like Lexica.