A Beginners Guide to Generative AI Art

Generative AI Art is revolutionizing the way we create and interact with digital content. offering endless possibilities for generating images from textual descriptions. In this guide, we’ll delve into how beginners can get started with popular platforms like DALL-E, Midjourney, and Stable Diffusion.

Introduction to Generative AI Art Platforms

Dall-E: Created by OpenAI, the creators of Chat-GPT, Dall-E can be used using OpenAI´s API, through Chat GPT on a paid subscription or using Bing Image Generator for free. on Chat-GPT, Dall-E tends to shy away from creating photorealistic images with humans. Particularly when recognisable faces are involved, leaning more towards artsy styles. However on Bing, it appears to be more liberal, creating superhero images with no problems.
Midjourney: Midjourney is another extremely popular Image generation service, similar to Dall-E. However Midjourney is currently only usable via the popular chat app Discord, where users can create images using chat commands and prompts. Like Dall-E it has a paid subscription to take full advantage of all the features. Unlike Dall-E, however Midjourney is excellent at creating photorealistic images, and has no problem generating well known faces.
Stable Diffusion: Stable Diffusion is an Open Source model that has been made available to the public for Free. Due to it´s open source nature many variants of the model exist, as it can be customised and trained to whatever use case you want. As long as you have the technical capacity to host it, set it up, and train it. Many Image creation services use Stable Diffusion as their back end, and the business applications are endless. Furthermore the developers of Stable Diffusion, Stability AI, also have their solution hosted and available for all to use via their own platform, also with a paid subscription. The nature of Stable Diffusion also allows it to be deployed to corporate hardware, and setup so that prompts and information can remain property of the company without sharing it with third parties.

Getting Started with Midjourney

Midjourney requires a Discord account to start creating Generative AI art. Here’s a brief guide to getting started:

Sign Up to Discord: If you don’t have one already, sign up and download the Discord app at https://www.discord.com
Join the Midjourney Channel: Discord uses channels for different communities. To access Midjourney, you need to join the Midjourney Channel, which can be found here. Use this link once you have signed up and installed Discord. Join Midjourney Here
Subscribe: Although Midjourney comes with a free trial, to make the most of Midjourney, a subscription plan will be required. This will give you access to a private bot, allowing you to generate images in Private. Please note, Images generated on the free plan, or on the Midjourney discord channel will be available for all to see.
Understand the Tools: Midjourney works by using /commands along with your prompt to provide additional parameters for the image, including specifying the seed, commands for consistent characters, and much more. Additionally, once the image is created, you can expand on it, by zooming in, out, panning, and more.
Using Commands: Start generating images with commands like /imagine followed by your prompt. Midjourney offers tools for upscaling, regenerating, and creating variations of images.

Tips for Midjourney:

Start Simply: Start with simple prompts and gradually introduce complexity.
Be Concise: Descriptive yet concise prompts tend to yield better results.
Experiment with Styles: Incorporating styles and mediums in prompts can lead to intriguing results.
Combine Concepts: For more targeted results, try combining different themes in your prompts.
Utilize Lighting: Adding lighting effects to your prompts can enhance the dynamism of the output.

Getting Started with Dall-E

Decide on Bing vs Chat GPT: Dall-E is useable from both. However using it in Chat GPT has the advantage of Chat GPT creating the prompts for you, and creating generative AI art within the context of the conversation you are having. As Opposed to Bing, where it will create an image exclusively from the prompt with no additional context. In both cases you will need to create an account or login.
Input Description: Enter a detailed textual description for the image you want to create. Specificity enhances the quality of the generated image.
Generate and Export: After entering your description, generate the image and fine-tune it as needed. Dall-E allows exporting in various formats for use in different applications.
Costs: Although Bing is Free, your request will be in a queue along with other users. However you can choose to purchase boosts, which will allow you to generate images more quickly, ahead of free users. On Chat GPT, DALL-E Is included as part of a premium subscription.

Best Practices for Dall-E:

Be Specific in Your Descriptions: The more detailed your text prompt, the more accurately DALL-E can generate your envisioned image.
- Example: Instead of “a dog,” specify “a golden retriever sitting on a sunny beach with a red frisbee.”
Understand the Limitations: Recognize that DALL-E may not perfectly interpret abstract concepts or generate highly complex images. Dalle-E in chat GPT, will also avoid using copyrighted characters or recognisable faces.
- Example: Asking for “a time-traveling car” may result in an abstract interpretation, reflecting the AI’s creative but non-literal approach to generative AI art.
- Example: Asking Dall-E (Chat-GPT) to create to create Tony stark wearing iron man armour, resulted in chat GPT pushing back with, “can create an image inspired by your description, but I’ll need to modify it slightly for originality. Let’s imagine a futuristic armored character, similar to a superhero, with his helmet off. This character will have a distinct look, not directly resembling any known characters but still capturing the essence of a high-tech armored hero. How does that sound?” Bing Still has copyright protectors but seems to be less stringent.
Use Clear and Concise Language: Clarity and conciseness in prompts prevent misunderstandings by the AI.
- Example: Opt for “a bluebird flying under a rainbow” rather than a long-winded, detailed description.
Experiment with Different Styles: DALL-E can generate images in various artistic styles, so experimenting with style descriptors in your prompts can yield diverse results.
- Example: Compare outputs for “a futuristic digital art landscape” with and without Van Gogh art style included in the prompt to see how DALL-E adapts.
Iterative Process: Use initial outputs as a starting point and refine your prompts for better results.
- Example: If “a cat playing piano” isn’t quite right, refine it to “a cartoon cat playing a grand piano on a stage” for more specificity.
Use Variations and Edits: Utilize the ‘variations’ and ‘edits’ features of DALL-E to explore different interpretations of your prompt or to refine the image.
- Example: For a prompt like “a medieval castle at sunset,” use variations for different artistic takes or edit to “a medieval castle overlooking a river at sunset” for a more specific scene.

Applications of Generative AI Art

Generative AI art in image creation has vast applications:

Creative Arts: Artists use these tools for quick conceptualization and generating visual ideas, blending AI creations with traditional techniques to form unique artworks.
Marketing and Advertising: Marketers leverage AI to produce original visuals for campaigns and social media, enhancing brand identity and engagement with minimal effort.
Education: Educators utilize these platforms to visualize complex concepts, making learning more interactive and accessible for students.
Personal Use: Individuals engage in creative exploration, using AI for personal art projects or just for fun and experimentation.

Platforms like DALL-E and Midjourney are reshaping creativity, offering easy ways to turn ideas into visual forms. As these technologies advance, they promise new avenues for innovation in art, marketing, education, and personal expression.

Are you ready to explore the world of advanced Image Generation and AI solutions for your business?

We offer tailored assistance in creating and implementing AI tools that align with your unique business needs. Our team of experts provides comprehensive training and support, ensuring that you can maximize the benefits of these cutting-edge technologies.

Take the first step towards unlocking the full potential of AI for your business.

Contact us for a consultation, and let’s discuss how we can help you integrate these innovative solutions into your business strategy. Our goal is to empower your organization with the tools and knowledge needed to thrive in the ever-evolving digital landscape.