🤯 Deep Dive: Create STUNNING Images with ChatGPT

Deep Dive into ChatGPT + Dall-E 3

Happy Tuesday and welcome back to How to Actually AI.

Today we’re doing a deep dive into DALL-E 3’s integration into ChatGPT.

What do we like about DALL-E 3 + ChatGPT?

ChatGPT will create VERY detailed image prompts for you. Once you have an image you like, you can simply say “give me 3 more styles” and it will create very detailed image prompts for you.

Examples of these kinds of prompts can be seen down below. The amount of detail is mind blowing.

Read time: 4 minutes.

What is DALL-E?

DALL-E 3 represents the newest advancement in OpenAI's AI-driven art creation tool. As a successor to the acclaimed DALL-E 2, DALL-E 3 elevates the art generation experience with a myriad of enhancements and novel capabilities. We'll delve deeper into these advancements later in the piece.

For those unfamiliar, "DALL-E" refers to a generative AI system that allows users to produce unique visuals based on text descriptions, termed as prompts. These prompts are then translated into visual artworks. At its core, DALL-E functions through a neural network, generating diverse and original images according to the directives provided by the user's prompts.

What is the difference between DALL-E 2 and DALL-E 3?

The two versions of the platform exhibit notable distinctions. DALL-E 3 showcases remarkable advancements in image generation proficiency, accuracy of visual results, and has even incorporated ethical refinements informed by user feedback.

Key improvements include:

  • Improved interpretation of prompts

  • Improved creative capabilities for image generation

  • Ethical awareness

  • Native integration with ChatGPT (see below)

How to Use DALL-E with ChatGPT

  1. Begin by accessing ChatGPT (visit) and logging into your OpenAI account. Ensure you have a subscription to ChatGPT Plus, which is priced at $20 monthly.

  2. Subsequently, navigate to the "GPT-4" model and choose "DALL-E 3" from the available options in the drop-down list.

    1. Example 👇🏻

  3. Input a prompt into ChatGPT to produce images utilizing the DALL-E 3 model.

  4. The system will seamlessly craft intricate prompts with GPT-4 and produce images on the spot.

  5. Click on the images it generates to see the specifc prompts ChatGPT used to generate each photo.

🧠 Go Deeper

Techniques discussed for Using DALL-3

  • Generate Images: Just type in a prompt like "create me a 3D cartoon cowboy riding on his horse," and let ChatGPT do the rest.

  • Modify Images: If you like an image and want to modify it, specify the changes you want. For example, "Make the sun in image one a moon."

  • Add Text: You can add text to your images. Just ask ChatGPT to "add text to this image that reads 'The Wild West.'"

  • Change Resolution: Specify the resolution size you want before typing out your prompt. You can choose from wide, square, or tall formats.

Examples of DALL-E Creations

Below are some example of DALL-E images with prompts. The key things to remember is that ChatGPT created these prompts. 🤯

Prompt: Photo of a neon sign glowing in blue and pink hues, set against a dark backdrop. The bright, vivid letters spell out 'Made with Dall-e' without any additional text below it, emitting a soft glow around the letters.

Prompt: Watercolor painting of a neon sign set against a blended night sky. The words 'Made with Dall-e' shine in vivid blue and pink tones, standing out starkly without any text below.

Prompt: Cartoon-style drawing of a neon sign, characterized by bold outlines and vibrant blue and pink colors. The sign prominently features the words 'Made with Dall-e' against a night sky backdrop without any additional text.

Prompt: Retro-style poster of a neon sign, reminiscent of 1980s aesthetics with a grainy texture and synthwave colors. The words 'Made with Dall-e' glow in neon against a dark gradient background, absent of any secondary text.