We’ve been discussing DALL-E, OpenAI’s text-to-image generative AI tool, for quite some time now. It’s been available via its web interface at labs.openai.com. However, OpenAI recently made it accessible from within ChatGPT. I decided to try it out and see what it can do.
Enabling DALL-E 3 within ChatGPT Plus:
To use the DALL-E feature, you need to have ChatGPT Plus. It is currently in beta and can be selected from the GPT-4 drop-down menu. Enabling DALL-E removes access to Browse with Bing, ChatGPT Plugins, and Data Analytics.
Using DALL-E:
To generate images, you need to provide clear prompts similar to ChatGPT. If you have a specific style in mind, you can provide stylistic information. In my first prompt, I asked for a photorealistic Mac mini on a marble pedestal. I received four renditions and selected the third one, which resembled a typical Mac mini.
Generating Laptops:
After successfully generating a Mac mini, I attempted to generate MacBook Pros but received a “no Macs for you” response from ChatGPT. So, I tried asking for laptops without any brand pedigree and got a set of images, including two laptops from the year 1900.
Using DALL-E for Presentations:
DALL-E in ChatGPT is not as powerful as the standalone version and (*3*). It doesn’t allow image uploads, panning, or selective editing. However, it is convenient for quickly generating images for presentations. The images may not be completely realistic but are suitable for PowerPoint use.
Closing Thoughts:
Getting started with the DALL-E add-on in ChatGPT is easy and recommended. While I will continue using my (*3*) account for advanced features, I will use DALL-E in ChatGPT for quick image generation. If you have experience with text-to-image tools, share your thoughts in the comments.
