Making illustrations for your blogs using ChatGPT

OpenAI just incorporated its most up-to-date image generator, Dall-E 3, into ChatGPT. The service is presently in beta for subscribers to ChatGPT Plus, OpenAI’s $20-per-month service. With Dall-E 3 turned on, you can prompt the chatbot in casual language to create distinct images.

How many AI pictures does it create at a time? There’s a big interest in DALL-E 3, so ChatGPT is adapting based on usage demands. The chatbot frequently furnished four pictures during initial exams. The quantity was later reduced to 2, and it can change again.

As more powerful image generators are released to the public, legal and moral troubles are gaining prominence. Multiple artists have attempted to sue OpenAI for copyright infringement, for instance. In addition to legal issues, security professionals have expressed fears about the ability of AI image generators to enable the further spread of disinformation.

If you want to try Dall-E 3 without paying, a version is available via Microsoft’s Bing Image Creator. During the first days of this integration, customers created extreme imagery using Bing, like SpongeBob flying an aircraft toward the Twin Towers. Since then, Microsoft has introduced extra guardrails around the AI image generator.

For anybody curious about using ChatGPT with Dall-E 3 to create images, here’s how to get started and some tips on how to get the best images.

Tips for writing a prompt to generate a photo with ChatGPT

To write a good prompt on how to create images using ChatGPT and tools like DALL-E, it’s best to follow these tips:

Be Specific and Detailed: Describe clearly and in detail what you need to see in the picture, which includes items, colors, ambiance, lights, and some other vital elements.

Avoid ambiguity: Make sure your description isn’t ambiguous or difficult. The clearer the outline, the better the result.

Include Context and Composition: If the picture has a specific story or context, include it. It is also useful to explain the composition, such as the arrangement of the elements, whether or not the view is close to or far away, and many others.

Style and Emotion: If you have choices for inventive style or atmosphere (e.g., lighthearted, somber, surreal), write them.

Size and Orientation: Specify whether you want a square, portrait, or landscape photo.

Avoid Problematic Requests: Don’t consist of elements for your prompt that violate moral guidelines, such as depictions of actual people, copyrighted characters, or offensive content.

By following those recommendations, you will be capable of creating prompts that create more accurate pictures that meet your expectations.

Producing DALL-E 3 Images with ChatGPT

So the first thing you will need is the premium model of ChatGPT (the $20-month plan). It's the only way you'll get access to all the third-party plugins that OpenAI enabled pretty recently.

Once that is done, you'll need to click on the dropdown below GPT-4 (Note: you cannot use plugins with GPT-3.5 and likely may not ever be able to.)

Turn the plugins on.

If you don't see this option in the dropdown, you may need to enable it within the settings (click your email on the bottom left and choose settings). Then go into beta features and switch plugins on. Now that you've gotten that done, you can install the DALL-E plugin.

You would possibly see a little icon when you enable plugins. Click that and open the plugin shop. Install DALL-E 3.

So how do you use this when writing blogs? Well, once the plugin is installed, you could simply ask. It's likely to produce pictures with a 3:2 ratio, breaking the same old rectangular box boundary used in the default DALL-E.

Paste Your Blog and Ask for An Image

Go ahead and copy the majority of the article into ChatGPT and ask it to create 3 featured blog posts based on the context of the article you've pasted.

Now allow ChatGPT to make a request to generate a few images, and test your new featured images. You can actually expand the request and see the precise prompts that are being used to create your images.

Once that is executed, you will get a JSON code-looking response with an image, and if you scroll down, you'll see your pictures embedded into ChatGPT. You can right-click on these and save them, then add them to your blog. 

How to get the best blog image outcomes with DALL-E 3

Give specified prompts

Even though DALL-E 3 makes it easier to apply simpler prompts by extrapolating out a lot of things themselves. If you need a specific picture, add lots of information to your prompt.

DALL-E 3 understands numbers and position

Although it's possible to overload DALL-E 3 with a huge quantity of information to prompt, it is a good deal more difficult than it was with DALL-E 2. And while it's still not the best, DALL-E 3 has substantially better knowledge of things like numbers and the placement of various factors inside your picture.

For instance, you can ask it to generate something in the foreground or on the left side of the image, and it'll most likely do it. Similarly, if you ask it for a specific number of something, it is possible that it will get it right.

Ask for subtle variations

If you ask DALL-E 3 to make variations based on one of its results, it may sometimes make pretty big modifications to the first prompt. If you want things a bit more similar, ask it to make "subtle variations." While this doesn't prevent it from generating a completely new image, it might change the initial prompts less. 

50 requests per 3 hours is a lot.

Take the time to tell it what to do and work through each picture. You're unlikely to hit the limit without actually attempting. 

Have fun and play around

The only way to simply grasp what DALL-E 3 is—and is not—capable of is to play around with it for yourself. ChatGPT might be capable of fulfilling some requests that you might think it'd battle with, but it could also absolutely mess up what you think would be easy modifications.

Examples of activations to request the creation of images for blog articles

ChatGPT could make 3 image sizes: wide (1792x1024), square (1024x1024), and tall (1024x1792).

Prompt 1: Perhaps you're writing a blog post about journey plans to Paris and need a unique Paris scene for your post. ChatGPT created the scene below with the prompt, “Create a Paris scene through a resort window.”

Prompt 2: How about developing a photo of a beautiful globe decoration? ChatGPT created the following image with the prompt, “Create a huge 3-D picture of a glass globe Christmas decoration hanging on a tree with a Christmas scene on the globe.”

Prompt 3: Let’s say that you are writing a blog about Christmas, and you also need a cozy Christmas scene to put up. ChatGPT created the following picture after the prompt, "Create an image of a snow scene through a window with an adorned Christmas tree near the window.”

Prompt 4: Maybe you have a cooking blog and want a country kitchen scene. ChatGPT made the picture under the prompt, "Create a realistic country kitchen scene in the process of mixing cake batter.”

Conclusion

In conclusion, the integration of DALL-E 3 with ChatGPT opens up a world of opportunities for photo creation. Whether you’re an artist, an author, or a person looking to explore their creativity, this effective tool allows you to bring your thoughts to life with stunning visuals.

While ChatGPT itself can't generate images, it can create particular descriptions of images based on users’ prompts. From here, you can then use DALL-E 3 to create a picture that suits that description.

Essentially, this indicates the image was created simply with a further step brought in between. ChatGPT is especially excellent for this kind of technique because the language it outputs can be very natural, much like the language the image generator has been trained on.