7 smart ways to use ChatGPT’s new image AI

ChatGPT’s native image generator has been the talk of the town for its ability to not only convert text prompts into images, but also refine images with various effects, including the now viral Studio Ghibli-like images. However, OpenAI’s latest tool has much more to offer than just ‘Ghiblifying’ or editing images, making it one of the best AI image generators out there.

But what exactly is native image generation and why is it such a big deal?

Native image generation refers to the ability of the Large Language Model (LLM) to directly generate and edit images using its multimodal capabilities, rather than relying on external models such as DALL-E 3. OpenAI has unlocked image generation capabilities directly in GPT-4o, allowing the chatbot to ‘refine images through natural conversation’. ChatGPT can now also handle between 10-20 different objects in an image, bringing more control and consistency to the images.

Let’s take a look at the 7 possible use cased of ChatGPT’s new native image generation.

7 ways you can use ChatGPT’s new AI image generator:

1) Cartoons

ChatGPT can now directly create cartoon characters and even convert real life images or popular memes into the theme of a particular cartoon. One user on Reddit was able to convert the popular ‘cheating boyfriend’ meme to Lego style, while a host of other effects are also possible.

2) Infographics

With the new capabilities of GPT-4o, there is no need to learn how to use any software to create infographics. Instead, users can simply ask ChatGPT to instantly create an infographic based on the given information.

3) Posters:

While most AI image generators on the market can already create simple posters, ChatGPT’s new capabilities take it a step further. It can now even generate advertisements for brands with a single prompt.

4) Slides:

ChatGPT can now create visually appealing presentation slides that includes bullet points and even concise summaries. It can also be used to generate entire slide decks while maintaining a professional design.

5) Illustrations:

ChatGPT can now help artists create more accurate AI-generated images. It can also easily turn their sketches into photorealistic images with text prompts for specific effects.

6) Stories:

Since its sudden popularity in 2022, ChatGPT has already been useful in generating storylines, even in the style of some popular authors. However, with its new capabilities, ChatGPT can convert these storylines into comic books, with each character having a unique personality and consistency in narrative.

7) Designs:

ChatGPT can now generate detailed graphics with a single text prompt. Moreover, the native image generation also allows users to tinker with the image until they get the desired result.

How to prompt ChatGPT 4o for great images:

There are a few key details to keep in mind when giving ChatGPT prompts to create images, such as describing the main subject, lighting, perspective, background and atmosphere, along with the art style for that particular image.

Here’s an example for a good prompt, straight from ChatGPT:

“A cyberpunk cityscape at night, filled with glowing neon signs and flying cars. The streets are wet from recent rain, reflecting vibrant blues and pinks. A lone figure in a trench coat walks under a flickering streetlight. High detail, cinematic lighting, digital painting style.”

If you are still confused about how to prompt ChatGPT for great images, just use another AI chatbot like Grok or Gemini to get suggestions for your prompt and refine it from there on.

Limitations of ChatGPT’s new image generator:

While GPT-4o’s new image generator is a step up from the previous DALL-E models used by OpenAI, there are still many problems that can occur with it. These include issues with cropping, hallucination (false or misleading information given by the chatbot), high binding issues, accurate graphing, multilingual text rendering, editing accuracy and dense information with small text.

What are some alternatives for OpenAI’s new tool:

At the moment, Gemini via Google AI Studio is the best free alternative if you want to do most of the above tasks.

However, if you want to jump on the Ghibli viral trend, Gemini may not be the best option. In such a scenario, Grok is a far superior alternative.

ChatGPT has also started to offer up to 3 image generations per day to free users from today.