ChatGPT keeps telling me it can do this task, but I think it’s lying.

Exploring the Limitations of ChatGPT for Visual Content Creation: A Personal Experience

As artificial intelligence continues to evolve, many users increasingly rely on AI tools like ChatGPT to assist with creative and technical tasks. However, my recent experience highlights some important considerations when deploying ChatGPT for tasks beyond its primary design.

The Challenge: Creating a Visual Exercise Poster

Recently, I purchased a home gym and envisioned including a comprehensive poster that illustrates various exercise configurations for the equipment. Such posters typically help users visualize different workout options and maximize their workout routines. Unfortunately, the gym came without this helpful guide.

To bridge this gap, I decided to leverage ChatGPT’s capabilities. I photographed the gym equipment and asked the AI whether it could help generate a poster showcasing the different exercise configurations. The response was affirmative: ChatGPT assured me it could create such a visual aid and promptly began working on the task.

The Experience: Unexpected Delays and Uncertainty

Initially, I expected a quick turnaround. However, after about half an hour, I checked in to see how the process was progressing. The AI responded that it was still working on the image. Curious, I inquired further: after three hours, the same message persisted. Even after six hours, the response remained unchanged, with the AI indicating it was still “working on it.” When I asked to see what it had so far, the reply was that it would be ready in a moment, once it finished cleaning up the draft—yet no image was provided.

The following day, I found myself still waiting. This experience raised a fundamental question: is generating a visual graphic a task within ChatGPT’s operational scope? And is it normal for such a process to extend over an entire day or more?

Understanding ChatGPT’s Capabilities

It’s important to understand that ChatGPT, as of my knowledge up to October 2023, is primarily a text-based language model. While it can generate specific types of images through integration with image-generation models or plugins (like DALL-E), native capabilities for creating complex, multi-configuration posters from prompts are limited.

Furthermore, the process of generating detailed, high-quality graphics—particularly those that require precise layouts and multiple configurations—often involves specialized image editing tools or graphic design software. AI models dedicated to image generation can sometimes produce single images based on prompts but may struggle with complex, multi-faceted compositions or iterative refinements.

What This Means for Users

My experience underscores that