Create an Infographic

Create an Infographic with GPT 4o? Here’s What Happened for Me: Artificial Intelligence Trends

I stumbled upon an article that told me I could create an infographic from an image containing data, so I decided to try it. Here’s what happened.

The article from Forbes (New ChatGPT: 12 Game-Changing Prompts You Won’t Have Tried Before, written by Jodie Cook and available here) discussed several interesting prompts that illustrate new functionalities in ChatGPT 4o to experiment with. This one in particular caught my eye:

Create infographics

Advertisement
Exterro

Upload an image containing data or sketches, and ask ChatGPT 4o to help you create a professional infographic.

“Create an infographic from the data in this image. Use the hex colours [include your brand colours]”

“How cool is that?”, I thought. I have to try that!

So, I did, using the statistics from my latest annual cybersecurity stats post that I publish every year on April Fools Day. I copied the stats into a Word document, formatted it so that they all fit on a single page, corrected the one stat that I intentionally published wrong (as part of the April Fools challenge), captured a screen shot of the stats and saved as an image. Took about 5 minutes. Here’s the image I used (reduced in size to fit within the blog post):

Advertisement
TransPerfect Legal

I then uploaded the image into GPT 4o and told it to “Create an infographic from the data in this image.” I didn’t worry about the color scheme as that didn’t matter to me.

GPT 4o went through its process to generate an image, and, in less than a minute, it gave me this:

Huh? Well, it looks like an infographic. But the data is absolute garbage. Yeesh.

On the off chance that was an anomaly, I decided to hit the regenerate button. This is what I got the second time:

Ick! No better. Nice design, but garbage data again.

DALL-E doesn’t do well accurately representing words on images, so I’m thinking that DALL-E is being used here to generate the infographics. Regardless, it can’t create an infographic that actually makes any sense. At least not for now.

Lesson learned here is don’t take at face value claims that an AI model can do certain things just because someone says so. Test it out yourself.

So, what do you think? Do you have any examples of failures from an AI model that you’ve experienced? Please share any comments you might have or if you’d like to know more about a particular topic.

Image created using GPT-4o’s Image Creator Powered by DALL-E, using the term “robot looking at a computer screen with an infographic on it”.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.


Discover more from eDiscovery Today by Doug Austin

Subscribe to get the latest posts sent to your email.

2 comments

  1. I’ve tried it a hundred times and ways, and everytime the text emerges garbled as you describe–similar to the words I want to some extent, but more hieroglyph than intelligence. If you find the secret, please share; because I’m tired of all the hype when there is a banal sameness to the imagery (Robots! Again? Really?!) and no ability to faithfully include essential text or numbers.

  2. I will certainly do that, Craig, though I think DALL-E must be involved as it does the same thing when I try to generate images with words in them, so I think it’s the culprit. Looks like we still have to do some things ourselves! 😉

Leave a Reply