“A car exploding through the internet”

It's the end of the day and I have been playing… I mean, working hard researching various AI tools for a client. I've already spent a lot of time with ChatGPT, Dall-E, Notion’s AI, Google’s Bard, all to help me and my clients in being more productive in day-to-day tasks. I recently trained ChatGPT to write like me. It's not 100% there yet. It still lacks the little human nuances and touches, but I tell you what, it's pretty damn close (so, can you work out if this post is authored by me or ChatGPT?).

And today I spent most of my time on Creaitor.ai — an AI app that helps you with many different writing tasks (blog post outlines, headlines, ad copy, social media copy, image captions, etc…), as well as SEO/SERP tools to help you create content that gets more easily found. And. get this, it also enables you to generate images.

That’s where the headine for this blog post comes in. No, it's not about some new 'Fast and Furious' spin-off, but rather about generative AI designed to create images. Today, whilst playing with Curaitor, I had the amusing task of feeding it prompts to generate some absolute corkers . You know, the stuff that makes you wonder if someone slipped something into your tea.

The Right Prompt Makes All the Difference

If you think saying "draw me a unicorn" to Dall-E will get you a simple, run-of-the-mill unicorn, you're in for a surprise. You might get a unicorn with wheels for hooves or one made entirely of spaghetti. Yep. Spaghetti.

And this leads me to my point about how the prompt you give these generative AIs massively influences the output. Want a creative twist? Ask for "a unicorn riding a Harley Davidson in space" and brace yourself for a visual extravaganza. The more descriptive and specific your prompt, the more vivid and interesting the resulting image tends to be.

Experiment and Explore

Feel like being a bit random? Why not ask Dall-E to generate "a two-headed giraffe wearing sunglasses," and just see what kind of madness you get? The possibilities are endless. It's like having a bonkers artist at your disposal, 24/7, ready to bring your weirdest ideas to life.

But, as you may have guessed, it's not all rainbows and unicorns. Generative AI, including Dall-E, is far from perfect. Sometimes, your prompts can yield totally off-the-wall or downright unsettling images. But that's all part of the fun, isn't it? It's like an art experiment with unexpected results.

A Few Tips for Prompting image generating AI

  1. Be Specific: Generic prompts tend to yield generic results. Spice it up with adjectives and additional context.

  2. Layer Concepts: Want a dragon that's also a librarian? Say so! Layering ideas often generates surprisingly coherent and creative images.

  3. Think Outside the Box: Feel free to mix concepts from different genres, like "a steampunk octopus playing the piano."

Final Thoughts

So, go ahead, have a romp through the colourful, bizarre world of generative AI. Whether you’re looking to design a unique logo, or you’re simply out to create meme-worthy content, the right prompt can make all the difference. And who knows, you might just end up with something as explosive as a car bursting through the internet!



Previous
Previous

What’s the difference between AI and ML?

Next
Next

Unearthing your business WHY