Like many of us I’ve been experimenting with generative AI images. The Photoshop version requires a prompt, and within a few seconds, it creates an image for you based on that text. But is it actually useful and what can you do with it?

Example 1 – “one large jigsaw piece being held by a woman”

For a client, I wanted to create an image of a woman holding a jigsaw piece. I had an image in my mind, and generated an image using the prompt above. The AI seems to understood the nouns and adjective, but not the verb

Example 2 – “a white puppy stealing food from a table”

This one using the prompt above is a bit more disturbing if you look closely. I suppose you could argue that the AI got the gist of the prompt, however, I don’t think you could argue that this image would be useful for any purpose

Example 3 – a more complex scene generated by AI

This example prompt was written by AI and used to generate the image. For some strange reason, there is a problem with the diner’s faces – you’d think that would be the easy bit!

Sunlight streams through the large windows, dappling the mismatched tables and chairs with warmth. Laughter mingles with the clinking of glasses and the soft murmur of conversation, creating a lively yet relaxed atmosphere. Four friends, their faces alight with anticipation, gather around a rustic wooden table adorned with a vibrant centerpiece of sunflowers and ferns.

In the center of the table sits a feast for the eyes and the senses. A colorful array of roasted vegetables glisten with olive oil and herbs, nestled amongst bowls of creamy hummus, vibrant red lentil salad, and fluffy quinoa tabbouleh. A towering stack of golden brown falafel sits proudly next to a platter of marinated tofu skewers, their smoky aroma hinting at their delicious char. Freshly baked crusty bread rests in a woven basket, alongside ramekins of spicy harissa and smooth tahini sauces.

One friend, a self-proclaimed “herbivore,” eagerly digs into the roasted kale chips, their satisfying crunch echoing across the table. Another, a recent vegan convert, marvels at the lifelike texture of the teriyaki glazed seitan, savoring the sweet and savory notes. The adventurous foodie of the group dives into the Ethiopian lentil stew, their brow furrowed in concentration as they discover the complex blend of spices. The final friend, a health enthusiast, beams with approval as they spoon up a vibrant acai bowl topped with fresh berries and granola.

As the meal progresses, stories are shared, jokes are cracked, and plates are passed around with generous abandon. The friends debate the merits of different cheeses made from nuts, reminisce about their favorite childhood vegetarian dishes, and exchange tips on local vegan grocery stores. Each mouthful is met with satisfied hums and appreciative nods, the delicious flavors a testament to the creativity and ingenuity of plant-based cuisine.

Sunlight fades, painting the room in a warm glow as dessert arrives. Plates adorned with rich chocolate avocado mousse, drizzled with homemade raspberry sauce, steal the spotlight. Conversation softens, punctuated by contented sighs and the clinking of spoons against ceramic. The meal ends with a shared plate of seasonal fruit, their sweetness a perfect counterpoint to the richness of the chocolate.

As the friends rise from the table, their bellies full and hearts content, they linger for a moment longer, savoring the memory of the laughter, the company, and the incredible vegan feast they’ve shared. They leave not just nourished, but inspired, their taste buds tantalized and their minds opened to the boundless possibilities of plant-based cuisine. This wasn’t just a meal; it was an experience, a celebration of flavor, friendship, and the joy of discovering something new.

Example 4 – something a bit simpler – “a close-up of a shallow pothole in a tarmac road”

This one certainly worked well and is a usable image

Example 5 – does AI understand more complex concepts? “a man, multi-tasking”

Oh dear. Almost everything about this image seems to have gone wrong!

Example 6 – something very abstract

AI seems to have had a good go at getting most of this right.

“a stripy green can standing in space on a small green planet with alien bids and spaceships flying around”

Example 7 – “show us something you would like to world to see”

Er, ok…

Example 8 – how does AI cope with abstract concepts?

This example has been a problem for the generative AI. I tried many attempts and rephrasing but they all came out similar to the following image

Show what freedom feels like but without using any text or writing

Example 9 – can AI talk to AI?

This example I asked AI to ‘Explain concisely and without images what a giraffe looks like’. The resulting generative AI prompt was:

Imagine the tallest animal you know, then tower over it with an impossibly long neck reaching for leaves unseen by others. Picture a tan canvas splashed with brown or reddish spots, all held up by spindly legs and topped with two bony knobs and a short mane. That’s the essence of a giraffe: a living tower adorned with nature’s unique artistry.

Example 10 – human hands

It’s been a running joke that AI has struggled with human hands since it first started generating images – I can only assume that this is because it creates an image of what hands are like as opposed to what hands actually look like. And because they’re so familiar to us, it’s all the more evident when it get’s them wrong. The following prompt produced the following disturbing image: “lots of human hands”

Example 11 – something abstract described in words by AI and then generated as an image by AI

A catalyst is a substance which influences a reaction, but is left unchanged. Catalysis is the name for this type of reaction using a catalyst. This is how it’s described by AI as a metaphor, and following that, is the image generated by AI based on that prompt.

Imagine two sluggish dancers, waltzing slowly, their movements hesitant and unsure. This is your chemical reaction before a catalyst appears. The dancers represent the reacting molecules, their slow pace indicating a low reaction rate.

Suddenly, a mischievous jester bursts onto the scene. This is the catalyst! They zip between the dancers, whispering secrets, demonstrating new steps, and gently nudging them closer. The dancers, energized by the jester’s encouragement and guidance, begin to move with newfound confidence. Their steps become quicker, their movements more synchronized. The dance floor, once filled with awkward pauses, now thrums with the energy of their accelerated waltz. This is the power of catalysis!

But here’s the amazing part: the jester never gets tired! After each successful dance, they simply step back, unharmed and unchanged, ready to repeat their magic with another pair of hesitant dancers. They are the ultimate recycling artist, their influence never diminishing.

Now, imagine this scene playing out on a grand scale, with millions of dancers and jesters on a colossal stage. This is the reality of countless chemical reactions happening around us, fueled by the invisible hand of catalysis. From the food we digest to the fuel that powers our cars, these tiny jester-like molecules play a crucial role in shaping our world.

So next time you witness a seemingly impossible transformation, remember the silent dance of the catalyst, working tirelessly behind the scenes to speed things up, leaving its mark on the world without ever losing its own spark.

Conclusion

The vast majority of Generative AI images on this page are completely unusable for any normal real-world application such as a blog post illustration. AI can certainly create images, but it seems to have no real understanding of how things look visually or how they work in the real world. That’s understandable, of course. However, if you are specific about what you need, and don’t require anything complex, or abstract, there’s a good chance you can create something usable.