AI image generation is now genuinely impressive

The other day my son asked me about help with a presentation he was doing for school. In the past I'd shown him how to find and embed GIF animations to spruce things up and he was looking for some more ideas. "

This time I asked if he'd checked out Gemini recently.

He shook his head.

Give it the story you're working on and get it to generate some cartoon imagery," I suggested.

He looked distinctly unimpressed.

I wasn't sure if I was over-committing on the promise of Gemini's Nano Banana.

Let's give it a go?"

We were both blown away.

The image(s) you see here was generated by Gemini to my son's specifications. (Bonus points if you can identify the English rugby player he was aiming for.)

The ability for the system to maintain backgrounds and context across multiple images was just staggeringly impressive.

"If you haven't tried it out, you've already got the capability to play with it if you've got a Google account. Head over to Gemini and try it out.