Introduction: My Little Experiment with AI & a Desi Ninja
Namaste, folks! I was scrolling through Twitter/X the other day and stumbled upon this interesting idea – what if we asked AI to create a “North Indian Woman Ninja”? And not just any ninja, but one depicted in the beautiful “ochre-garnet Sansuiga ukiyo-e” style. I was immediately intrigued and decided to give it a shot using different AI image generators. Let me share my experience and the fascinating results!
What is Sansuiga Ukiyo-e?
Sansuiga Ukiyo-e is a Japanese woodblock printing style that flourished during the 19th century. It is known for its focus on landscapes, often featuring panoramic views with a deep perspective. The “Sansuiga” name translates to “perspective picture,” emphasizing this key element. These prints typically depict natural beauty, historical events, or scenes from everyday life, rendered with vibrant colors and intricate details. Ochre and garnet are colors that could be found within the palette of a Sansuiga Ukiyo-e, though not necessarily dominant or defining characteristics.
For suggesting Sansuiga Ukiyo-e, a hat tip to Duncan (theartofweb) on X.
The Prompt That Started It All
Here’s the exact prompt I fed to the AI generators
steampunk style image showing a well trained north Indian woman Ninja, rule of thirds, Meadow Sunset, oil painting by klimt, solar punk style
A Look at the AI’s Creations
Image 1: Tengr.ai gave me a picture of a woman in a cool steampunk outfit, complete with goggles, against a sunset backdrop. The colours were mostly yellows and browns, giving it a nice earthy feel. While the steampunk part was spot on, I felt the “North Indian” and “Ninja” aspects were a bit missing. She looked more like a steampunk adventurer than a stealthy ninja, you know?
Image 2 : DALL-E 3 presented a woman in a gorgeous, traditional North Indian outfit. There were subtle hints of steampunk in her jewellery and clothes, which was a nice touch. The sunset and meadow background were there, and the colours were rich and vibrant, just like I imagined. She didn’t have any obvious ninja weapons, but her expression and stance gave off a powerful and confident vibe, like a true ninja warrior. This image, as well as image3 below, were generated using You.com's image generator tool.
Image 3 : This one from DALL-E 3 was quite interesting! It showed a fully masked figure in full-on steampunk gear, wielding a sword and other gadgets against a golden background with a sunset. The ninja part was definitely clear – the action pose and weaponry were spot on. But the “North Indian” element was a little hidden, maybe because of the mask and all the steampunk stuff.
Image 4 : Tengr.ai’s second attempt gave me a woman in a teal Japanese-style outfit with a katana, standing in front of mountains and a river. It felt more like a samurai than a ninja, and the “North Indian” part was completely missing. I think Tengr.ai might have gotten a bit confused with this one!
Image 5 : Leonardo.ai went for an anime-style image, with a woman in a red samurai-inspired outfit and two katanas. Again, the ninja and North Indian aspects weren’t really there, and the anime style wasn’t what I was hoping for.
Image 6 : Grok’s creation was a woman with her back turned, wearing a beautiful but revealing black and gold outfit. The sunset and poppy flowers in the background were lovely, but I couldn’t really tell if she was a ninja or from North India. It seemed like Grok focused more on making a visually appealing picture rather than sticking to the prompt.
My Thoughts on the Experiment
It was fascinating to see how each AI interpreted my prompt in its own unique way. Some did a great job capturing the steampunk and sunset elements, while others struggled with the “North Indian” and “Ninja” aspects. And honestly, none of them really nailed the “Sansuiga Ukiyo-e” style, which shows how challenging it can be for AI to understand and recreate specific art styles.
This whole experiment was a great reminder of how AI is constantly evolving, and how it can be a powerful tool for creativity and exploration. But it also showed me that AI still has a long way to go in terms of understanding complex concepts and specific artistic styles. It was a fun journey, and I’m excited to see what AI can do in the future!