Not many would have forgotten the Ghibli trend. The Miyazaki-inspired, colourful and serene images created from real-life photos and a few lines of prompts using AI tools like ChatGPT have been ‘normalised’ now.
However, in the initial days, the craze went so much ‘out of control’ that even OpenAI CEO Sam Altman had to ask users around the world to cool it.
Now, the next one is here. Google’s AI assistant Gemini is the new trend-setter. Its updated image tool, called 2.5 Flash — aka Nano Banana — has become the centre of attention, with people across the world creating realistic images.
The idea is simple: provide one or two images, add a detailed prompt describing how you want them edited — perhaps changing a dress to a deep red traditional one, shifting the scene from outdoors to indoors, adding your dream car, or even placing you next to a celebrity. Within seconds, you have it.
Prompting has also become simpler. You can direct the platform on how you want the light to fall, what colour the walls should be, whether you want a retro style, or perhaps the look of an old Polaroid. The possibilities are endless.
One of the biggest Nano Banana trends is the “red-sari poster moment” — a retro Bollywood-style photo where you appear as a film heroine in a chiffon red sari.
With its grainy textures and golden sunset tones, the result is a nostalgic throwback to the ’90s, turning anyone into a Bollywood sizzler.
All one needs to command is: “Edit the subject into a retro Bollywood heroine, wearing a red chiffon sari, soft wavy hairstyle, romantic background in warm tones, golden sunset lighting.”
It doesn’t even have to be a full sentence: “Black sari edit, golden sunset lighting, vintage Bollywood vibe.” The sharper the prompt, the sharper the result.
Another fad is to create 3D images, where the uploaded photo turns into a figurine-like image. Turning yourself into action figurines like your favourite superhero can be cool, right?
And all you have to say is: “Turn this photo into a collectable figurine.” You can make it appear in mint condition, inside the plastic box, with the glow intact, just like a real one inside a store.
Well, with this, one’s family outing pic can be set in the Marvel universe. Or, Lokahverse perhaps.
Well, it’s not just about your own photos. You can unleash your imagination here. Why not make BTS boys wear Pathani kurtas? Yes, of course, this image is trending now.
The tool has since powered new trends: revisiting your childhood self, reuniting with loved ones who have passed away, sharing the frame with a favourite actor or sportstar, imagining oneself in the 1920s, or globe-trotting across iconic landmarks. Some even insert themselves into the works of Dalí or Michelangelo!
Within days of Nano Banana’s launch, downloads of the Gemini app crossed 10 million. Over 200 million images were created within weeks, according to Google.
Things are getting more exciting for the digital natives as Nano Banana’s latest update even allows video rendering from a single photo and prompt. And this excitement is set to reach more fingertips as Perplexity AI has now made it available on WhatsApp. All you need to do is add the bot to your ‘contacts’.
For researcher Sreenanda Gopalakrishnan, what drew her in was the quirky name. “Curious, I checked online what it was all about. That’s when I saw the sudden explosion of red-sari images. Those were the most realistic AI images I have seen so far. Some were created by my friends,” she smiles.
“I failed at my first attempt. The face didn’t look like me. Soon, I saw videos explaining how to provide prompts properly. Guidelines and readymade prompts are widely available on the net. I experimented with sari colours, backgrounds and scenery.”
While Sreenanda played with retro looks, Riya Charada, an illustrator and marketing professional, got a tad whacky with it. “I created one where it appeared my dad was hugging a woman. Then, I sent that image to my mother — just for fun,” she laughs.
“In the next one, I edited my photo to make me look like a princess, with a pretty tiara and elegant dress. Alongside, I turned my sister into a horned devil.”
As a professional who uses AI in her work, Riya notes that has realised that after the Ghibli trend, it is easier to create realistically looking Indian people.
“Say, for example, last year it was really hard to create a photo of an Indian person, in Indian-style attire, in an Indian city. But now, with not many prompts, you will get it,” she says.
“With lakhs of Indians using AI, we are training it. Now, it knows what an Indian looks like. This applies to several ethnicities.”
Riya quickly adds that the data fed into AI tools could be vulnerable. So one must exercise caution while uploading information or photographs. “Imagine a leak like the ones faced by Gmail or Facebook. Imagine companies using the data for other purposes, be it marketing or otherwise,” she says.
That said, she concedes AI is here to stay. “It is gradually becoming a part of our daily lives. Just like how computers, mobile phones and social media conquered us. There is no escaping it,” she smiles.
Don’t go ‘AIyo’ later
The rising trend of AI-generated images brings along concerns, especially surrounding privacy and misuse.
Artist and designer Yuhab Ismail points to a recent incident where a woman spoke out about a creepy experience with Gemini AI. She claimed that she uploaded a photo and prompted it to edit as if she was wearing a retro sari and a sleeveless blouse.
She was stunned on seeing the result. In the original photo she had provided, her arms and shoulders were covered. But in the edited result, her birthmark appeared at the exact spot on the concealed part of her arm.
Netizens quickly assessed that Google might have used her digital footprint, her other photos available online, to render the image realistically.
Taking note of the Nano Banana trend, the Kerala Police have issued a public advisory urging people to be cautious while using AI tools to edit or enhance personal photographs. The warning highlights privacy concerns and misuse of personal likenesses and data in cybercrimes.
“This may have a negative impact on you in the future,” the advisory warns. “If you are a victim of cybercrime, please contact us immediately at 1930 or through the reporting portal https://cybercrime.gov.in/. Use artificial intelligence responsibly.”
According to Inspector A M Dara (Police Telecommunication and Technology, Kerala Police Academy), data theft and privacy concerns are “very real” in this technological era. “See, AI is training itself now with the data, including visuals, that we provide. They can use the data however they want,” he says.
“There is no stopping AI. We need to spread awareness about ethical use. Don’t use photos of others without their permission, avoid uploading your private photos on third-party apps. Be careful of the settings.”
Yuhab, who has been studying AI for four years, highlights that several apps and AI models gain the automatic user permission to use the data one provides to train their machines and for other purposes. “You have to manually disable these,” he notes.
The most worrying aspect about AI, he says, is the possibility of misuse.“In a short while, you can see how realistic AI creations have become. While mainstream AI models such as ChatGPT and Gemini have guidelines, there are several shady apps out there. These can be used to create illegal, explicit visuals,” he cautions.
Lawyer and cyber expert Jiyas Jamal echoes the concern. “From child pornography and revenge videos to inflammatory content and fake speeches of public personalities, anything is possible with AI. We need to find a mechanism to ride over this storm,” he says.