Death To Photoshop?
4 Sept 2025 · Aditya Batra · reaching-singularity · ai-tools

The AI wars have begun, and the first casualty is Photoshop, toppled by Google's newly released Gemini 2.5 Flash Image, codenamed Nano-Banana. Launched in late August 2025, Nano-Banana is available through the Gemini app, Google AI Studio, and the Gemini API.
Nano-Banana is a fast, affordable image model that currently dominates LMArena's leaderboard. After a while, it feels like Google has made something genuinely cool again. It climbed to the top of the image-editing and text-to-image rankings shortly after launch, and in the process pretty much killed Photoshop for casual users. Production use is still limited by the SynthID watermark baked into outputs. Skipping the nerdy stuff, here is what you can actually do with it.
Making a LinkedIn profile picture professional
One sloppy two-line prompt ("change the attire to something more professional, like a suit") and it swapped a kurta for a suit with the face, hair and lighting intact. No masking, no layers, no lasso tool.
Chilling with the homies
It put me on a yacht mid-conversation with Elon, Sundar and Zuck. Consistent faces and lighting from a two-line prompt with zero prompt engineering. I also asked for Altman, but Gemini did not feel like adding him for some reason. See the chat.
Bringing maps to life
Prompt: "Generate the image of what the red arrow sees." It turned a flat map screenshot near the Taj Mahal into a believable street-level scene. I don't know how accurate it is, but turning a map into a photo is a party trick worth knowing. Gemini chat.
Turning photos into AR explainers
Nano-Banana has access to Gemini's world knowledge, so you can upload real photos and ask it to annotate them. A shot of India Gate came back labeled with the architect, the year completed, and what each part of the monument is. A static photo becomes a lightweight AR-style explainer. Credits and cooler AR visuals: Bilawal Sidhu on X.
Making your drawings real
Even a genuinely bad doodle becomes a photorealistic image with "give real world representation of this person". The better the input drawing, the more refined the output, but the fact that a sloppy sketch turns into something that looks real is quietly wild.
Creating an entire brand in minutes
We started with a text prompt for a logo, slapped that logo onto a coffee cup, then built a full branded coffee-shop identity: avatars holding the cup, lifestyle shots, Instagram-ready images. Nothing to full brand identity in minutes. Gemini chat.
This is the tip of the iceberg. For a deeper tour, this Futurepedia video covers more use cases in detail.
Want to use AI like this for your own work? Keep an eye on Singularity's workshops and events. One email a week, only the stuff worth your attention.
want in on the next one? upcoming events →