Google Gemini’s recent AI-generated image of a bustling North Indian street has sparked intense debate on social media. Initially gaining attention for its “nano banana” feature, Gemini was prompted by user Abhishek (@ImAbhishek7_) to visualize a busy street during peak hours. The generated scene featured auto-rickshaws, potholes, stray dogs, vegetable carts, and political hoardings, evoking a gritty urban atmosphere.
However, social media users questioned the authenticity of the image, suggesting that specific elements like haze and political banners were influenced by the prompt. Some comments indicated that earlier outputs didn’t include certain features, raising concerns about AI interpretation.
We tested the same prompt and found a different result, missing potholes and hoardings but including similar street elements. AI models like Gemini rely on extensive datasets and algorithms to generate images, leading to varying interpretations of prompts. This incident highlights the complexities of AI-generated visuals in representing real-world scenes.
Source link
