Online users have started employing ChatGPT to identify the specific locations depicted in photographs. This phenomenon has sparked a debate on privacy concerns as AI technology advances.
How ChatGPT Guesses Locations from Photos
The AI company OpenAI has introduced new AI models, o3 and o4-mini, that can extract details from photographs such as building structures, signage elements, and vegetation to determine geographical locations. By analyzing images, the AI can identify dining styles from restaurant menus, aiding in pinpointing the restaurant's location without relying on location tags.
Why This Is Fun But Risky
Users on social media have been enjoying the challenge of testing ChatGPT with various photos to play a location-guessing game similar to GeoGuessr. While it may seem like harmless fun, there are risks involved as individuals can potentially locate someone's residence or workplace without consent, a practice known as doxxing.
How Good Is ChatGPT at Guessing?
Tests have shown that ChatGPT's o3 model demonstrates improved accuracy compared to its predecessors in identifying locations from photos. However, the AI is not infallible, sometimes making incorrect guesses or getting stuck. Despite its occasional errors, ChatGPT's ability to pinpoint specific locations raises concerns among privacy experts.
What Can You Do to Stay Safe?
While OpenAI has not yet incorporated tools to prevent location guessing, users are advised to exercise caution when sharing images publicly to protect their privacy until such features are implemented.
Company Policies
© 2024 The Tech Basic INC. 700 – 2 Park Avenue New York, NY.
Sign in to your account
