We Tried the Shocking ChatGPT ‘Reverse Location Search’ Trend—Here’s What We Discovered!

ChatGPT users have recently found an interesting way to use the AI chatbot: as a reverse location search tool. Just like the game Geoguessr, you can upload a picture, and ChatGPT can often figure out where it was taken.

Mashable tried this out, and the results were surprising. When testers uploaded different photos, ChatGPT got quite close to the actual locations. For instance, it mistakenly identified a rooftop hotel in Buffalo instead of Rochester but still generated specific addresses.

OpenAI, the company behind ChatGPT, recently updated their models with improved visual reasoning. This upgrade is making waves on social media, with users turning their pets into cartoon characters or themselves into action figures. However, the reverse location feature raises some privacy concerns.

This trend caught fire when users noticed ChatGPT’s ability to identify locations even without image metadata. Expert Ethan Mollick shared an example on social media, showing that the AI could guess his driving location despite his efforts to remove any identifiable data.

Mollick highlighted that this illustrates the power of "agentic AI," which can think critically and perform complex tasks like searching the web.

So, how good is ChatGPT at pinpointing places? In one test, it identified a flower shop photo as being in Brooklyn, though it was off by about seven miles. In another, a photo from Japan was accurately identified as Arashiyama, Kyoto, near the famous Togetsukyo Bridge.

Testing continued with images from a well-known Instagram model. ChatGPT not only guessed the general area but also suggested specific high-rise buildings and even a home address. While the house was fairly famous among influencers, the level of detail was both impressive and worrying. It serves as a reminder that we need to think carefully about what we post online, as AI can expose our locations.

OpenAI acknowledges the mixed feelings around these tools. They believe that the new reasoning models can assist in various fields, like emergency response and accessibility. They promise to work on privacy safeguards and actively monitor misuse of their technology.

As we embrace these advancements, it’s crucial to balance innovation with caution. Keeping personal information safe is more important than ever in this digital age. For more on OpenAI’s latest updates, check their official blog.

Check out this related article: Synology Mandates Use of Branded Drives for High-End NAS Products: What You Need to Know