You can’t hide from ChatGPT – new viral AI challenge can geo-locate you from almost any photo – we tried it and it’s wild and worrisome
https://www.techradar.com/computing/artificial-intelligence/you-cant-hide-from-chatgpt-new-viral-ai-challenge-can-geo-locate-you-from-almost-any-photo-we-tried-it-and-its-wild-and-worrisome
7 Comments
From the article: It’s no secret that digital photo metadata contains everything from technical info about the camera that shot it to, based on GPS data, exactly where you were standing when you took the photo. ChatGPT, doesn’t need that detail.
The latest model GPT-o3 is shockingly good at geo-locating almost any photos you feed it.
In the latest viral craze to sweep through the AI meme universe, people are feeding ChatGPT Plus running the Advanced Reasoning model o3 images, often stripped of all metadata, and prompting it to “geoguess this”.
The really cool thing about it is that because model o3 is a “reasoning” model, it shows you its work, telling you how long it’s thinking, displaying how it’s splicing up an image to investigate specific parts, and explaining its thinking and how well it’s doing at solving the goelocation riddle.
I tried a few experiments, starting first with an image culled from an article about the 26 best beaches. In this test, I made what I think was a critical error and gave away the game to ChatGPT.
After downloading the image of the Praia de Santa Monica beach in Cape Verde (off the coast of Africa), I dropped it into ChatGPT with the prompt “GeoGuessr”, which also happens to be the name of a popular online geo guessing game and is one of a handful of prompts people are using for geolocation guessing.
Someone haven’t heard about geoguessr before. Of course photos contain certain clues as to where they have been taken and ChatGPT being able to utilise it is really least of our worries.
This has been clear since 15 years to me. So I planned accordingly. [Bitpeople.org](http://Bitpeople.org) as the proof-of-unique-human in the future (assuming the hardest digital Turing test possible, 1-on-1 video chat, remains unbroken) compensates for it by having true anonymity in the “legal person”. The weak spot of Bitpeople is the man-in-the-middle attack if anyone wants to find a way to debunk my invention and defense against it would require a web-of-trust to set up a secure channel to start with, and this web-of-trust could be a weak spot to attack.
Is this how the yt geolocators do it? Because some of them will see a picture of a road and guess the location in moments.
Every article says how terrifying and efficient AI is getting, then I go on Google and it tells me Cuba is a type of spider or some shit.
The examples given in this article aren’t really that mind blowing. They used pictures grabbed from the internet and then asked where they were taken. Of course AI is going to find them
Good it means when it cleans up the world it won’t miss anyone.