Uhm, a new ChatGPT version just dropped and GeoGuesser is now a solved problem.
PSA: When posting any outdoors photos, update your threat model from "someone skilled and dedicated could theoretically locate this" to "any stalker can do this for 20€/mo"
https://xcancel.com/arithmoquine/status/1912671688874926575/
@piegames You can ask ChatGPT also to make an estimate to the position of the photographer.
See ALT--Text of the image...
@mxk @piegames Google Lens liefert mir die Landmark, aber nicht das drumherum, vor allem führt es darauf keine Analysen aus.
Weshalb ich das für besonders halte: für ein Taschengeld bekomme ich eine Fähigkeit, die sonst Spezialisten vorbehalten war.
ChatGPT konnte mit dem Bild den Standort des Fotografen mit einer Genauigkeit von 1km bestimmen.
@mxk
Der für mich wichtigste Unterschied ist, dass niemand mehr stundenlang suchen und nachdenken muss, sondern eine einigermaßen genaue Antwort per Knopfdruck erzeugt wird.
Datenanalyse und Verletzung von Datenschutz sind immer dann am potentesten bzw. gefährlichsten wenn sie automatisiert und massenhaft passieren können.
Das ist ein Geschenk für Mega-Konzerne und unfreundliche Regierungen, und potentiell enorm gefährlich für alle anderen.
@mxk
Positiv: gefälschte Nachrichten mit plausibel aussehenden Fotos von anderen Orten oder vom falschen Zeitpunkt lassen sich schneller enttarnen.
Gaanz schlimm: Google, Meta, Microsoft, Amazon, Apple... werden früher oder später alle Fotos die sie greifen können auf diese Weise analysieren, und ebenso die US-Regierung, die chinesische, russische... und sie werden alles was sie finden zum eigenen Vorteil einsetzen.
@masek @piegames
@jeremy_list I can manage to get it in the right country with actual shot of Cambridge station, which it calls London. On the train I'm in Canada. In York it's Durham, and my boots are sneakers. Brighton is London in another, and I like smoking and gambling (nope).
So, like a lot of this stuff, all very well if you don't mind that the answers might be nonsense, don't use it to answer any important questions.
Hopefully some CEO too
Fuck AI New: https://social.nouveau.community/@june/116189020908193236 --- https://insinuator.net/2025/09/stealing-emails-via-prompt-injections/ https://flausch.social/@piegames/114352447253793517 https://pod.geraspora.de/posts/17342163 https://pivot-to-ai.com/2024/11/30/meet-the-underpaid-workers-in-nairobi-kenya-who-power-openai/ https://www.codastory.com/stayonthestory/nursing-ai-hospitals-robots-capture/ https://tante.cc/2024/04/22/ai-as-support-systems-for-diagnostics/ https://social.saarland/@fedithom/112797547775951493 https://jascha.wtf/blogmojo-ai-plagiat-im-jahr-2023-wenn-kuenstliche-intelligenz-texte-klaut/ https://chaos.social/@leah/112871670828981320 https://mastodon.green/@akshatrathi/112445979071434307 https://danmcquillan.org/chatgpt.html https://www.brusselstimes.com/world-all-news/1042696/chatgpt-consumes-25-times-more-energy-than-google https://www.vox.com/climate/2024/3/28/24111721/ai-uses-a-lot-of-energy-experts-expect-it-to-double-in-just-a-few-years https://pluralistic.net/2024/05/15/they-trust-me-dumb-fucks/#ai-search https://www.takahe.org.nz/heat-death-of-the-internet/ https://www.bloodinthemachine.com/p/for-tech-ceos-the-dystopia-is-the + eigentlich alles von https://tante.cc/
@piegames What version of ChatGPT would that be?
GPT models I have access to have literally guessed wrong twice already what flat complex I'm living in, and it's actually known around the world, and we have yearly excursions of architecture students. And it was rather bad at guessing. The screenshots are nice, but it only shows the "greatly successful" identifications.
(Yes, if I ask long enough in a mall, I can get 3 guys that believe you can get pregnant from oral ,,,, that proves nothing.)
@piegames Ah, GPT-4.5 managed with only 3 images and in the 4th query asking it to "try again".
Not sure if that counts as a "solved problem".
@yacc143 Confusingly the model is called "o3", while its predecessor is "4o". YMMV on the quality, often times it is "only" within the range of 20km.
But given the current rate of progress it is only a matter of time before someone hooks in Google Maps into it and then be able to locate any visible rooftop.
@piegames Yes o3 tends to be better at that, let me retry with it, it seems to be one the reasoning models.
But no, o3 seems to be worse at guessing where I live, sigh, it insists that my French is significantly better than it is in reality. I would probably need 3 months in France before I'd be able to order myself a lunch, to reactivate my High School French.
@piegames this is really really scary.. a friend of mine posted a picture of one of the inner roads inside our college and guess what the model guessed it accurately..
There are no publically available images of that road and still the AI model identified it..
This is really really scary
Stuart Henderson
piegames
DerSchulze
Martin Seeger
Shortman
mxk
Mr. Teatime
Jeremy List
Major Denis Bloodnok
Kat O’Brien
Joshua Barretto
Aedius Filmania ⚙️🎮🖊️
Lord Kelvin
fedithom
cache, Δ-4b0082
Andreas K
Scott 🏴😷
p0mf
HD
Ross Burton
arrbee
sijmen
Matthias
Wolf480pl
Björn ❄️ Starkimarm
Furbland's Very Cool Mastodon™
zl2tod
AI6YR Ben
Ian K. Rogers ikr?╭ರ_ಠ
ExperiencersInternational 
🇵🇸
i’m Mi
Kailas Idavanmuzhi