piegames @piegames@flausch.social 额，新版 ChatGPT 刚刚发布，感觉 GeoGuesser 已经变成一个已解决的问题了。

温馨提示：发布任何户外照片时，请将你的威胁模型从“理论上，有技能和有决心的人可以定位到”升级为“任何跟踪者都可以花 20 欧元/月来做到这一点”。

https://xcancel.com/arithmoquine/status/1912671688874926575/ Nitterhenry (@arithmoquine) alright Apr 17, 2025, 08:51 AM·Public·Web 上次编辑 Apr 17, 09:17 AM 955boosts·611favorites Quiet public15hInfoseepage @Infoseepage@mastodon.social @piegames 记得以前 Andrew Sullivan 的 Dish 网站上有一个常规功能，他会发布一张从窗外拍摄的照片，然后他的观众会尝试根据他们看到的内容确定位置。猜得最接近的人获胜。很多时候，他们正确地识别出了建筑物和拍摄照片的确切窗户。 0 Public15hDerSchulze @DerSchulze@elbmatsch.de @piegames 20€？呸，我猜这在免费帐户中也有效，不是吗？ 1 Quiet public15hpiegames @piegames @DerSchulze 我没有 ChatGPT，但根据我的理解，你付的钱越多，结果质量就越好 0 Public15hMartin Seeger @masek@infosec.exchange @piegames 你也可以要求 ChatGPT 估计摄影师的位置。 请查看图像的 ALT--Text... Estimating the Photographer’s Position: Given the angle and perspective: The view is looking southeast toward the Mediterranean Sea at sunset or sunrise. The W Hotel is centered in the frame and lit by low sunlight, likely sunset (since it's west-facing). The rooftops in the foreground suggest the photo was taken from a rooftop or terrace in the Ciutat Vella (Old Town) or El Raval district—potentially somewhere around La Rambla, Gothic Quarter, or Raval, looking toward the waterfront. The prominent tower to the left of the image is the Torre de Jaume I, a cable car tower that connects Montjuïc with Barceloneta, which further confirms this orientation and location. Rough Coordinates Estimate: 📍 Approx. 41.378° N, 2.170° E (Within a few blocks west of La Rambla) ALT 隐藏 3 Public15hShortman @Shortman@sueden.social @masek @piegames Du auch in Barcelona? Frage für ChatGPT 1 Public15hMartin Seeger @masek@infosec.exchange @Shortman @piegames Ich bin vier Wochen auf Workation hier in Barcelona. 1 Public15hShortman @Shortman@sueden.social @masek @piegames 10 Min ago ;-) 隐藏 1 Public15hMartin Seeger @masek@infosec.exchange @Shortman @piegames Da waren wir Sonntag. Ich sitze jetzt im Hotel und arbeite 😄 1 Public15hShortman @Shortman@sueden.social @masek @piegames falscher Tag bzgl des Wetters 0 Public13hmxk @mxk@hachyderm.io @masek @piegames to be fair, this picture also can be located using Google Lens.Using Google lens and a bit of skill, a lot more outdoor photos already could be located lots faster, than most people would assume. 1 Public13hMartin Seeger @masek@infosec.exchange @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. 1 Public13hmxk @mxk@hachyderm.io @masek @piegames ja, ich bestreite ja nicht, dass es das einfacher macht, aber auch mit Lens und manueller suche danach auf Google Maps ist das verorten von Fotos halt lange keine exotische Fähigkeit mehr und Menschen verraten schon lange mit Fotos viel mehr über ihren Standort, als ihnen bewusst ist. 1 Public13hMartin Seeger @masek@infosec.exchange @mxk @piegames Dem zweiten Teil des Satzes stimme ich zu. Für den ersten Teil nur zum Teil: max 10% der Menschen konnten bisher eine solche Analyse durchführen. ChatGPT gibt die Fähigkeit nun an sehr viel mehr Menschen und senkt die Hürde beträchtlich. 1 Public13hmxk @mxk@hachyderm.io @masek @piegames ich glaube der Unterschied besteht darin, ob man glaubt, dass es einen wesentlichen Unterschied macht, ob 5% der Bevölkerung ein Foto durch ihre Fähigkeiten verorten können oder jeder der sich einen ChatGPT-Account leistet. 0 Public4hKat O’Brien @obrien_kat@mastodon.world @masek @piegames it’s definitely in Raval from the angle of the shot. 0 Public15hJoshua Barretto @jsbarretto@social.coop @piegames Women are going to die because of this 2 Public15hAedius Filmania⚙️🎮🖊️ @Aedius@lavraievie.social @jsbarretto @piegames Hopefully some CEO too 1 Public15hLord Kelvin @kelvin0mql@mastodon.social @jsbarretto @Aedius @piegames Now you’re talkin’. 0 Public13hPasswordsarehard4 @passwordsarehard4@mastodon.social @jsbarretto @piegames child pornographers as well, silver linings and all. I’ve sent in dozens of hotel window views to help find where videos may have been shot at and when. 0 Quiet public15hfedithom @fedithom@social.saarland @piegames Goes straight to my list: https://social.saarland/@fedithom/112455121136790239 social.saarlandfedithom (@fedithom@social.saarland) AI can fuck off. New: 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 https://mastodon.green/@gerrymcgovern/112456092679172352 https://www.swr.de/swrkultur/wissen/clickworker-ausgebeutet-fuer-kuenstliche-intelligenz-swr2-wissen-2023-08-24-102.html + eigentlich alles von https://tante.cc/ 0 Quiet public15hcache, Δ-4b0082 @cache@catcatnya.com @piegames jesus fucking christ 0 Public15hAndreas K @yacc143@mastodon.social @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.) 3 Quiet public14hpiegames @piegames @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. 1 Quiet public13hAndreas K @yacc143@mastodon.social @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. 0 Public14hAndreas K @yacc143@mastodon.social @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". 1 Quiet public14hpiegames @piegames @yacc143 Update your threat model accordingly 0 Public11hp0mf @p0mf@fuck.church @yacc143 @piegames Habitat 67? 0 Public14hHD @hareldan@mapstodon.space @piegames are we sure it's not just reading the exif data? 2 Quiet public14hpiegames @piegames @hareldan Would need some confirmation, but I doubt it. At the very least it can extract non-trivial information out of an image regardless of EXIF: "Left‑hand house has stucco scrollwork and window surrounds typical of late‑19th/early‑20th