Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Watching o3 guess a photo's location is surreal, dystopian and entertaining

simonwillison.net

Post reply on HN
Print view View original

161–170 of 453 posts

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#161

Post by SamPatt »

raincole wrote: post #108
SamPatt wrote: post #46

Earlier quoted context omitted.

They aren't fake. I'm a Master I level Geoguessr (the penultimate competitive ranking) and what the pros can do is very real. I looked at the image in the post before seeing the answer and would have guessed near San Francisco. It seems impressive to someone if you haven't played Geoguessr a lot, but you'd be surprised at how much information there is about location from an image. The LLMs are just verbalizing what i…

The fact some humans can do that doesn't make it any less impressive to me. I knew Terence Tao can solve Math Olympia questions and much much much more difficult questions. I was still very impressed by AlphaProof[0]. [0] https://deepmind.google/discover/blog/ai-solves-imo-problems...

I couldn't agree more. It's very impressive. I'm just countering the claim that it might be cheating. Of course, sometimes it might be, but knowing what I know now, it's completely possible.

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#162

Post by simonw »

SamPatt wrote: post #160
wongarsu wrote: post #147

Earlier quoted context omitted.

Also things like follow cars (some countries had government officials follow the streetview car), the season in which coverage was created, camera glitches, the quality of the footage, etc. There is a lot of "legitimate" knowledge. With just a street you have the type of road surface, its condition, the type of road markings, the bollards, and the type of soil and vegetation next to the road, as well as the presence…

Spot on. Nigeria and Tunisia have follow cars. Senegal, Montenegro and Albania have large rifts in the sky where the panorama stitching software did a poor job. Some parts of Russia had recent forest fires and are very smokey. One road in Turkey is in absurdly thick fog. The list is endless, which is why it's so fun!

Do you have a feel for how often StreetView published fresh imagery?

When that happens, is there a wild flurry of activity in the GeoGuessr community as players race to figure out the latest patterns?

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#165

Post by brundolf »

SamPatt wrote: post #63

I play competitive Geoguessr at a fairly high level, and I wanted to test this out to see how it compares. It's astonishingly good. It will use information it knows about you to arrive at the answer - it gave me the exact trailhead of a photo I took locally, and when I asked it how, it mentioned that it knows I live nearby. However, I've given it vacation photos from ages ago, and not only in tourist destinations eit…

I find this type of problem is what current AI is best at: where the actual logic isn't very hard, but it requires pulling together and assimilating a huge amount of fuzzy, known information from various sources

They are, after all, information-digesters

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#166

Post by is-is-odd »

brundolf wrote: post #165
SamPatt wrote: post #63

I play competitive Geoguessr at a fairly high level, and I wanted to test this out to see how it compares. It's astonishingly good. It will use information it knows about you to arrive at the answer - it gave me the exact trailhead of a photo I took locally, and when I asked it how, it mentioned that it knows I live nearby. However, I've given it vacation photos from ages ago, and not only in tourist destinations eit…

I find this type of problem is what current AI is best at: where the actual logic isn't very hard, but it requires pulling together and assimilating a huge amount of fuzzy, known information from various sources They are, after all, information-digesters

it's just all compression?

always has been

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#167

Post by SamPatt »

simonw wrote: post #162
SamPatt wrote: post #160

Earlier quoted context omitted.

Spot on. Nigeria and Tunisia have follow cars. Senegal, Montenegro and Albania have large rifts in the sky where the panorama stitching software did a poor job. Some parts of Russia had recent forest fires and are very smokey. One road in Turkey is in absurdly thick fog. The list is endless, which is why it's so fun!

Do you have a feel for how often StreetView published fresh imagery? When that happens, is there a wild flurry of activity in the GeoGuessr community as players race to figure out the latest patterns?

Google updates Street View fairly frequently, but most of the updates are in developed nations and they're simply updating coverage with the same camera quality and don't change the meta.

However every once in a while you'll get huge updates - new countries getting coverage, or a country with older coverage getting new camera generation coverage, etc. And yes, the community watches for these updates and very quickly they try to figure out the implications. It's a huge deal when major coverage changes.

If you want an example of this, zi8gzag (one of the best known in the community) put out a video about a major Street View update not long ago:

https://www.youtube.com/watch?v=XLETln6ZatE

The community is very tuned into Google's street view plans - see Rainbolt's video talking to the Google street view team a few weeks back:

https://youtu.be/2T6pIJWKMcg?si=FUKuGkexnaCt7s_b

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#168

Post by kavith »

parsimo2010 wrote: post #36

I’m sure there are areas where the location guessing can be scary accurate, like the article managed to guess the exact town as its backup guess. But seeing the chain of thought, I’m confident there are many areas that it will be far less precise. Show it a picture of a trailer park somewhere in Kansas (exclude any signs with the trailer park name and location) and I’ll bet the model only manages to guess the state c…

I just tested the model with (exif-stripped) images from Cork City, London, Ho Chi Minh City, Bangalore, and Chennai. It guessed 3/5 locations exactly, and was only off by 3kms for Cork and 10kms for Chennai (very good considering I used a slightly blurry nighttime photo).

So, even outside of California, it seems like we're not entirely safe if the robot apocalypse happens!

edit: it didn't get the Cork location exactly.

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#169

Post by HarHarVeryFunny »

simonw wrote: post #95
HarHarVeryFunny wrote: post #87

Earlier quoted context omitted.

I don't see how this is mind blowing, or even mildly surprising! It's essentially going to use the set of features detected in the photo as a filter to find matching photos in the training set, and report the most frequent matches. Sometimes it'll get it right, sometimes not. It'd be interesting to see the photo in the linked story at same resolution as provided to o3, since the licence plate in the photo in the stor…

If you want to be impressed I suggest trying this yourself on your own photos. I don't consider it my job to impress or mind-blow people: I try to present as realistic as possible a representation of what this stuff can do. That's why I picked an example where its first guess was 200 miles off!

I'm not a computer. I expect a computer to also do better than me at memorizing the phone book, but I'm not impressed by it.

Re: Watching o3 guess a photo's location is surreal, dystopian and entertaining

#170

Post by pphysch »

simonw wrote: post #130
casey2 wrote: post #127

surreal and dystopian is realizing that the US military has likely had (much) better tech than this for at least a decade.

I wonder if they have? My current intuition is that the US military / NSA etc have been just as suprised the explosion in capabilities of LLMs/transformers as everyone else. (I'm using "intuition" here as a fancy word for "dumb-ass guess".) I'd be interested to know if the NSA were running their own GPT-style models years before OpenAI started publishing their results.

You don't need a LLM to do this. A dedicated image->coords model would likely perform much better, and that's old school ML at this point.
Post reply on HN