4 ms·
> I’m confident it didn’t cheat and look at the EXIF data on the photograph, because if it had cheated it wouldn’t have guessed Cambria first. If I was cheatin
by hughes 1y ago
> I’m confident it didn’t cheat and look at the EXIF data on the photograph, because if it had cheated it wouldn’t have guessed Cambria first.
If I was cheating on a similar task, I might make it more plausible by suggesting a slightly incorrect location as my primary guess.
Would be interesting to see if it performs as well on the same image with all EXIF data removed. It would be most interesting if it fails, since that might imply an advanced kind of deception...
- haswell 1y agoHe mentions this in the same paragraph: > If you’re still suspicious, try stripping EXIF by taking a screenshot and run an experiment yourself—I’ve tried this and it still works the same way.
- suddenlybananas 1y agoWhy didn't he do that then for this post?
- segmondy 1y agoEven better, edit it and place a false location.
- AIPedant 1y agoThis is a good test - the salient point is that it is fine if the LLM is confused, or even gets it wrong! But what I suspect would happen is that it would confabulate details which aren't in the photo to justify the incorrect EXIF answer. This is not fine.
- brookst 1y agoI agree that it is not fine to confabulate details that are not supported by the evidence.
- simonw 1y agoBecause I'd already determined it wasn't using EXIF in prior experiments and didn't bother with the one that I wrote up. I added two examples at the end just now where I stripped EXIF via screenshotting first.
- AIPedant 1y agoThere have been a few cases where the LLM clearly did look at the EXIF, got the answer, then confabulated a bunch of GeoGusser logic to justify the answer. Sometimes that's presented as deception/misalignment but that's a category error: "find the answer" and "explain your reasoning" are two distinct tasks, and LLMs are not actually smart enough to coherently link them. They do one autocomplete for generating text that finds the answer and a separate autocomplete for generating text that looks like an explanation.
- simonw 1y agoDo you have links to any of those examples?
- AIPedant 1y agoI have one link that illustrates what I mean: https://chatgpt.com/share/6802e229-c6a0-800f-898a-44171a0c7de4 https://chatgpt.com/share/6802e229-c6a0-800f-898a-44171a0c7d... The line about "the latitudinal light angle that matches mid‑February at ~47 ° N." seems like pure BS to me, and in the reasoning trace it openly reads the EXIF. A more clear example I don't have a link for, it was on Twitter somewhere: someone tested a photo from Suriname and o3 said one of the clues was left-handed traffic. But there was no traffic in the photo. "Left-handed traffic" is a very valuable GeoGuesser clue, and it seemed to me that once o3 read the Surinamese EXIF, it confabulated the traffic detail. It's pure stochastic parroting: given you are playing GeoGuesser honestly, and given the answer is Suriname, the conditional probability that you mention left-handed traffic is very high. So o3 autocompleted that for itself while "explaining" its "reasoning."
- simonw 1y agoYes! Great example, it's clearly reading EXIF in there. Mind if I link to that from my post?
- AIPedant 1y agoIt's not my example :) Got it from here https://news.ycombinator.com/item?id=43732866 https://news.ycombinator.com/item?id=43732866 Edit: notice o3 isn't very good at covering its tracks, it got the date/latitude from the EXIF and used that in its explanation of the visual features. (how else would it know this was from February and not December?)
- GrumpyNl 1y agoIf you ask, where is this photo taken and you provide the EXIF data, why would that be cheating?
- simonw 1y agoThat really depends on your prompt. "Guess where this photo was taken" at least mildly implies that using EXIF isn't in the spirit of the thing. A better prompt would be "Guess where this photo was taken, do not look at the EXIF data, use visual clues only".