This doesn't really deal with the issues involved in "fake news". What it does is let you know with confidence that you have what someone is asserting - namely, that they say they took this photo at this time and place with that smartphone.
This doesn't touch on all the other ways of influencing the narrative that news organizations have. Editorial discretion is a huge factor. Someone high up in, say, the WSJ, can decide that manufacturing importance for a story is in their best interest, and now we see these facts brought to our attention instead of those ones.
It provides a baseline from which to build. If I have a news article that cites sources, and eventually those sources are verified in this kind of manner, then I can "know" something.
I think this is a cool project. But I dont think this fixes fake news. Like 2xlbuds says above, "I doubt the sort of person who would believe in conspiracies will take the time to understand what digital fingerprinting is" [1]
It's a decent start, and works for quite a few stories. However, it'll need to be used with other solutions in order to cover various other cases where this doesn't help. Such as:
1. Where it's the object in a photo that needs to be confirmed as real. For example, a fair few sites will take pictures/scans of magazine or newspaper articles, and this wouldn't help in those cases (since it only says the photo is of a real object).
2. Where the photo isn't taken with a camera. For example, a lot of entertainment journalism (like for movies and TV shows) and almost all gaming journalism requires capture cards to get the screenshots, not a camera. So something like this that could be adapted to these use cases would be nice to see as well.
3. And finally, where the image is real but used in a misleading way. Time and place stamps help here, but still leave room for manipulation by careful camera angles and what not.
But still, it seems like an interesting tool none the less.
I think this is trying to combat a phenomenon that isn't based on logic with something that is - I doubt it will work. Fake news is such a powerful tool because it plays into people's biases and reinforces their beliefs. I doubt the sort of person who would believe in conspiracies will take the time to understand what digital fingerprinting is.
I still think that it is incredibly important that we make tools for validation of primary source material easy to use and friendly for non-technical people.
Tech for faking video is getting more powerful day by day.
this is a pretty cool project, but the title is misleading at best.
tldr; So when you enable proof mode on your android phone, it grabs your pgp key(auto generated), gps, datetime, hash of photo, etc; and signs your picture.
This could be useful in court, or any time you need proof that the photo is 'real'. As in, you were there, at that time, etc.
It doesnt address taking things out of context [1], presenting only certain facts, etc. Those are the tools/weapons that I am worried about.
rant:
Am I the only one getting tired of seeing some technical tool which is essentially trying to prevent lying or bias? Or the fact that people believe ridiculous things if it confirms their worldview? This is not a technical problem imo... Look at lie detectors. If you can control your pulse, then you appear to be telling the truth. These days those devices are no longer trusted. I feel like these types of stories are lie detector 2.0(not this one, just the title). Sorry for the rant.
/rant
As far as I can tell, the app proves nothing, except that the data comes from you. At the end of the day, you're taking a bunch of data and signing it with a PGP key you made up. That data can be anything - it doesn't prove that you were actually there, or that it's a real photo.
Submissions are flagged by members; if the submission accrues enough flags, the '[flagged]' tag appears. Enough HN members thought the piece was inappropriate for HN. Off the top of my head, members may be feeling "fake news" meme exhaustion, or have come to the same conclusion you did, "the app proves nothing", and thought it wasn't worth discussing.
14 comments
[ 3.2 ms ] story [ 46.0 ms ] threadThis doesn't touch on all the other ways of influencing the narrative that news organizations have. Editorial discretion is a huge factor. Someone high up in, say, the WSJ, can decide that manufacturing importance for a story is in their best interest, and now we see these facts brought to our attention instead of those ones.
1. https://news.ycombinator.com/item?id=13729304
1. Where it's the object in a photo that needs to be confirmed as real. For example, a fair few sites will take pictures/scans of magazine or newspaper articles, and this wouldn't help in those cases (since it only says the photo is of a real object).
2. Where the photo isn't taken with a camera. For example, a lot of entertainment journalism (like for movies and TV shows) and almost all gaming journalism requires capture cards to get the screenshots, not a camera. So something like this that could be adapted to these use cases would be nice to see as well.
3. And finally, where the image is real but used in a misleading way. Time and place stamps help here, but still leave room for manipulation by careful camera angles and what not.
But still, it seems like an interesting tool none the less.
Tech for faking video is getting more powerful day by day.
http://graphics.stanford.edu/~niessner/thies2015realtime.htm...
https://www.youtube.com/watch?v=ohmajJTcpNk
tldr; So when you enable proof mode on your android phone, it grabs your pgp key(auto generated), gps, datetime, hash of photo, etc; and signs your picture.
This could be useful in court, or any time you need proof that the photo is 'real'. As in, you were there, at that time, etc.
It doesnt address taking things out of context [1], presenting only certain facts, etc. Those are the tools/weapons that I am worried about.
rant: Am I the only one getting tired of seeing some technical tool which is essentially trying to prevent lying or bias? Or the fact that people believe ridiculous things if it confirms their worldview? This is not a technical problem imo... Look at lie detectors. If you can control your pulse, then you appear to be telling the truth. These days those devices are no longer trusted. I feel like these types of stories are lie detector 2.0(not this one, just the title). Sorry for the rant. /rant
1. http://imgur.com/CMykxFC
Needs a timestamping sever
God know why the article was flagged, though.
Submissions are flagged by members; if the submission accrues enough flags, the '[flagged]' tag appears. Enough HN members thought the piece was inappropriate for HN. Off the top of my head, members may be feeling "fake news" meme exhaustion, or have come to the same conclusion you did, "the app proves nothing", and thought it wasn't worth discussing.