if you want to have some fun, there’s already a product for that:
I wonder if Randall’s been watching NileRed on YouTube?
Ha! One for which I didn’t even have to reach into deep memories of long-ago uni lectures to get the most obscure reference!
I’ve been saying this for a while. Photographs, or audio recording, or video, or whatever, isn’t to prove that someone’s not lying. They’re to prove that someone’s not misremembering. If someone wants to lie, all they need to do is not have any video at all.
We’re certainly no worse off than we were before video was invented, and we’re much better off in many ways.
Impressive. I don’t even know if you’re talking about Ea-Nisir or the Cottingley Fairies! I had to look them both up.
I would say the Ea-nasir Complaint Tablet is far more obscure, I mean, they even made not one, but two films inspired by the Cottingley affair. But I’m a huge Bronze Age nut, and a big fan of Irving Finkel, the cuneiform expert at the British Museum.
Now, who are Frances (9) and Elsie (16)?
The two young girls who were behind the Cottingley hoax.
“Good grief! It’s been nearly 4,000 years. Let it go already!”
– Ea-nasir
thx.
I knew the basics of Cottingley; didn’t remember the names of the girls, and my brain was too caught up in Ea-nasir to think to search the bloody obvious. D’oh!
If it was just the one complaint letter, maybe. But when the other letters found in you house are also complaints, you go down in history as the shiftiest copper ingot trader of all time…
True, but it became a popular meme and punchline last year on Reddit and Imgur, so not quite as obscure among a certain population. A typical example:
And tumblr. Which never forgets a meme
That is somewhat problematic if recordings are normally kept but are conveniently absent at some specific juncture; just ask the Secret Service.
I disagree with the stance taken both by Munroe and his imagined opponent.
Liars and lie detectors are in an arms race. Lying has been around far longer than humans have; camouflage is just a form of lying, after all. Lie detectors develop tools to see through the lies, whether by finding inconsistencies in the story, or conflicts with the evidence, or strange body language, and so on. Liars also use all the tools at their disposal, including technological ones, allowing them to lie in any media. Lie detectors have likewise had to up their game by spotting the particular artifacts introduced by the particular technique in question.
So it goes on and on. That doesn’t mean the bottleneck is the willingness to lie; if the lie detectors hadn’t improved alongside the liars, the liars would simply have free rein.
Furthermore, just because this is an ongoing arms race, it doesn’t mean that all developments are the same as they ever were. Deepfakes and AI generated imagery are–at least plausibly–the equivalent of nuclear weapons. On one hand, sure, it’s just one more thing that requires a response. And yet it’s such a powerful weapon that it feels like a step change and that we may have to reevaluate some old assumptions.
Also, Doyle was a dingbat and a bad lie detector, even for the time.
No, they’re not nuclear. In the olden days, if Bob said “I saw Jim robbing the store”, then we had to decide whether to trust him. And we had relatively primitive tools with which to make that decision: Mostly, based on how reliable we’d found Bob to be in our past interactions with him.
Now, we fast-forward to the present, or even beyond, to a time when recording video and deepfaking video are both easily within the means of the common person. Now Bob says “I saw Jim robbing the store, and here’s video of it”. We still have all the same primitive tools for evaluating how reliable Bob is. That hasn’t changed at all. We can, if we choose, completely ignore the video, on the grounds that it might be faked. So we’re no worse off than we were before, if we don’t trust Bob.
But what if I know, from my past experience, that Bob is an honest man, but that his memory, or his skill at recognizing people, is sometimes unreliable. That’s hardly surprising; all humans are at least somewhat unreliable in our memory and our recognition ability. In this case, if Bob has video, now we’re a lot better off than we were before. Because we know Bob is honest, we can trust that he’s not deliberately lying, and so we know that the video isn’t faked.
Or what if we know that Bob is often dishonest. He might have faked the video… but there might still be technological evidence of that fakery. Maybe there’s cryptographic documentation of when the video file was modified. Maybe there are temp files on his computer that show he’s been editing video. Maybe there’s a record of him buying deep-fake software. And maybe he’s not as good at he might be at covering his tracks. And so, if someone of unknown honesty shows a video, and we try and fail to find the tracks of deepfakery, we at least have some information, that’s not completely unreliable.
Deepfake video might be a big step, but it’s not as big a step as video, itself, was. The only way that deepfakery could result in a nuclear outcome would be if the technology existed, but it were not widely known that it existed. But we’re already past that hurdle.