Photographs and video footage has been staged, doctored, manipulated and fabricated for almost as long as there has existed the technology to create photographs and video footage. Most often this has been done for entertainment purposes or to create genuine art. However, sometimes this has been done for nefarious purposes, to fabricate evidence, to make such footage appear to show something incriminating or damaging to someone or some group.
One of the most famous early examples of this is the so-called Cottingley Fairies photographs from 1917, which were staged using paper picture cutouts. At the time these photos were so convincing that they even fooled the famous writer Sir Arthur Conan Doyle.
The thing is, for literally a century this was something very non-trivial to do. It required skill, knowledge and a lot of work to create convincing fakes. Photographs, being just single still images, have always been easier to fake than live action footage, but even faking photographs has always been difficult, and quite often has left telltale signs of fakery, manipulation or staging.
It has always been easier when what's being depicted is something non-human, like purported alien spacecraft, cryptozoology creatures, etc. Fakery is much more difficult and involved when trying to depict humans, especially a celebrity.
Detecting doctored photographs and video footage is an entire forensic science, very often used in court cases involving such material. Just like there are literally thousands of ways to fake such material, there are also thousands of ways to detect if such material has been manipulated or faked in some manner.
Modern technology, however, is throwing a spanner in the works, for two main reasons: Firstly, the ever-advancing AI algorithms are becoming better and better at creating photographs and even videos that are literally indistinguishable from the real thing, no matter which method you try to use to detect fakery.
Secondly, and perhaps even more crucially, it's allowing anybody to trivially create such faked pictures and video, without the need of any skill nor knowledge.
This means that anybody can create fake photos and fake footage, and even fake audio recordings, of for example celebrities, and it's becoming easier and easier, and more and more convincing, by the month. We have already reached a level where most of that material is literally indistinguishable from genuine footage, and it's only going to become more and more advanced in the near future. And it requires literally zero skill and knowledge to create it!
There will very soon become a point (if we haven't already reached it) when anybody can create eg. defamatory footage of celebrities (such as politicians they don't like), and there will be no way of corroborating if the footage is real or AI-generated.
This, rather obviously, has to major implications: Firstly, the amount of fake claims about celebrities and other people will increase exponentially in the near future, causing an endless amount of fake scandals.
Secondly, it will completely erode the credibility of actual genuine footage. If there's literally no way to distinguish whether the footage is real or AI-generated, then even real footage can be dismissed as fake. After all, according to our justice system, the accused is considered innocent until proven guilty beyond reasonable doubt (as it should be). Well, in the very near future "this is AI-generated" will be reasonable doubt, even for genuine footage, if there's no way to distinguish between the two.
This means that innocent people will be involved in fabricated fake scandals, and actual criminals (eg. politicians who commit crimes or abuse their power) will be able to get away with their crimes by simply casting doubt on the evidence.
Ironically, the same technology that's making our lives better will also be abused to make our lives worse.
Comments
Post a Comment