Hoax birth certificates. Faked movie posters. Imaginary computer products. It's become ridiculously easy to create passingly-believable hacks of just about any image. The same will soon be true of video. It's time we have some new rules for dealing with images as evidence.
I don't mean legal guidelines, per se, although it's likely that the era of defendants being convicted on the basis of photos or videos will be drawing to a close soon. I mean the rules that we use in our day to day lives to grapple with the flood of information, online and off, or the common sense heuristics we go by to avoid being lied to. In a world of easy manipulation of images, it pays not to be too credulous.
Here's a draft of five rules for the photo-fakery era. I welcome suggestions, corrections, and updates:
- A single photograph is evidence of nothing. No matter how much you want to believe (in aliens, in Apple iPads, or in foreign-born U.S. Presidents), a single image is no better than a nicely-illustrated urban myth. This doesn't mean that it's a lie, only that you should trust the image no more than you'd trust someone just describing it in words.
- Multiple images from the same source is no better. Anyone who can fake a single photo can fake a series of photos.
- Just as with photos, a single video or source of video should be regarded with extreme skepticism. It's getting easy to do professional-level composites with inexpensive or free video tools that run on even modest laptops. It's not yet as easy as working with Photoshop or the GIMP, but that goal gets closer with each passing year. Moreover, the relatively low-resolution video images typical of even new cameraphones make some kinds of hacks simpler, especially when shown via Flash video. If you can't clearly read a speaker's lips, changing what they have to say becomes a matter of a small edit of the soundtrack. This is something of an artifact of the technology of the moment, however; as HD becomes more prevalent, along with 4G or faster upload speeds, it should temporarily get harder to hack video.
- Conversely, the more sources, the more believable. This comes down to human behavior--when you have lots of people involved in a hoax, the faster it will be revealed. As Benjamin Franklin said, "three can keep a secret if two of them are dead." Even well-disciplined cadres and true believers tell their friends and spouses; the less-well-disciplined will brag on Facebook.
- With technology as well as people, the more diversity of sources, the more believable. If there's video of, say, a UFO, there should be photos, too. With consistent GPS tags. And people Twittering about seeing it. And mobile phone records of a burst of phone calls in the area as people call to say "you wouldn't believe what I'm looking at right now!"
Realistically, you won't always have situations where a wide array of sources and materials will be readily available for a given item or moment. So ask yourself: is the person (or group, or TV network) showing you the images or video trying to get you to do or believe something that's in their interest? Or, more bluntly, is there a reason why the person (etc.) might want to fool you?
Nowhere is this more important than with politics.
If you're annoyed by the "birther" churn, get used it--this kind of political hack is here to stay. It's easy and effective. Cheap digital tools make the work of faking official documents, "candid" images, and behind-the-scenes videos readily possible, even for rough amateurs.
Moreover, the hacks don't have to convince skeptics--they only need to strengthen believers. Faked materials just need to be convincing enough to cause doubt in the minds of people already inclined to believe a lie. For people trying to undermine political opponents, uncertainty is both easy and useful. Imagine if the hoax Obama birth certificate had been produced in October of 2008, instead of August of 2009: it's all too likely that the chaos surrounding the document could have cut his percentage in closely-contested states.
Would it have been enough to throw the election? Possibly, possibly not. We'll likely have parallel cases to study in the U.S. in next year's Congressional election, and almost certainly in the Presidential election in 2012. Don't be surprised to see similar political hacks showing up during elections elsewhere in the Internet-dense parts of the world. Our ability to evaluate what we see and read will be put to the test, time and again.
In a Photoshopped world, only the skeptical eye prevails.
Read more of Jamais Cascio's Open The Future blog.
"JCKenya" by Jamais Cascio, Creative Commons licensed.