No, It's Not A Deepfake. Yes, The Language Matters

The greatest trick the devil ever pulled was convincing the world he didn’t exist.” -- Verbal Kint / Keyser Soze, "The Usual Suspects"

There was a moment, back in 2016, when we had a common understanding of the term “fake news.”

Per Wikipedia, “fake news… is a form of news consisting of deliberate disinformation or hoaxes spread via traditional news media (print and broadcast) or online social media.”

Here’s an example: In July 2016, a website called “WTOE 5 News” declared that “news outlets around the world” were reporting  the Pope had endorsed Donald Trump. He hadn’t, they weren’t, and “WTOE 5 News” was not, in fact, a news channel. Deliberate misinformation. A hoax.

Fake news is scary because it’s deliberately designed to be so shocking our natural instinct is to share it. And it can spread around the world like wildfire, with the truth having no hope of catching up.



But then…

January, 2017. President-elect Trump has a heated exchange with a CNN reporter at a press conference -- at the end of which, he hits on a stroke of genius. Pointing from the podium, he tells the reporter, “You’re fake news.”

Three years on -- at least in certain circles -- the term's definition has been expanded to mean any media you distrust, particularly mainstream media. For Trump, it means any media that is critical of him. It no longer just refers to the particular brand of outrageous falsity that it once did.

This is useful if you actually do peddle misinformation. Nowadays, if you call Alex Jones “fake news,” you’re lumping him in with CNN and The New York Times.

If “fake news” is the devil, Trump’s co-option of it has led us to believe that it doesn’t exist.

After the State of the Union last week, Nancy Pelosi famously tore up her copy of Trump’s speech.

Two days later, Trump tweeted a video that had edited the speech-ripping to intersperse it with footage of the Tuskegee Airmen being honored, or military families being reunited.

Over on our sister publication, Digital News Daily, Karlene Lukovitz reported that “Facebook, Twitter Refuse To Ban Trump's Deepfake Of Pelosi's Speech-Rip Moment.”

Yes, the video was chronologically fake, making it look like Nancy Pelosi was snubbing the stories of the Americans being honored, rather than snubbing the speechgiver. But no, it was not a deepfake.

In a similar fashion, last May a video of her got circulated that had been slowed down to make it look like she was drunk. Despite this dramatic headline on CBS News: -- "Doctored Nancy Pelosi video highlights threat of 'deepfake' tech" -- that video also was not a deepfake.

A “deepfake” uses artificial intelligence to generate video of someone doing or saying something they never did. Here’s an example: Trump, Obama, Trudeau, Putin, and other world leaders singing John Lennon’s “Imagine.”

Deepfakes have only been technologically possible for a few years, but they’re improving dramatically.

In December, Timothy Lee from Ars Technica decided to play around with the tech, and was able to generate this video swapping Mark Zuckerberg’s face with the face of Data from “Star Trek.” It took two weeks, and cost just over $550.

If fake news is scary, deepfakes are terrifying. Not only are we rapidly approaching the point where anyone can create convincing video of anyone doing anything, we are rapidly approaching the point where everyone has plausible deniability of everything they’ve ever done -- even if there’s video evidence. All you have to do is claim it’s a deepfake, and BAM! reasonable doubt.

Lukovitz and CBS News were not intending to be inaccurate with their use of the term. My aim here is not to single them out, but to make a broader point: In a world already full of misinformation and lies, definitions matter.

Fake news has already convinced us it doesn’t exist. If deepfakes succeed as well, we are lost.

Next story loading loading..