Adam Chapman

Deep Fakes And The Future Of Fake News

Do you ever get the feeling that you can see what’s going to happen? Not due to some all-seeing superpower but because it’s so blindingly obvious. It’s like witnessing an overworked waiter trying to balance that last plate upon his arm only to watch it come crashing to the ground.

This is the feeling I have when it comes to what is known as “Deep fakes”. Now we discussed deep fakes on one of our podcasts earlier this year (hopefully returning soon) and the impact they could have but as sure as the sun rises and sets every day, technology continues to develop.

For those not in the know, I’ll give a brief outline of what deep fakes actually are. Deep fakes are created using a program called “Generative adversarial network” or GAN for short. Developed by Ian Goodfellow in 2014 this program creates new data sets from existing data.

What this allows is what’s being exploited by fake news connoisseur and horny men around the globe. By uploading thousands of pictures of a celebrity or public figure the program can create realistic looking footage of said person. In other words, it can make it seem like someone is in a situation they never were. 

Of course much like the rise of VHS (as it was cheaper than Betamax) elements of the porn industry jumped upon the chance to live out their fantasy of seeing whatever celebrity they choose, engaging in whatever their minds desire. This was of course before the porn sites out there had a clampdown on anyone uploading deep fakes but in this modern world, it’s clearly impossible to 100% eradicate anything.

The threat of deep fakes isn’t just solely about watching celebrities in embarrassing sexual escapades, politically the effects could be catastrophic.  

If fake footage of someone was to be teamed with fake audio developed on Adobe’s program VOCO, that does a similar thing as GAN but in regards to someone’s voice, you can basically create believable footage of anyone in the world, saying whatever it is you want them to say. Threats of war, revelations of fake evidence, character assassination, it’s all going to become possible, the fallout of which could change the political landscape as we know it. 

What’s even more worrying is how this can be exploited in this era of “Fake News”. When people are already sceptical of the media and are already disregarding evidence placed right in front of them, it’s going to be a hell of a lot easier for people like Trump to disregard genuine footage as just another deep fake. The footage doesn’t even have to have been made for the impact to be seen. Genuine footage disregarded and fake footage believed, old footage altered to fit a modern narrative. It could technically rewrite history. 

Trump is already claimed fake news when footage of him from an interview with that shit rag The S*N came out earlier this year. The media left in complete bewilderment at how a man can claim he didn’t say something there is footage of him saying. It’s not going to be a fun place when he starts using the term deep fake. That is a world I am not looking forward to being a part of.

I’m not 100% sure what the trust levels are of the readers of The Bickering Press but I’d like to feel when we add our opinions to a topic it’s not through the fabrication of evidence or in the hopes of trickery. I’m very upfront with my opinions of why I think Brexit is a terrible idea, The S*N is the scourge of our nation’s media and why The Tories go against most of the things I stand for on a moral basis. Then again, I’m not sure why someone would want to create fake footage of me going against all this, not sure the reward would be worth the effort. 

So what can we do if we’re going to see an influx of deep fakes slowly filling our news feeds? 

As with anything technological you would expect the laws and technology to catch up to the new threats created. In the same way speed cameras became more adept as cars got faster and how drink driving tests became more about breathalyzers and less about reciting the alphabet backwards (something I’d struggle with sober) you’d imagine there are people already working on counter programs to be able to detect deep fakes. 

Indeed Facebook is already developing a machine learning program to counteract deep fakes and further steps are being researched to combat this ever-growing threat. It’s clear to see that as each side raises the stakes and develops newer technologies the other side must react to stay ahead.  

I’m getting images of an 80s training montage where the lowly hero must fight against all odds to overcome the enemy. Much like how Rocky Balboa overcame Ivan Drago in Rocky 4 this battle between deep fakes and counter deep fake technology could be the 2018 equivalent.  

The future looks fun doesn’t it…