“DeepFakes” between entertainment and misinformation
Technology advances with artificial intelligence at the head, and this generates the appearance of other phenomena. The case of Deepfakes is a good example.
What are deepfakes? Their name comes from the combination of deep Learning + Fakes and basically consists of manipulating human faces in video, using artificial intelligence.
Like all new developments, it provokes surprises and fears. They were born as jokes where we could see Barack Obama or Vladimir Putin giving speeches they had never said before, or versions of movies like “The Shinning” or “Terminator” starring by other actors. However, with the advances, those faces look more and more real.
Many specialists in communication warn us that it is a weapon that can be, in their own words “devastating” in terms of disinformation in the future.
The greatest fear is that fake videos will be viralized, and that these could influence, for example, the results of a political campaign. As always, it is not the technology that is dangerous, but it’s in “what” and “how” they are used.
We are coming into a new paradigm, understanding that what we see with our own eyes may not be real.
A bit of history
On December 28th 1895 in Paris, the Lumiere brothers gave the first cinema screening in history, and the image showing the arrival of a train terrified those present who thought they were going to be crushed by it.
On October the 30th 1938, Orson Welles (Citizen Kane, among others) during an episode of “The Mercury Theatre on the Air” broadcasted the War of the Worlds by radio. This caused many of those who were listening to believe that the earth was being invaded by aliens was real.
This was a paradigmatic case to demonstrate the massive power of the media in the first decades of the 20th century.
In both cases, the people who consumed the content believed that the fact was real.
In the first, because it was a new medium they didn’t know. In the second case, because they did not have the mindset, that the medium was capable of transmitting something that was not true.
“This medium transmits a speech, and that speech does not stand a chance of manipulation”.
As happened decades ago with the written word, with oral discourse and photography, humanity understood that they could be manipulated for a specific purpose (although today it is still the biggest motor of disinformation) it is necessary again, to realize that an audiovisual image can be false, in terms of a raw shot; manipulation through editing as old as the initial advances in cinema in the first decades of the last century.
As I am passionate about the topic and have been following it for some time, I decided to conduct a survey among my colleagues in the field of communication and audiovisual design to find out if they were aware of this issue.
I was surprised to find that almost 50% of the more than 40 people I consulted didn’t know what Deepfake was.
The alarming thing is that we are talking about people from that industry. Imagine in cases where people belong to industries farther away from the theme.
The first controversial cases
It did not take long to arrive, there was a case a few months ago in the United States that generated quite a stir, it was viralized in networks a video of Nancy Pelosi (Speaker of the House of Representatives of the Democratic Party) which she seemed to show difficulties of diction by being drunk. Even Donald Trump and Rudolf Giuliani (former mayor of NY) tweeted about the news and were disgusted by the state of Nancy.
But that wasn’t the truth: the video had been manipulated showing very specific moments from a press conference in which Pelosi was stammering, and although it wasn’t a Deepfake, because his face wasn’t manipulated, it was slowed down the time of the video.
It served to demonstrate the power and how easily something of this style can be viralized and how even people of high power can fall into the trap.
Virals and Entertainment Content
But not everything is dark in the future of deepfakes, they are a great resource for entertainment and content.
Even in movies like “Rogue One: A Star Wars Story”, the appearance of actor Peter Cushing, after 22 years of his death, playing again the character of Grand Moff Tarkin, showed us the power they have and the possibilities for the future.
Allowing us to imagine and make futurology, if we think for example of the videos that were made virals of Jim Carrey in The Shinning or Silvester Stallone in Terminator, could exist in the future movies where you can have different options of actors starring them? With just one click we could choose the actor we like the most to play that character.
There are also educational projects being developed with holograms, and if it is done with ethical responsibility, the possibilities can be infinite.
The State of the art
Back to the current time, Samsung is working on technology where you can generate videos with just one photo.
And in China the application “Zao” that was the most downloaded in 2019 in that country, allows you to put your face in famous scenes of Hollywood movies and series.
Even the porn industry took note of this technology, where they published fake videos “starring” by famous actresses and models, such as Gal Gadot (Wonder Woman).
Although this technology is still just being born and the videos are still quite precarious and you can see their fakes, not long before it is very difficult to notice that it is real and not.
Fortunately, there is also work on software that can detect videos that were manipulated, even as easy and simple as a Browser extension.
The advance cannot be stopped, but we can be more and more aware that all speech is permeable to be manipulated, and that it is our critical nature that has to prevail to challenge it and not fall into the trap.
