What should marketers be aware of about video fakes?

When you hear the word "deepfake," you might instantly think of fraud, ethically questionable or downright disturbing videos which have swept across the web over the past few years. As fake video content - and the AI technology behind these videos continue to get sophisticated, it's imperative that both creators and marketers in every sector understand the ways they function and how they can be applied to get and stay ahead in the ever-changing climate.

For the uninitiated Deepfakes are fake media which have been digitally manipulated to replace the appearance of a person convincingly, resulting in being able to make films that look and feel real yet aren't. Therefore, it's not surprising that they could receive a negative rap for distributing false information or exploitation of individuals' appearances.

As with any other technology it's in the way you use it. Innovative marketers and creatives have already begun using the technology of deepfake and ethically in a responsible and ethical mannerethically and responsibly -to make new artistic forms and tell fresh stories and improve their own video campaigns.

In this piece we'll look at some instances of how deepfake technology is being utilized to good ends and useful purposes, along with suggestions on how to experiment on your own with this technology.

What's a deepfake?

A deepfake is a film, or audio recording, of someone whose facial or body has been digitally changed. Deepfakes employ AI to create likeness, using pattern recognition to identify facial tones, expressions and movement.

Some other definitions for a deepfake can include synthetic or artificial media, or artificially-generated content.

 A short history of deepfakes

The creation of Generative Adversarial Networks (GAN) began the trend of realistic deepfakes back in the year 2014. GANs comprise two AI agents which create fake images and identify forgeries, allowing the AI to develop over time.

The creation of Deepfakes is also possible by using a deep-learning computer network known as a varial auto-encoder (VAE). VAEs can be trained to encode images in low-dimensional representations of a subject, and later decode those representations and convert them into images that move.

The term "deepfake" wasn't coined until 2017, and in 2018 media massed alerted the public about deepfakes and viral videos, with the first deepfake videos featuring Barack Obama and Donald Trump appearing across social media.

Deepfakes are also used for purposes which are becoming more useful to ordinary marketers and not only hackers or online trolls who try to spread false information.

How do deepfakes work?

Machine-learning AI is a crucial element of creating a fake. Deepfakes use this technique to detect trends in images and information.

To make a deepfake video, a developer needs to feed these machine learning algorithms with hours of real footage. This then trains the deep neural networks to identify the patterns in tone, facial expressions, and more. The next stage is to integrate these learnings with images.

It's not difficult to make a fake, simply existing videos or audio of the person you want to replicate. While it could seem difficult at first making a fake, it doesn't require complicated tools - only basic graphic design knowledge and video editing skills.

Examples of artful video deepfakes

Marketers are at the beginning stage of using deepfakes and other AI technology to help with video marketing and digital. These fake examples won't necessarily fit in a marketer's toolbox at this point however, they demonstrate the capabilities of these AI technologies right now.

 1. Chris Shimojima's "Dolche Big Man"

This breathtaking Staff Picked music video from director Chris Shimojima takes deepfake technology and flips it over its head, using faces from 14 different performers (and 40 contributions) to tell a single tale. The result is a stunning, unexpected combination of tech as well as human voice.

 2. David Beckham's numerous languages

Malaria must die used AI to influence soccer superstar David Beckham to speak in 9 different languages. The campaign leveraged fake technology to make a big splash and dramatically increase their campaigns' reach.

 3. Salvador Dali's museum greeting

It took more than 1000 hours of machine learning to enable the Dali Museum to create their deepfake version of Salvador Dali precisely right. This is a new approach to museum visitors that gives them a new perspective: they get to learn about the art of Salvador Dali from him himself!

3 everyday uses for deepfake video technology

While some applications of deepfakes might be out of the common marketer's grasp however, there are a variety of creative and interesting ways to use the technology of deepfakes in your job.

• Repair sloppy lines in posts: If you have even cursory editing experience You've experienced the struggles and frustrations of putting together polished audio from a more informal interview. If your interviewer misspoke, or didn't respond with the full length of a phrase, using the technology of deepfake to fill in blanks is a great way to keep your post-production procedure moving forward with no need to reshoot. (Just ensure you have the permission of your subject first, of course!)
   
  
• Customize videos for customers in a mass scale Marketers are able to implement an easy way to personalize their welcome or product videos with names of prospects or their company names. You only need their names, and some sound from the camera to use deepfake technology in incorporating it to the video.
   
  
• Localize your videos Deepfake technology introduces a new world of easy translation. Instead of using subtitles, artificial intelligence can insert translated, spoken audio, either created from an audio bank or from the voice of the actor who recorded it.

New technology, opportunities for new technologies

It's impossible to know for certain what the future of AI can provide however one thing is certain: deepfakes won't go away. Similar to other AI-powered technology (chatGPT perhaps? ) Anyone who is willing to experiment with deepfakes with eyes wide open to the potential for pitfalls are well-equipped to succeed in the evolving video landscape.