If you haven't tried it yet, perhaps the mobile app I've been having the most fun with this year is Reface. The mobile application allows you to take your face and replace the face of anyone in another photo or video within their database.
Why Is It Called Deepfake?
Deepfake is a combination of the terms Deep Learning and Fake. Deepfakes leverage machine learning and artificial intelligence to manipulate or generate visual and audio content with a high potential to deceive.
The Reface mobile application is simple to use and the results can be quite funny. I'll share some of my results here. Side note… they're not very deceitful, just embarrassing, horrifying, and hilarious.
Are Deepfakes More Frightening Than Funny?
Unfortunately, we live in a world where disinformation is prevalent. As a result, deepfake technology is one that may not always be utilized in something as innocent as making me dance or star in a movie… they could be used to spread disinformation as well.
Imagine, for example, images, audio, or video footage that utilizes deepfake technology to set up a politician. Even if it's identified as deepfake, the result could travel at the speed of social media to manipulate voter opinion. And, unfortunately, a significant percentage of voters – while minimal – could believe it.
Here's a great video from CNBC on the topic:
As you might realize, regulators and detection technology is becoming quite popular to attempt to fight deepfake technology. No doubt it's going to get interesting…
How Can Deepfakes Be Used for Marketing?
The technology to generate deepfake media is open source and available throughout the web. While we see it in modern film (footage of Carrie Fisher from the 1970’s was used in a deepfake in Rogue One), we haven't seen them in marketing… but we will.
Trust is critical in any relationship between the consumer and brand. Aside from legal ramifications, any business looking at deploying deepfake technology in their sales and marketing efforts is going to have to step lightly… but I do see opportunities:
- Personalized media – brands could generate media for the sole purpose of having their customers insert themselves. Imagine fashion designers, for instance, enabling a person to insert their face and body likeness into a runway video. They could see how the fashion looks visually (in motion) without ever having to try on the outfit.
- Segmented media – recording and editing videos can be especially expensive and brands are paying more and more attention to the representation of demographics and cultures that are depicted. In the near future, a brand could record one video – but utilize deepfake technology to segment the message to represent different demographics and cultures within it.
- Video Merge – brands could have their sales representatives or leaders star in videos that are deepfakes but that are personalized to communicate directly with a prospect or client. This type of technology is already available with a platform Synthesia. While I believe brands should disclose the deepfake, this is an eye-catching method to speak directly to each person personally.
- Translated media – brands can utilize influencers across languages. Here's a fantastic example of David Beckham – where his likeness will attract attention, but the message is properly translated. In this case, they use other voices and deepfake technology for mouth movement… but they could have also used deepfake to replace the audio.
In all of these examples, the deepfake isn't there to deceive but to improve communication. It's a thin line… and businesses are going to have to be careful walking it!
Let's end this on a good note…
Disclosure: I'm using my affiliate link for the Reface app. I highly recommend the paid version which provides a ton of additional media to mess with.