Star-studded celeb David Beckham participated in a charity video a few months back raising awareness for malaria. That isn’t unusual by itself, but the technology behind the video made heads turn.
In this Malaria Must Die campaign video, Beckham’s lips sync with people speaking different languages. I didn’t think it was a perfect syncing, but it was close enough that people freaked out.
How did Beckham’s face move like that? He clearly wasn’t mouthing along as other people spoke. What’s going on?
It’s called synthetic content, and it’s taking the world by storm.
In this video, the crew used 3D modeling and certain algorithms to match Beckham’s lips with the audio. While this was done to send a powerful message, this is an example of sythetic content used for good.
What happens when this technology is in the hands of not-so-good people?
What is synthetic content?
Before we freak out, it’s important to understand what synthetic content is. It’s most popular in videos, where editors traditionally have to spend hours of work adjusting a video. With synthetic content tech, these professionals can delete or add dialogue to speed up the editing process.
Sounds pretty mundane, right? Well, the issue is that you can devise completely false videos with synthetic content. And it’s not a blurry SIMs video, either. Because of the advancements in 3D technology, these videos look clean and realistic.
Synthetic videos are also called “deepfakes” and that’s a pretty accurate name. They’re hard to pinpoint unless you already know you’re looking at a fake. While it’s a great technology for marketing and video, it’s still concerning, and for good reasons.
How synthetic content works
Synthetic content can be used in a few ways. As we saw in the Beckham video, it can be used to alter visuals. However, it can also be used to regenerate fake audio that sounds like it came from a real person. Basically, the technology allows people to “rewrite” a video.
Once upon a time, the synthetic content algorithm needed 40 minutes of audio to replicate someone’s voice. But thanks to technological advancements, they need just a few minutes of audio.
It’s important to remember we see synthetic content every day, though. Photoshopping is a form of synthetic content that, while people generally don’t love it, we accept it.
But when you have the power to alter, say, a political speech into something else entirely, and you have an environment of angry folks on the internet, synthetic content becomes less acceptable. As the tech becomes more affordable, we’re seeing more bad players enter the field, too.
Why would you use synthetic content?
It’s easy to whip out the pitchforks when we hear about a creepy, new technology. I think it’s helpful to understand how synthetic content helps marketers before we all start panicking. I found that companies are using synthetic content for good in 3 ways.
-
Customer service
While it’s not the same as the synthetic content used in videos, brands are using similar algorithms for customer service. Chatbots mimic the speech of real people, answering basic customer questions to free up brands’ time. Although chatbots pretend to be real people, most of us know “Sally” is an algorithm attached to a database. This tech is pretty innocuous, and even helpful, for both customers and brands.
2. Personalization
Synthetic content is out of reach for SMBs and small agencies, but it opens up a lot of possibilities. Only enterprises will be able to implement it right now, but synthetic content has big implications for marketing and personalization.
For example, you could have a synthetic content host run your promotional webinars. Or you could send customers a seemingly personalized video message about their account, using their name and history (without being creepy, of course).
3. Save time
Time is a big, big factor in marketing. Marketers are tasked with to-do lists the size of Mount Rushmore, and they’re expected to do it with just two hands, nearly zero budget dollars, and an intern if they’re lucky.
Synthetic content opens doors to time-strapped brands. I foresee small agencies finally being able to do video content without the need to film lengthy, complex scenes. Videographers and video pros are already using it to cut down on editing time, which means this technology is saving everybody more time and money.
When it comes down to it, I don’t think synthetic content will replace marketers, but it will augment our capabilities.
This sounds creepy. What are the downsides?
Okay, synthetic content isn’t purely used for good. This technology is really cool, but it has 3 big downsides that are changing everything from user agreements to security.
1. Consumer trust goes out the door
People already distrust digital content. You read one breaking headline and the “fake news” accusations are flying not long after publication.
Today’s consumers are already skeptics. Can you imagine a world where we can’t tell reality from fiction? That could be our future with synthetic content.
There’s nothing wrong with creating synthetic content for entertainment purposes, like spoofing a movie, but when it crosses into reality, it could hurt people. You could take politicians’ speeches and make them say anything you want.
So yeah. Synthetic content can be really damaging. It mimics authority figures and warps reality. In a world where consumer trust is already so fragile, this could be a big hit to the digital realm.
2. Everyone needs to change their policies
Obviously, the internet isn’t going to stand for this. Thank god. But that means that platforms are coming up with solutions for these deepfake videos.
For example, Twitter is creating a policy to combat synthetic content. Google (which I’ll go into more in a sec) is also combatting synthetic content with new policies and initiatives.
While this is a good thing, it means there’s a whole industry cropping up around synthetic content and protecting the internet. That’s a good thing, but it does increase costs in unnecessary ways.
3. It can fool voice authentication
This one is really freaky. The problem with synthetic content is that it can almost perfectly match someone’s real voice. Thanks to text to speech technology, synthetic content could fool voice authentication methods.
That means synthetic content will be able to fool what we previously thought was a (fairly) trustworthy biometric tool.
How to combat synthetic content fraud
Good ol’ Google is leading the charge against fraudsters using this technology. While it’s still in development, Google is working on anti-spoofing measures. These will make it easier for users to tell the difference between real and synthetic content.
Industry leaders are kicking around other solutions, including:
- Watermarking videos that have been edited
Digital fingerprinting
Another way to combat this is to educate ourselves as consumers. Now that we know synthetic video is hitting the airwaves en masse, we need to be informed consumers. Think twice before retweeting a video, especially if it’s outrageous or featuring famous people. It just could be a fake.
Is the future synthetic?
I know this sounds scary, but we have to weigh the pros and cons of any new technology. As creepy as it can be, there are some very good reasons to use synthetic content. When it comes down to it, this is a tool. Tools can be used for good or evil, depending on the human being behind the keyboard.
As far as synthetic content and the economy around marketing, this technology won’t be stealing anyone’s jobs. If anything, synthetic content will speed up content production and give marketers time to focus on other tasks.
The question now is how the industry will adopt anti-spoofing measures to protect our customers. In the meantime, we can all become educated consumers and think twice before we overreact to online videos.