Welcome to the Deepfake Universe
Because fake news is so yesterday!
In 2019 we live in a world where Deepfake videos and even images of people can be created by AI that are totally manufactured.
Meanwhile OpenAI’s GPT-2 is being hyped for its ability to write convincingly and deceptively.
In this new world AI can mimic human content in a variety of ways while potentially being used by bad actors and state-sponsored propaganda campaigns to influence public sentiment in a variety of ways.
In February 2019, we are witnessing an explosion of deception online. In an era when even Facebook refuses to be called a Media company, what exactly are deepfakes?
Invasion of the DeepFakes
Deepfakes are essentially deep learning fakes where via an AI based human image synthesis technique. It’s used to combine and superimpose existing images and videos onto source images or video using GANs (generative adversarial network) tech.
So this combination of existing and source videos results in a fake video that shows a person or persons performing actions or saying something that never occurred in reality. In 2019 we are also seeing an explosion of fake faces whereby AI is able to conjure people that don’t exist in reality which has a certain creep factor.
Fascinatingly, NVIDIA Software Engineer Phillip Wang has created a website demonstrating the ability of a technique using generative adversarial networks (GANs) to generate images of fake people.
You can check some of the faces out here.
- Thispersondoesnotexist.com generates images based on a new method of using StyleGANs developed by NVIDIA, which makes it possible to train a system to build high-quality artificial images with up to 1024 x 1024 resolution.
- If you keep refreshing you will get a sample of these “fake faces”, whereby digital humans and fake personas could become bots on the internet that could influence us in a variety of ways. Here is a showcase of some of the faces (these people do not really exist).
As you can see there is no way to tell that these are not real people. Fake people and fake text are the latest potential fronts in the deepfake debate that is just beginning and is yet another way AI can be harnessed to weaponize and alter collective sentiment through digital trickery.
Essentially GPT-2 and GANs type tech will only get smarter. Russia’s interference in the 2016 election cycle was a wake-up call for many but it’s likely just the beginning. Digital humans, AI News anchors, virtual personas, anything is possible on the new internet. The market for online deception is ripe.
This Face Does Not Exist
Creating a deepfake reality is half-way to the matrix! Think about the potential misuse of deepfake videso, identities and fake news combined.
At first glance, the two rows of portraits at the top of this article just look like a bunch of average-looking people. The catch is, none of them exist.
- The rise of ‘deepfakes’ could open up a new front in the disinformation war.
- Deepfake infiltration would be relatively cheap and easy to weaponize.
- If AI-generated images, news and videos have reached a level where they’re often indistinguishable from real photographs, human content and videos, we are on the cusp of a new era of AI-information wars.
Not being able to tell what is real online is already severely problematic in an era of social media and Facebook, but it’s only going to get worse. Considering young people spend more time online than ever, they perhaps will become more acquainted with AI with each passing year with unexpected consequences on society.
When Deep Learning Augments Fake News
The danger of Deepfakes means another kind of cybersecurity menace threat where what’s possible is out-racing how to combat it. These generative adversarial network (GANs) developed by Nvidia, could easily be classified as a mis-use of deep learning if such regulatory bodies existed.
You might have seen this video on YouTube which by the way is a deep fake.
Obama could have said this, but he didn’t.
In 2019 not only can your identity be stolen, you could be framed, smeared and trolled online with deepfake videos that would be difficult to disprove. You could be scammed, frauded and left vulnerable online and the enemy would be AI.
The ease of using deepfakes implies a new kind of hypocrisy and illusionary content online is coming. This is no longer about fake news, but memes, spam and content that’s progressively becoming less human, for good or for ill.
- Fake text generators
- Fake imagery
- Fake video content
- Voice cloning
- Persona cloning
The combination of smarter AI will create very real human problems that go so far beyond fake news or hate speech, that Facebook will be unable to keep up.
Before I die I’ll be able to clone myself into an AI persona online to retain not just some of my deepest thoughts but my behavioral patterns for my descendants.
DeepFaking ourselves Down the Virtual Rabbit Hole
Meanwhile the internet will continue to evolve at the intersection of AI in a way where how humans “augment” themselves might be unrecognizable even to us. We are quickly speeding to a world of neural implants and being surrounded by not just robots, but AI entities (call them what you will).
- The Gen Z cohort is spending so much time “online” their reality is more inclusive of online events than face-to-face events.
- Misinformation will rise in an incremental manner where Facebook and Deepfakes will spawn a mythical world of immersion, entertainment and a surveillance capitalism that will be deeply disturbing for our mental health and political stability.
- AI regulation will lag and many dangers will surface due to new technological developments and the uses of new tools.