A design response to AI
Duration: 1 months
Role: Individual contributor
Tools: Cinema 4D, Photoshop.
Artificial Intelligence (AI) is a study in computer science that emphasizes the creation of intelligent machines that behave and react like humans. Currently AI’s used in speech recognition but the fear is that one day, computers will be able to out think humans. This project responds to a likely dystopian future where AI has caused significant damage to our society. A future where it is difficult to believe if the news is real.
I was interested in exploring the rising fear of “Fake News”. Fake News has become an increasingly popular media topic with our current president. Today, it is common to see on social media articles form false websites publishing incorrect written content about an event or individual.
But with AI, it is possible to create fake videos that resemble a realistic individual saying manipulated content.
Below are two research documentations that show how easily videos can be manipulated to portray whatever content is desired.
Deepfake is a technique for human image synthesis based on artificial intelligence. It is used to combine and superimpose existing images and videos onto source images or video using a machine learning technique known as generative adversarial network. Deepfake is often used negatively in the pornographic industry to superimpose celebrity faces onto bodies.
The above videos show how this technique can additionally be used on governmental figures to instill fear into the public. As this AI work continues to grow, I fear that there will be large consequences and lack of trust towards individuals that we rely on to survive as a society.
My design solutions is a physical object that is used to project a
‘3D watermark’ around the speaker. This object projects a UV light that time stamps and gives coordinates to where the individual is located. Viewers watching the speech from the internet would be able to toggle on/off the UV layer to confirm that the video has not been altered.
The UV light would not interfere with the speaker but give confidence to the public that what they are watch online is real. If the video has been manipulated, they would be able to see a gap in the ‘3D watermark’.
The viewer can see that this video is ‘fake’ and can act accordingly in their belief of the video.