In April 2018, a fake video showing a generated video of former President Obama went viral. It was the start of a new era of synthetic realities, in which our eyes and ears can no longer distinguish authentic content from synthetic ones. The threat of synthetic realities is that they break the link between authenticity and truth, which is fueled by concerns about the disinformation era we live in.

These generated synthetic images are possible because of a technique called GAN (Generative Adversarial Networks ). Generative Adversarial Networks (GAN) are a relatively new concept in Machine Learning, introduced for the first time in 2014. Their goal was to generate artificial samples, such as images, that are indistinguishable from authentic images.

An Airbnb-like rental website where the rent does not exist. By training a machine learning model to look at millions of pictures of bedrooms, millions of pictures of people, and hundreds of thousands of Airbnb listings, it was able to create a synthetic Airbnb listing.

Link : This Rent Does Not Exist

A resume generator that generates a resume of a person that does not exist. It uses a nonexistent person’s face and uses a TextgenRNN model that was trained upon millions of CV.

Link : This Resume Does Not Exist