Designed to Deceive: Do These People Look Real to You?

Designed To Deceive

Source: Kashmir Hill and Jeremy White
Affiliation: The New York Times

This article describes how machine learning algorithms (specifically Generative Adversarial Networks or GANs) are able to create incredibly realistic-looking human faces. The article describes how GAN technology works and the implications of being able to create fake faces. It also discusses some of the flaws in facial recognition technology, including how biases in training data can lead to biases in algorithms. Finally, the article concludes with some tips for how to spot GAN-generated fake faces.

Keywords: Computer Science , Deepfakes , Tech Ethics