Deepfakes: Is it good or bad? 

Deepfakes are a kind of artificial intelligence (AI) media in which a person's likeness is substituted for another person's in an already-existing image or video. Combining "deep learning," a subset of artificial intelligence, with "fake," is how the word "deepfake" is created.

As a machine learning doctorate candidate at the University of Montréal, Ian Goodfellow created deepfakes for the first time in 2014. (At Google's AI research center, DeepMind, he works as a research scientist nowadays.) His machine learning approach, known as generative adversarial networks, or GANs, is what allowed him to produce deepfakes.

A deep learning computer network known as a variational auto-encoder [VAE] combined with a facial recognition algorithm allows an artist to create a deepfake film by swapping out one person's face with another. VAEs are trained to convert low-dimensional representations of pictures back into images by encoding and decoding the representations.       

For example, if you wanted to transform any video into a deepfake with Oscar-winning movie star Leonardo Dicaprio, you’d need two auto-encoders — one trained on images of the actor’s face, and one trained on images of a wide diversity of faces.

Deepfakes can have advantages and positive aspects. For example, when a student creates a presentation about Issac Newton, Deepfake algorithms can animate historical photos and footage, allowing influential figures to give speeches and presentations as if they were in the classroom. The resulting videos are significantly more entertaining and participatory than traditional lectures or textbooks. Aside from that, Deepfakes became easy tool for hackers. It can be used for malicious purposes, such as making bogus content featuring people without their knowledge. This poses severe privacy concerns, prompting requests for further rules and legislation to protect people from deepfake media.

Apart from the good and bad impacts that deepfakes have, we as Onsmartizen have to be wise about their use! 😉