Moving forward in time, we need to use a lot of discernment about what is real and what is fake, especially what we see on TV, movies or other media.
If they can create a convincing fake video of a person, then how can we trust what we see on video? If they can copy a person’s voice patterns well enough, they can release an audio recording that says whatever they want them to say. The idea of Microsoft creating an avatar chat bot of a person based on their social media and other data is terrifying, given the volumes of data that they have on each person. Microsoft has access to so many things in a person’s life now. They own LinkedIn, Skype, Teams, XBox and Kinnect, OneDrive, Office 365, Sharepoint along with the Windows OS, Edge Browser, etc.
As someone who has gone through the Microsoft Professional training and certification for Data Science, I’m familiar with the how they use machine learning and image, text and video processing to do predictive analysis and to classify people for different purposes. With enough data, they can get very accurate results. There are some ways to thwart the algorithms, but they are getting more sophisticated everyday. I believe the end goal is a Chinese-style social credit system and that only approved information will be available. But even in that information, how do you tell a true video from a Deep Fake?