Deep Fake AI put to good use

On Apr 9, 2019, there is a somewhat unconventional media buzz across the world. The #MalariaMustDie campaign (https://www.malariamustdie.com) is being launched, with the slogan “Malaria must die, so millions may live”. You may wish to read about malaria (resources provided by the World Health Organisation) and join the campaign. The campaign face is David Beckham, and here is the lead video:

What is notable about this campaign is the use of Deep Fake AI technology. Deep fake content is content created by advanced AI technology. What it means in simple terms is that you can take a real video if a living person, usually a celebrity or a politician, and produce 3D models of how they move and change their facial expression when they speak. By analysing how they pronounce different sounds and combinations of sounds, the technology adjusts the video layer accordingly to create a seamless video in which the person will “say” anything that the producers want them to, and in any language.

See here an excellent description of how the video was created:

Deep Fake AI content category is a profoundly effective manipulation tool. It was introduced to the wider public in 2017 by researchers at the University of Washington who had produced a photorealistic former US President Barack Obama. In April 2018 this was repeated and taken further by examples and a walk-through of the process by Jordan Peele and BuzzFeed. In November 2018, the China’s Xinhua news agency announced that it has created, using AI, the world’s first virtual news anchor. A more extended coverage of the story is available from BBC.

Back to the #MalariaMustDie campaign, it is the first time that the technology is used in such a scale in the public domain with objectives different than research or intent to mislead and harm by producing disinformation. And for that reason it is worth mentioning. Some say that Deep Fake AI opened the Pandora’s Box. Other point that technology can not be blamed for the intentions and motives of the humans who use it and this is an old and probably everlasting debate (e.g. Dewey’s lectures on Ethics ca. 1900). While we are at this, we recommend — to those who might be interested in going deeper into the ethics domain — a more comprehensive presentation on the Ethics of Invention in this 2016 video of Sheila Jasanoff from Harvard’s Kennedy School speaking at Talks at Google.