“The era of synthetic content is here” | Μy key takeaway from #ONA19 conference
What is synthetic content, you may ask. Well, as it turns out, the current level of Artificial Intelligence has allowed the creation of a whole new segment for the media market. Using initial corpora and training algorithms, we can now generate voices, text, photos, videos, and even…people.
It is entirely possible that within the next decade media consumers will start engaging more and more with synthetic content.
Editor: Radu Lisita*
SYNTHETIC PEOPLE ARE HERE
Miquela Sousa (@lilmiquela) is a 19-year-old Instagram influencer with 1.6 million followers. She posts about her Brazilian heritage, her passion for music and fashion. Brands like Opening Ceremony and Prada have worked with Lil Miquela, and her followers' database is growing day by day.
The only thing is, Miquela is not real at all. She is a digital robot, a computer-generated influencer. But people don’t really seem to care. The creator behind Miquela and her Instagram presence is anonymous, and it’s unclear who acts as her “voice” for the photo captions. Is he a human or is this another algorithm at work?
But the more interesting question is — can you imagine a social media landscape with tons of such influencers, customized according to your needs & preferences?
SYNTHETIC VOICE IS HERE
• Canadian AI startup named Lyrebird claims it can recreate any voice using just one minute of sample audio. You can already listen to Barack Obama, Donald Trump or Hillary Clinton’s fake voices: https://soundcloud.com/user-535691776/dialog
• Voicery.com is another AI voice service generator that delivers a product based on the input text. The product uses algorithms to create realistic artificial voices based on real people’s voices and is capable of conveying emotional undertones & controlling intonations.
The results are already quite astonishing, showing that in the nearest future it would be hardly possible to distinguish a fake voice from a real one.
SYNTHETIC MODELING
• David Beckham spoke nine languages in a new short film calling for an end to malaria. In the short film, Beckham appears to speak fluently in Spanish, Arabic, Hindi, and Mandarin. The voices heard in the video in addition to Beckham’s belong to men and women around the world who have been affected by malaria.
Using Artificial Intelligence video synthesis technology, the producers of the campaign created a 3D model of Beckham which was re-animated to make it appear as if all the voices featured in the video are being spoken by the former footballer. https://www.youtube.com/watch?v=QiiSAvKJIHo
• Now, can you imagine Mark Zuckerberg openly admitting to something many people already suspect him of doing e.g manipulating users’ data? “One man with total control of billions of people’s stolen data, all their secrets, their lives, their futures,’’ Facebook CEO says in a video. The provocative video is actually a ‘deepfake’ created by artists Bill Posters and Daniel Howe using artificial intelligence. https://www.instagram.com/p/ByaVigGFP2U/
The video has been posted to Instagram. The company previously said it would not remove this type of video, but how will Facebook be dealing with manipulated footage going forward?
• • •
The larger concern, of course, is how all of us (both corporate and private entities) will be affected in this new space of AI-generated content that looks shockingly realistic. What safeguards do the platforms have against bad actors willing to use this technology to cause harm or disruption? It’s certainly something that already needs to be addressed at this stage.
#DCNfall2019 #DigiComNet #WorldLearning #ExchangeOurWorld
Slides are scrapped from Amy Webb’s presentation on Future Trends: https://www.dropbox.com/sh/ugp70199v7198yz/AADR0qAblSJgJcwVSRU2ElYCa?dl=0
*Radu Lisita is digital and social media manager and he lives in Chisinau, Moldova. He is one of DCN’s 16 exchange participants. Find him on Facebook and on Linkedin.
