The media agency Reuters and an AI startup called Synthesia today they unveiled a new project they collaborated on: it uses deepfakes technology to generate automated news in real time.
Designed as a proof of concept, the system takes football match score data in real time and generates deepfake news reports complete with photographs and videos.
Synthesia and Reuters use a neural network that mixes news from the network and pre-recorded elements to develop a hyper realistic anchorman that provides updates on results and news.
To each his own TG
The big idea here is that you could have, for example, 20 different “live” videos streaming showing the same “person” announcing real-time score updates for each of the teams in the soccer championship.
Reuters pointed out that the deepfake news prototype is only a prototype and not necessarily a feature that it plans to implement.
The global manager of the company's core information products and services, Nick Cohen, said in a note: “Reuters has long been at the forefront of exploring the potential of new technologies to deliver news and information. This type of prototyping is helping us understand how AI and synthetic media can be combined with our real-time photography and reporting feeds to create new types of products and services.”
Deepfake news: the implications
Applications of this technology (assuming we end up developing a news program hosted by a "person" indistinguishable from those in the flesh) there could be many.
Aside from media coverage, it's easy to imagine airports full of monitors showing the same face with updates for flights from different airlines, or any updates on request for a wide variety of topics.
What is true, what is false?
On the other hand, we may be headed towards a dystopia where the massive use of artificial intelligence and a face people can trust become the main deciding factor in whether the public considers something fake or not. What once was "it's true, I read it in the newspaper", and later "it's true, I heard it on TV" could be "it's true, the AI journalist said it". Be horrified by the prospect of living in a future where humanity and AI are indistinguishable.
That said, Reuters could be a positive use case for the use of Deepfake technology. Paired with the right translation services, this tool could be used to generate emergency reports in locations where language barriers in both oral and written communications might otherwise complicate the rapid dissemination of information.