Aare Deepfakes a scary new reality or an opportunity in disguise? To date, we’ve primarily seen the technology used to make politicians appear to say comedic or even controversial things that they never have. Most recently many were shocked when actress Jennifer Lawrence’s body and voice was paired with actor Steve Buscemi’s face.
What is a Deepfake?
Miles Brundage, a policy expert who co-authored a recent report on the malicious uses of AI, said the term “deepfake” does not have distinct boundaries, but generally refers to a “subset of fake video that leverages deep learning […] to make the faking process easier.”
– The Verge
The Jennifer Lawrence video represents one of the more concerning directions Deepfaking can take us. So where is the positive? For those in charge of e-learning in the life sciences, it’s called “Native Dubbing”. Native Dubbing is being led by a company called synthesia.io and they describe it as a “..new method of translating video content that utilizes AI to synchronize the lip movements of an actor to a new dialogue track.”
”Native dubbing is a new method of translating video content that utilizes AI to synchronize the lip movements of an actor to a new dialogue track.-Synthesia
This technology creates exciting possibilities for localizing video content. Imagine a single video shoot translated to multiple languages versus one shoot done in multiple languages. Here are a few initial thoughts on how it can be used in life science training:
- Allowing a US-based sales trainer to share a product launch plan with her companies’ Japanese sales representatives or vice versa
- Making the insertion of green screen narrators/instructors in an e-learning piece more cost-effective and engaging
- Expanding the audience of a KOL, Marketing, or Commercial team videos meant for one local to another
- Helping sales leadership share messages and messaging globally at POAs, workshops, and national sales meetings (NSMs)
At CLD we believe that green screen narration can be used anytime and give e-learning deliverables a more personal feel while also helping walk learners through more complicated subjects like clinical data. It can also bring an added “wow” factor. Combine this with the potential to localize these videos in an automated fashion and sales trainers will have a powerful content option in the future. In the example below we even able to take video from an pre-existing etutorial (English) and natively dub it to Spanish:
Green Screen Example with Background Removed