Attack of the clones: Should we fear generative video?

marie's avatar
marie
Share
A vintage television screen displays an image in triplicate of a person raising their hands to their face in fear. It looks like a still from a vintage horror movie. The word ‘play’ and a play video icon is located in the upper lefthand corner of the screen.

A new breed of AI platform is
pioneering the ability to create
high-quality videos without
cameras. Should we be excited
or worried?

Generative video means affordable access to more
polished content for creators, but actors are currently
striking to prevent this kind of AI from being used.
One fear, for example, is that the tech will take their
digital likeness, using it in perpetuity despite only
paying for a once-off 3D scan.

The tech itself is interesting. Joshua Xu, co-founder
and CEO of generative video company HeyGen,
explained that it can “replace the camera layer,
unlocking 100 times more creativity”.

But how does it work? Users record a short video of
themselves speaking, which trains an AI avatar clone.
Typing any script generates a video of the resulting
avatar delivering those lines in their own voice. No
filming required.

The AI then layers in music, graphics, backgrounds
and more to produce a polished video from just a
text script. “There have been tons of technology
advancements in the past six to 12 months,” said
Joshua, mentioning AI art generators and voice
cloning as specific examples.

Is generative video about to get its ChatGPT moment?

Joshua predicts a “ChatGPT moment for video” in
the next two years, with AI-driven video generation
becoming mainstream. But with this would come a
raft of potential misuses, including the cheaper and
faster creation of deepfakes (AI video footage created
using someone’s likeness without their consent).

What are the positive implications? For content
creators, it massively reduces production time.
An animated explainer video that takes weeks
to storyboard and animate could be generated
instantly. Joshua shared examples of HeyGen
creating five videos per week for YouTube without
any actual filming.

It also enables personalised video at scale. Sales
teams could generate customised outreach videos for
each lead by integrating AI video into CRM systems.
Education or training videos could feature different
avatars tailored to the viewer. According to Forbes,
the global generative AI video market was worth
around US$470 million in 2022, and is expected to grow
by 20 percent year-on-year between 2023 and 2030.

The downside of all this is potential loss of jobs. As a
striking background actor told the Washington Post, this
kind of AI is “erasing jobs from the market, and being
done so in a way that doesn’t financially compensate”.

As the GenAI market matures, it promises to
democratise quality video production for creators
and businesses. However, we must thoughtfully
navigate concerns around fair compensation,
likeness rights and potential job loss, lest the clones
instigate a rebellion against their masters.

To get more digital marketing tips and expertise, download our full Marketing and media insights report here.

Main image of a vintage television screen displaying an image in triplicate of a person raising their hands to their face in fear: Web Summit

Related
Image of Kabir Barday, founder and CEO of One Trust
Tech

The future of AI regulation: From data to algorithm deletion

OneTrust founder and CEO Kabir Barday warned that impending regulations may soon require AI companies to...

August 12, 2024 - 3 min read
Related
A photograph of a person (Daniel Yanisse, co-founder of Checkr) speaking onstage at Web Summit. They are sitting on a chair and wearing a headset microphone, while gesturing with their hands. The Web Summit branding is visible behind them.
Tech

Are people with criminal records an untapped resource for tech?

People with criminal records are often shut out of the labour ma...

October 24, 2023 - 1 min read