
Advertisers and agencies will one day use a new
form of generative AI (GAI) to create TV commercials, as technology increasingly brings non-traditional companies like Dassault Systemes into the industry.
Dassault's 3DExcite already provides
agencies with 3D digital images of cars -- including ones that don't actually exist -- to help build campaigns for automakers, but using GAI for TV commercials is still on the horizon, CEO Tom Acland
tells "Media 3.0."
The shift to TV will require much higher performance data and materials to drive customer attention, he explains.
GAI has significant implications for marketers
because it enables a new form of product storytelling ensuring authenticity and personalization.
Marketers showcasing a new car heating system for a brand, for example, require different
approaches in different locations across the world. Consumers in Sweden are likely to care about different weather-related capabilities than those who reside in the United States.
advertisement
advertisement
“It
matters where the consumer lives,” Acland explains, citing other factors such as “whether or not they have a family or dog.”
With updates made to its system, companies can
now use 3DExcite to generate accurate content about their products in any local setting.
They can also demonstrate any product feature without burdening the graphics department of the brand or
the agency.
At Nivida’s recent GTC 2024 conference for developers, 3DExcite showcased a scientifically accurate rendering of Hyundai's IONIQ 6, which appeared in alternating AI-generated
environments, as well as in a variety of trim colors.
The print-ready images pulled data from the car’s engineering specs, and rendered images of the car depicted in mountains and
on city streets using simple text commands that are familiar to anyone who has used ChatGPT.
Acland says the platform also uses voice commands, and shared a video image of the Hyundai
IONIQ 6 during a Zoom call. He instructed the program to create a background described as a "breathtaking" Swiss landscape in a cinematic afternoon setting.
The technology generated the
environment — featuring water, mountains and trees — that reflected off the car’s paint in a moving image.
The models were trained on Shutterstock images. Shutterstock took
the commercial data sets and used video technology to create a large GAI model, which created the scenes.
3DExcite took the scene and an image of the car and generated a background for
Hyundai.
“It will enable car manufacturers like Hyundai to tell a story about the vehicle’s features,” Acland says, adding: “You can show the same feature to different
people in different ways. It will help to drive programmatic advertising and, we think, help to build better products.”
Acland stopped short of sharing additional names of brands, but
said 3DExcite works with many automakers worldwide as well as in other industries such as aerospace, and with industrial equipment such as robots.