NVIDIA founder and CEO Jensen Huang: Generative AI will bring accelerated momentum to creators

Source: Gyro Finance

NVIDIA founder and CEO Jensen Huang recently stated at the Cannes International Creativity Festival on the French Riviera that generative AI will bring accelerated momentum to creators of all types of content in all walks of life.

Talking to Mark Read, CEO of global marketing and communications services company WPP, Jensen Huang said: “We are now able to enhance the creative process like never before through content generation, and it can be text, images, 3D, video, any form of content. "

Jensen Huang and Read backstage at Cannes Lions International Festival of Creativity

Thousands of creators, marketers and brand executives from all over the world participated in the Festival of Creativity. At the event, Huang Renxun outlined the impact of AI on the $700 billion digital advertising industry. He also talked about the ways AI can improve the capabilities of creators and the importance of responsible AI development.

Huang Renxun said: "Although we can now generate content on a large scale, unlimited content does not mean unlimited creativity. We must guide AI with our own ideas to generate content that conforms to values and brand tone."

Not long ago, Jensen Huang gave a keynote speech at COMPUTEX. The speech mentioned that NVIDIA and WPP announced the joint development of a content engine based on generative AI and the NVIDIA Omniverse platform (for building and running metaverse applications).

The driving force that promotes the development of the era of generative AI

NVIDIA has been driving innovation in graphics technology for 30 years and has been at the forefront of the AI revolution for nearly 10 years. With expertise in both graphics and AI, NVIDIA is uniquely positioned to advance this new era of generative AI applications.

Huang Jensen said that the "big moment of modern AI" can be traced back to an academic competition in 2012. At the time, a team of researchers at the University of Toronto led by Alex Krizhevsky demonstrated that it was possible to use NVIDIA GPUs to train an AI model that outperformed all previous computer vision algorithms in object recognition.

Developers have since taught neural networks to recognize images, video, speech, protein structures, physics, and more.

Huang Renxun said: "People can learn the 'language' of everything. After learning this language, they can apply this language, and the application of this language is 'generation'."

Generative AI models can create text, pixels, 3D objects, and lifelike motion, giving professionals the power to bring their ideas to life more quickly. The user is like a creative director working with a team of artists, guiding the AI model with hints and fine-tuning the output according to their own ideas.

“You need to give the machine the feedback you would a top-notch creative director,” says Read.

Jen-Hsun Huang emphasized that these tools will not replace human creativity, but will enhance the skills of artists and marketing professionals, helping them create content that meets the needs of customers for different audiences at a faster speed and in multiple formats.

"We're going to democratize content generation," he said.

Using AI to reimagine the way we live, work and create

The main benefit of generative AI for the creative industries is its ability to scale up content generation, quickly producing text and visual proposals that can be used in advertising, marketing and film.

Huang Renxun said: "In the past, it was necessary to create hundreds of different advertising program options, which would be retrieved on the media. In the future, there will be no retrieval, billions of different advertisements can be generated, and each of them meets the requirements of the media. The tone of the brand and fit the brand perfectly.”

To be used by professional creators, these AI tools must also be able to produce high-quality visuals that meet or exceed the standards achieved by content produced through traditional methods.

And the foundation of all of this is a digital twin that can realistically simulate real-world physical assets. The NVIDIA Omniverse platform creates stunningly realistic visuals that accurately represent the physics and materials of images, videos, 3D objects, and immersive virtual worlds.

In this regard, Huang Renxun said: "Omniverse is a virtual world. In this virtual world we create, AI can learn how to create AI based on physics and based on physics."

"Thanks to an interface called 'USD', this virtual world can extract assets and content created by any tool," Jensen Huang said. The USD he mentioned refers to the Universal Scene Description Framework for 3D Collaboration. With it, artists and designers can combine assets developed using common tools from companies like Adobe and Autodesk with virtual worlds developed using generative AI.

NVIDIA Picasso is a foundry of custom generative AI models for visual design launched earlier this year. It also supports premium image, video and 3D generative AI capabilities developed in collaboration with partners such as Adobe, Getty Images and Shutterstock.

Huang Renxun said: "We have built a platform that allows our partners to use data authorized by companies such as Getty, Shutterstock, and Adobe for training. They all respect content owners very much. The training data comes from this, and the output income is also It will be given back to the creators.”

Both Read and Huang said that, like any groundbreaking technology, it is critical that AI is developed and deployed with careful thought. Technologies for watermarking AI-generated assets and detecting whether digital assets have been altered or counterfeited can help achieve these goals.

Jen-Hsun Huang said: "We should invest as much energy in AI safety as in AI function. In the advertising industry, safety refers to brand consistency, brand integrity, correct tone and authenticity."

Working together to create a digital advertising content engine

As a leader in digital advertising, WPP is embracing AI as a tool to drive creativity and personalization, helping creators across the industry craft great content that reaches their target audience.

Huang Renxun said: "From the creative process to the client, there must be an advertising company that understands this technology. The entire process in the middle requires human participation, so it is necessary to understand the requirements of the brand to be presented."

Through the Omniverse Cloud, WPP's creative experts can use a brand's specific product design data to build physically-compliant digital twins of products. This real-world data can be combined with AI-generated objects and digital environments licensed from partners such as Adobe and Getty Images to create virtual scenes suitable for marketing content.

Huang Renxun said: "There is no doubt that WPP will become an AI company and will build an AI factory that can input ideas, ideas and prompts and output content."

Powered by responsibly trained, NVIDIA-accelerated generative AI, the content engine will increase the speed and efficiency of creative teams, helping them deliver on-brand ad content quickly and at scale.

Jen-Hsun Huang concluded by saying: "The content that can help customers generate this way is limitless. In the past, it was necessary to create hundreds of content examples for a certain brand or a certain campaign, but in the future, billions of content can be generated for everyone."