AI: The Reality of Virtuality

One area where we are going to see tremendous innovation through artificial intelligence (AI) is in entertainment. Video games, for example, already employ AI in numerous ways, from creating armies of opponents who act like real people to building out realistic open worlds faster than a person could possibly create them. AI manages the laws of physics, moving everything in the environment like it’s in the real world. Rather than having to spend the majority of design time describing interactions between objects, programmers are now free to focus on building out a compelling reality. Programmers can set the initial conditions and then allow the AI to take of care of running the world.

AI is also disrupting existing technologies. For example green screen technology, which underlies many movie special effects, is currently undergoing dramatic changes through applications like Zoom in response to COVID-19. Zoom doesn’t require a green screen behind a person to screen out the background. AI determines what is presenter, background, and objects of interest that should be visible. Now workers at home can appear to be in front of the Eiffel Tower or on the moon. It won’t be surprising to see future enhancements like pajama-to-suit transforms so you don’t have to change clothes for a meeting.

The money behind blockbuster movies has also helped drive significant innovation, and AI is seeing the benefits as well. Consider leading-edge deepfake face swapping technology being developed by Disney. Deepfakes were the subject of the movie Looker  back in 1981, where models were scanned by computers and their digital models used in commercials. Today, deepfakes are on the verge of transforming how movies might be filmed. For example, a stunt person or stand-in could film a scene and the star’s face could be transferred in later. Or consider films containing scenes with a deceased actor. Dead Men Don’t Wear Plaid (1982) is the classic of this genre, using old footage from other films. With deepfakes, however, directors won’t be limited to existing footage; they will be able to create wholly new scenes starring actors long since departed. One intriguing application for deepfakes will be in inserting the viewer into the film. Imaging starring as the hero in your favorite action movie and having your significant other be the movie’s love interest. How about that for feeling like you’re part of the story?

And that’s just the beginning of what’s possible. Combine all of these technologies and you’re just a few steps away from achieving completely interactive entertainment with massive immersion and realistic world interactions, all in real-time. High-end video games are already very much like movies and have similar budgets and gross revenues – Call of Duty grossed $1.132 billion in 2019. It’s not unreasonable to expect to see a whole new breed of interactive game/movie – virtuality – in the coming decade.

As illustrated, AI can be implemented in many layers. The above examples are all what could be considered mechanical or procedural processes. For example, deepfakes still need a reference actor upon which to transfer facial expressions and manage body movement. The creative aspect of acting – the decision of what expression to show and when to show it – still depends upon a person.

Creative processes could be considered the next frontier for AI. Acting out a script. Playing music. Reciting poetry. And then the next level of creativity: Writing a script. Writing music. Writing poetry. While there has been progress in teaching computers how to perform creative endeavors such as composing poetry, there’s still quite a ways to go before they can guarantee high marks on a Turing Test. It’s clear AI is fundamentally changing the way we generate entertainment content. And with any major change in processes, there’s a shift in the tools we need to complete them. AI requires a different mix of processing and storage resources compared to traditional programming models.

At Dell Technologies, we understand the challenges of AI. After all, we just don’t design and build the infrastructure that underlies artificial intelligence: we use it ourselves. Working with NVIDIA, we have solutions that combine best of breed NVIDIA GPU accelerated compute with high-performance scale-out PowerScale storage. Our approach offers flexibility and informed choice including build your own options with PowerScale and the ultra-dense GPU accelerated PowerEdge C-series as well as prepackaged Dell EMC Ready Solutions for AI: Deep Learning with NVIDIA for organizations that prefer to buy solutions. We also offer advanced reference architectures such as the Isilon All-Flash F800 and NVIDIA DGX-1 servers AI reference architecture that reduce risk and compress the time needed for training and testing analytical models for multi-petabyte data sets.

Dell Technologies is at the forefront of AI innovation to help you make tomorrow possible. Learn more about your AI infrastructures needs from proof-of-concept to large scale production deployments from my NVIDIA GTC 2020 session here. And you can learn even more about how AI is changing the world, through our sessions at Dell Technologies World on October 21-22.

About the Author: Chhandomay Mandal

Dr. Chhandomay Mandal is the Director of Solutions Marketing at Dell Technologies. He leads solutions across artificial intelligence, analytics, business applications, VDI and HPC as well as industry-specific solutions for healthcare, media & entertainment, semiconductors and smart manufacturing. Prior to his current role, he led Dell’s all-flash storage solutions marketing efforts for desktop virtualization, server virtualization and private cloud. Dr. Mandal has been awarded 13 patents. He has a PhD from University of Florida, MBA from Indiana University, and BTech from Indian Institute of Technology.