Delivering on the Promise of AI

How Digital People rise above other technologies.

The Questions Every Decision Maker, Investor and Innovator Should Ask.

From Avatar to King Kong, advanced computer- generated imagery (CGI) and motion capture technology created hyper-realistic, high quality, animated experiences for movies, television, games, and augmented reality.

Technology now makes it possible for these same characters to interact, learn and express themselves in human ways – and in 1:1 interactions – to varying degrees. But adding artificial intelligence (AI) to CGI – and suggesting this creates Digital People – is far from what’s real or even possible.

Different approaches to creating Digital People will result in different outcomes – both for the user and for the brands and businesses investing in them. A fully autonomous Digital Person is radically different in experience and functions than a Digital Puppet. Seeing through the smoke and mirrors of high-tech marketing requires a clear definition and taxonomy for under standing the new generation of Digital People that will soon surround us.

Some are either 2D video recordings or relatively low-quality CGI characters with plastic-looking skin and non-life-like muscle structures. These are presented as so-called “interactive” performances with poor and disjointed animation. What is presented is a mash up of decidedly non-AI animation techniques – largely discarded by Hollywood and the Games Industry.

A new generation of Digital People, pioneered by Soul Machines, represent the next evolutionary advance in combining the quality of hyper-realistic CGI with a fully autonomous, fully animated CGI or digital character.

Download Whitepaper

01. Foundational Elements

CGI and Animation are two foundational elements used to create Digital People.

CGI has given us digital characters like the amazing Pandoran creatures from Avatar, Kong and a raft of cartoon characters and superheroes. Building highnquality characters that we as humans can relate to and engage with comes down to CGI expertise, time, and cost.

Animation in its earliest versions was the illusion of creating movement, but as techniques became more sophisticated and high quality animation became possible you could describe it as the art of creating life-like emotional connections. Think of the difference between the hand drawn cartoon sequences that Walt Disney used to make Mickey Mouse move for the first time compared to the actor driven, motion capture recording that became the basis of animating James Cameron’s Avatar movie.

Then there is bringing the CGI and the Animation together to create the performance and the interaction.

Low quality CGI and poor animation equals a terrible experience and represents failure in the movie and games industry. So why should we, and why would we, accept it as the right way to create a world where we are imbuing the machines we are creating with more human-like Artificial Intelligence?

02. Creating a HumanOS™

Different Approaches Result in Different Outcomes

Differing approaches to animation result in different outcomes for both those creating them, and the user experiencing them.

  1. High Quality Movie Animation using actors and motion capture suits and cameras – and based
    on pre-recorded story lines and a linear content flow – are high cost and time intensive. Hero animation per a character per minute is typically 20-30k for photoreal film, 30-40k for Pixar quality hand-animation, and 10-15k for AAA game titles. The result is high quality animation that takes significant time to create, but does not have the ability to be modified without re-recording, nor will it be interactive or deliver dynamic nonlinear content and experiences. The technique works well for movies and games but isn’t affordable or scaleable to meet the constantly changing interactive content of businesses andtheir brands, products andservice experiences.
  2. Blended or Procedural Animation is borrowed from the lower-cost end of the games industry.
    Here, lips and eyes move in a static image of a person or character. Procedural Animation is generated from simple rules-based systems and algorithms. Blended Animation takes pre-existing animation cycles and blends them together to make a performance. These approaches combine motion capture and embedded lip animation with pre-recorded head movement and an emotional expression like a smile on some form of loop playback. When the character needs to speak the video is interrupted, the face freezes and the lips start to move. This creates more of a Digital Puppet where the lips move speaking words fromscripts but the character’s face is rigid and unnaturally still and devoid of emotional expression.Again, the result is poor animation delivering a poor-quality customer experience. It can’t cross the uncanny valley and there is no emotional engagement which is the ultimate objective. It’s an experience that a customer is hardly likely to look forward to again – and an approach that degrades brands rather than enhances them.
  3. Synthesized (Deep Fake) Video is highly dependent on pre-recorded video
    content resequenced by AI algorithms to generate new video content which can be played back on demand. As seen in Samsung’s recent demonstration of Neon, it suffers from many of the same issues that Blended Animation does. While increasingly of a high quality, Synthesized Video is still pre-recorded content and not dynamic, responsive or interactive in the experience. The ethical use of Deep Fake technology has created a very high-profile debate which is a major inhibitor of its use and adoption by business.
  4. Autonomous Animation combines the quality of hyper-realistic CGI with a fully autonomous, fully animated CGI or digital character. Our human brain enables us to be autonomously animated in everything that we do. Soul Machines’ Digital
    BrainTM – driven by deep research into neuroscience, psychology and cognitive science – is the only way to autonomously animate digital characters. It’s a scaleable, cost effective way to animate for customer experience (CX) where face-to-face interactions need to be personalized, dynamic and unique in the same way each and every human interaction is.

Today, only Soul Machines’ Digital People harness the power of Autonomous Animation to create economics that meet the scale, quality, and interaction needs of users, businesses, and brands.

03. Understanding the Five Levels of Autonomous Animation

The intersection of AI and CGI (and eventually AR/VR) requires a new approach to animation – one based on the presence of a Digital Brain to autonomously animate Digital People in real-time. Digital People express responses and information based on both the context created by the user and the content being delivered.

In the same way there is a taxonomy for the different levels of Autonomous Vehicle technology, Soul Machines believes our industry needs a clear taxonomy for Autonomous Animation to clearly differentiate between the real and the fake. As a result, we have come up with five levels of Autonomous Animation in an effort to create a standard for measuring innovation in this field.

Actor driven, pre-recorded video or motion capture.

Enabling Technology:
Motion Capture Cameras

Possible Solutions:
Movie and Games Characters

Actor driven, pre-authored animation delivers pre-recorded movement based on simple triggers.

Enabling Technology:
FAQ text driven conversational database with pre-recorded voice content responses.

Possible Solutions
:
Digital Puppets

Video manipulation, and/or voice/text based partial face animation with encoded triggers.

Enabling Technology:
NLP – scripted conversational content Mirroring and simple emotion analytics.

Possible Solutions:
DeepFake.

Ability to learn from each interaction. Full human-like emotional responsiveness in facial animation including a conversational driven personality.

Enabling Technology:
Soul Machines™ Digital Brain V 1.0
NLP – scripted conversational content.
NLG – dynamically created conversational content combined with synthetic voice creation.

Possible Solutions:

Digital Employees, Digital Companions, Digital Influencers and Digital Celebrities

Full body motion control provides the ability to interact with digital environment. Creative behaviors based on learned experiences and agency to discover and set new intentions, plans, and goals. Ability to train through interaction.

Enabling Technology:

Soul Machines™ Digital Brain V 2.0

Possible Solutions:
Interactive 3D AR/VR characters

Independent Digital Heroes with strong semantic and contextual understanding of self’s actions.

Enabling Technology:

Soul Machines™ Digital Brain V 3.0

Possible Solutions:
Virtual people for virtual worlds, non linear story telling

04. Future of Customer Experience

CGI and animation’s new frontier is being redefined by a range of AI applications and the way this is relevant to the future of customer experience.

Digital Heroes – such as those created by Soul Machines – will fully integrate with Enterprise systems for data, conversational AI and other AI based applications, democratizing and enabling businesses and brands to achieve never-seen-before scale. Furthermore, they will embrace diversity and choice, ranging from hyper-realistic humans to cartoon characters.

05. Five Questions Every Investor, Business and Brand Should Ask

  1. How long will it take to create a diverse digital workforce, and how much will it cost?
    Hyper-realistic digital characters capable of being animated at the highest quality level are expensive, take time and require the best CGI artists in the world.
  2. What does “powered by AI” even mean?Today, every software company is using AI in some form. But few are combining the latest research in cognitive science, neuroscience and developmental psychology to create a new form of biologically inspired AI.
  3. What is the quality of the experience?Is it really just a Digital Puppet with low quality CGI and poor animation? Or is the experience offering something more?Is the Digital Human and its animation platform scaleable and dynamic?Even if you think it’s a starting point to learn and experiment – will it actually scale across your organization meeting the diverse needs of your internal staff and your clients?Will it scale and improve exponentially as AI applications like voice, content creation and image recognition get exponentially better and 5G networks provide the latency and bandwidth for super high quality delivery
  4. Is it a 3D Digital Character?Does it have a full body?Will the experience continue to grow and adapt as digital worlds and AR/VR become more accessible?
  5. Does it have a personality?Creating a face and attaching a conversational corpus and a TTS Voice gives you a talking head – a chatbot with a face – but unless you have the ability to define a personality that matches your brand values, in the role you need, in a way that appeals to your target client base, you won’t create a Digital Person.

06. Reading & Research

01. The Innovation Series : The Digital Brain
02. The Innovation Series : Autonomous Animation
Soul Machines – Changing the Face of AI
How Far is Too Far? | The Age of A.I.
When Roman Met Sam | Ep 1: This is a bit awkward!
Introducing Mia – A Soul Machines Digital Hero for Madera Residential

Next Steps

About Us

Learn More

Soul Machines™ is an astonishing AGI company whose unparalleled Human OS platform allows the world to leverage the full capabilities and goodness of human and machine collaboration in a responsive, relatable and unprecedented way. Soul Machines is the only company with a patented Digital Brain resulting in the most incredibly lifelike dynamically interactive Digital People.

Soul Machines works with the world’s leading brands to reimagine and accelerate AI as a platform for changing the customer experience.

Soul Machines is headquartered in San Francisco with R&D based in Auckland, New Zealand.