Synthetic realities

Audio - 00:00

There’s a new kind of reality on the block. Generated and mixed realities are blurring the boundaries of “truth” and challenging how we value it.

As synthetic realities become more normalized in 2019, organizations should look past the drama and fear associated with them. Instead, they should hone new strategies to capitalize on their creative potential and manage the risk of unwittingly being featured in a synthetic reality created by someone else.

Key sectors:

  • Communications & Technology
  • Health & Public service
  • Travel
  • Retail & consumer goods
MIT’s Deep Angel software tool erased the astronaut from this image.
MIT’s Deep Angel software tool erased the astronaut from this image.

What’s going on?

In 2018, synthetic reality, most often generated by artificial intelligence (AI), reached new heights of sophistication, sparking controversy, and also fascination, about its creative possibilities.

In April 2018, a video apparently showing former President Obama calling President Trump names went viral. It was part of an explosion of scandalous photo manipulations created for hoaxes and propaganda using “deepfake,” an AI-powered, face-swapping technique

The technology behind face-swapping can now map any image style to another. It is already able to generate moving faces, bodies and objects from simple outline drawings. And it can swap the characteristics of fruit or animals, such as turning a video of a horse into one of a zebra.

Synthetically-generated brain scans are being used to help analyze real ones much more accurately.
Synthetically-generated brain scans are being used to help analyze real ones much more accurately.
This AI-generated artwork sold for $432,500.
This AI-generated artwork sold for $432,500.

A new tool from MIT can erase anything – or anyone – from old photos. And in November 2018, China’s state-run press agency Xinhua News Agency launched its first AI anchors – digital composites created from footage of human hosts that read news using synthesized voices.

Unsurprisingly, a common response to all this has been one of serious concern and calls for legislation. While we must be vigilant about potential pitfalls, we’re also seeing a wide range of positive applications of synthetic realities across entertainment, health care, mobility, security, automation, art and design.

In medicine, synthetically generated brain scans have been successfully used to train other machine learning models, with 14 percent better accuracy than ones trained on actual data. And it doesn’t come with any privacy issues because the data isn’t real.

The CGI-generated character “Lil Miquela” has 1.5M followers on Instagram.
The CGI-generated character “Lil Miquela” has 1.5M followers on Instagram.

Art project come Instagram influencer, CGI-generated robot Miquela has risen to fame on social media through her unique style and activism. Miquela has made her way into fashion magazines and is a musician/songwriter whose first single was released in 2018.

Understandably, fear of synthetic realities stems from the fact that they break the link between authenticity and truth, which is also fueled by broader concerns about the post-truth, disinformation era in which they are happening. While these are legitimate concerns that need to be dealt with and acknowledged, the emergence of synthetic media is simply following the same path as Photoshop and CGI: First it was scary; next it was familiar; then it was accepted. We expect a similar path here, once many of the questions and concerns are properly managed.

Lil Miquela is a CGI-generated pop sensation – this is her first single “Hate Me," produced by Baauer - Baauer, YouTube.

What's next?

Brand owners will need to consider what role a brand has, can have, or should have, in a world in which we question the authenticity of everything. For example, how can and should a brand evidence the value of authenticity when the fake might be more interesting or just as valuable?

Customers will increasingly expect brands to meet them halfway to supplement the realities they desire. Next, they will expect reality to adapt to them in real time without any conscious request. And the role of designers will be to set the stage on which these experiences happen.

Baby Dove UK advertising used AI to create a supposed 'perfect mum' from unrealistic depictions of motherhood in the media.
Baby Dove UK advertising used AI to create a supposed 'perfect mum' from unrealistic depictions of motherhood in the media.

Most leading AI providers will soon offer tools and libraries for building AI-powered natural language generation, image manipulation and other generative use cases. This will be in addition to those already available, such as generative graphics, photos, audio, video, text, code and materials like 3D printing and CRISPR.

The Adobe Project Cloak tool can remove people or objects from videos.
The Adobe Project Cloak tool can remove people or objects from videos.

In 2019, simulations will help break further ground in research and development and offer new ways to educate people and AI systems. And there will be opportunities for synthetic reality to make us better humans in the real world.

What you should do

01

Reframe “authentic”

In a world of synthetic realities, authenticity – something consumers value highly – will be more important than ever. Understand how to be authentic and communicate that authenticity effectively.

02

Be clear, be prepared

Continue to distinguish your brand by having a clear purpose and a platform that can be built from, rather than an entity that can be copied and manipulated. Be prepared for when things do go wrong, and have policies in place to handle them quickly.

03

Explore synthetic realities as a creative tool

Don’t be put off by the fear of being accused of bending the truth. Audiences already accept CGI in film; soon people will accept synthetic realities in everyday life. Once it’s familiar, no one will call it new or different or scary anymore.

04

Harness the power of AI-generated images for learning

Explore ways that this technology can be used to educate people and other AI systems in scenarios that were previously not possible. High-fidelity simulations offer a whole new way to test and train.