Skip to main content

The results of a new study have detailed the four types of “digital humans” that we can expect to interact with in the near future: virtual agents, assistants, influencers, and companions.

There are four types of digital people that humans will interact with, including Kardashian-style celebrities that are custom-built to influence consumers and companions for elderly people receiving care in the community, according to a new study.

As a sales assistant, a corporate trainer, or a social media influencer, these digital people can mimic human behaviour and are used as computer-generated digital characters. Taking Lil Miquela as an example, she has almost three million followers on Instagram and is an entirely online influencer.

In the last decade, Soul Machines, one of the leading names in the animation software industry, has deployed around 50 digital humans in a variety of organizations around the world. With the help of this technology, consumers can interact with chatbots and other computer-based interfaces in a way that is more relatable and human-like.

For the past seven years, Alan Dennis, the John T. Chambers Chair of Internet Systems at the Indiana University Kelley School of Business, has been closely observing this emerging trend, together with colleagues at the University of Sydney and Iowa State University. It is in this article published in the Harvard Business Review – AI with a Human Face: The case for – and against – digital employees – that they take a closer look at how digital humans interact with us and when they are most appropriate to be used.

Providing personal services at scale is one of the most significant advantages that digital humans have over their anthropomorphic counterparts. It will enable them to offer tailored support and assistance one-on-one to customers and employees, allowing them to interact one-to-one. It is highly advisable for businesses that are looking to improve their customer relationships to take advantage of this level of personalisation, as it can improve customer satisfaction and loyalty.

Digital humans, however, do come with a number of potential drawbacks as well. The use of a virtual assistant may not be comfortable for some people, and it could lead to a negative experience for these customers as they interact with the assistant. In addition, there is concern about data privacy and security as well as the risk of bias being built into the algorithms that power the digital humans, as well as other privacy issues.

READ MORE: The FTC warns: Be careful with AI claims

In their article, Dennis and his colleagues wrote that firms that embrace the new technology will reduce costs, increase revenues, and have a sustainable first-mover advantage that slower adopters might find hard to overcome as customers become more attached to their digital counterparts over time. There is no doubt that within a decade, we expect the majority of managers at most companies will have an assistant or an employee who is a digital human.

Mike Seymour, a co-director of the Motus Lab at the University of Sydney, Dan Lovallo and Kai Riemer, both professors at the University of Sydney, and Lingyao “Ivy” Yuan, an assistant professor at Iowa State University, are the other authors of the study.

While we are aware that the digital human doesn’t exist in the real world, our minds are “attuned to facial signals and respond emotionally to them,” they wrote in their article. The thing is, even though we understand that what we see on the screen is an artificial construct, we still connect instinctively with it, and we do not need to turn into computer experts to interpret the facial signals and make the exchange work as it should.”

Humans are already doing some things in a better way with the help of digital technology

Having worked on projects for companies that create digital humans – including Pinscreen, Epic Games and Soul Machines – and consulting for them on those projects, Dennis and his colleagues have studied the design and appearance of digital humans, when they should be deployed, and in what contexts they should be placed in.

Virtual humans can be divided into four main groups: virtual agents, digital assistants, influencers, and companions.

  • In the future, virtual agents will be able to accomplish specific, one-time tasks like chatbots and will also be able to act as digital instructors in videos and presentations just like chatbots. In his work, Dennis and his co-authors cite how digital humans are used in international airports to provide travelers with flight information.
  • As a virtual assistant, you can assist a user with specific tasks and often develop personal relationships with them, which is why they are often used as rehabilitation therapists, personal assistants, and coaches in rehabilitation programs.
  • There have already been successful implementations of virtual influencers in the fashion industry that provide their human followers with experiences, but they are not “personalised.” In other words, just as people may follow the Kardashians and see pictures of their lavish experiences, “any relationship a person might feel with them [virtual influencers] comes from that person’s projection and not from any individualisation.” 
  • In addition to elder care, virtual companions have great promise in early childhood education, as well as elder care. It is known that older people who use virtual companions have an increased chance of staying in their homes for a longer period of time, which is better for their physical and mental well-being. Compared to assisted living facilities or nursing homes, they are also much more affordable,” they wrote. There are similar opportunities in the education sector as well. Watching other children engages a child more than watching their own. In other words, a digital human child-aged teacher could at times be more effective than a human adult teacher in terms of teaching children.

“Digital humans are much more likely to be able to convey complex instructions, or describe features of a product, than human beings,” the researchers write in their report. “This is why YouTube instruction videos do so much better than pages and pages of text,” they write.

In order to get a better sense of how the pieces go together and whether the outfit reflects who they are, a person who is searching online for clothes might consider seeing the outfit on someone who looks just like them. When this is the case, the customer will be engaged more, the sale will be completed more smoothly, and the likelihood of product returns will be lower.