Digital humans poised to transform workplace within decade, researchers predict

Within the next decade, most companies will likely have hired someone who doesn't need a paycheck, doesn't call in sick, and won't ask for a promotion. Researchers from the University of Sydney Business School, working with colleagues at Indiana University and Iowa State, are predicting that digital humans—AI-powered avatars with realistic faces and voices—will become standard workplace fixtures across industries.

These aren't the chatbots people already know. Siri and ChatGPT operate as disembodied voices or text. Digital humans take that conversational technology and anchor it to a face and voice, creating something that feels fundamentally different to the human brain. Mike Seymour, a senior lecturer at Sydney's business school, explains the psychological mechanism at work: we're wired from birth to respond to faces, particularly to the faces of people we trust. Digital humans exploit this deeply embedded instinct. They can't actually feel emotion, but they're becoming remarkably skilled at performing it convincingly enough that our brains find the interaction natural and pleasing.

The technology is already moving beyond the laboratory. At EY, the professional services firm, executives have created deepfake versions of themselves that deliver messages to international staff in languages the real executives don't speak. Lil Miquela, a digital human with nearly three million Instagram followers, operates as a commercial influencer, selling products and experiences with the same apparent authenticity as a human creator. The economics are obvious: once deployed, digital humans never tire, never complain, never demand raises, and always comply with company policy.

But not every job is equally vulnerable. Seymour and his colleagues argue that roles involving genuine collaboration—where humans work together to solve problems that require judgment, creativity, and real emotional intelligence—will likely survive. A digital assistant might join a Zoom call to check schedules and answer routine questions, freeing humans from administrative busywork. But it cannot decide what information matters or navigate the ambiguity of a genuinely novel problem. Jobs that require people to work together with other people will be augmented by digital agents, not replaced by them. The roles most at risk are those built around emotional connection and personal assistance: customer service, coaching, mentoring, certain aspects of healthcare and education.

Medical students could soon train with hyperrealistic virtual instructors. Shoppers could see how clothes fit on a virtual body resembling their own while receiving encouragement from a digital sales associate. Elderly people struggling with new technology could learn from a familiar digital face—a form of companionship that might help them stay independent longer, living at home with family rather than in institutional care.

Seymour is careful to frame this not as workforce destruction but as workforce transformation. As some jobs disappear, others will emerge—though he acknowledges the transition will be disorienting. The crucial message, he argues, is directed at students and educators: understanding this technology and its limitations will become essential to remaining competitive. Workers who grasp how digital humans work and where they fail will have an advantage over those who don't. The alternative is being outpaced by someone who does.

The underlying tension is real. Digital humans are statistically plausible—they generate responses that sound natural because they're built on patterns extracted from vast amounts of internet data. But plausibility isn't the same as truth or understanding. As these systems become more sophisticated and more prevalent, the question isn't whether they'll transform the workplace. The question is whether we'll know how to work alongside them without mistaking performance for genuine comprehension.