“What is the real advantage of being human?” Alex Swarstel, director of perspectives at Jobs for the Future (JFF), asked herself this question after Intel, one of the major US technology companies, engaged her services. The company wanted her to help determine which areas of its operations could be enhanced by technology and which should be left to machines.

“Generative artificial intelligence (AI) can simulate human interaction, sometimes even better than a person,” says Swarstel. That is why it is more urgent than ever to answer the question of what the real human advantage is. In her view, there is no single answer. It might be “knowledge and experience in a particular field, which enables you to know the difference between being 85% certain or 100% certain on an assembly line.” Or it could be the ability to “bring that personal, human touch to certain services.” In both cases, she says, “AI can get close, but it will never quite reach that point.”

This is one of the major questions of the day: every industry, bank, construction company, institution, hospital, university, NGO, retail chain and even transport operator will have to decide anew which tasks to allocate to people and which to machines – and must do so with unprecedented urgency if they do not wish to concede an advantage to their competitors.

“This is a moment none of us will ever forget. We are going to be the last CEOs to manage a workforce made up solely of humans. From now on, we will also have digital workers,” said Marc Benioff, chairman and CEO of Salesforce, one of the leading providers of customer management technology services.

This is not the only solemn pronouncement to have been made recently; 2025 has already seen more than a few. NVIDIA founder and CEO Jensen Huang believes we are on the brink of the “ChatGPT moment for general robotics.” He sees it taking three main forms: AI agents, which he calls “information workers,” able to move freely, learn, analyse and make their own decisions; autonomous vehicles; and humanoid robots.

Of these, only the first are already a reality for many large-scale organisations, albeit still in testing. Autonomous cars are now driving in US and Chinese cities, though they remain subject to strict regulatory restrictions (Europe continues to take a tougher line) and must operate in designated areas where they have undergone extensive training. As for humanoids, robots in human form, for the time being they are more effective with wheels than with legs.

In Huang’s vision of the future, AI agents will constitute a form of digital labour, working alongside people and taking decisions on behalf of companies. That is why he suggests we design an onboarding process for them akin to that used for human employees. “The IT department of every company will be the human resources department for AI agents in the future,” he says.

Dario Amodei, co-founder and CEO of Anthropic, one of the leading firms in generative AI, declares: “We are moving towards artificial systems that are better than almost all humans at almost all tasks.” He describes his vision of a “country of geniuses”: a data centre filled with millions of AI agents. “What limits will there be to solving all the world’s problems at once? We could make 100 years of progress in areas like biology in just five or ten years.”
The company is currently developing a “virtual collaborator – not necessarily more intelligent than a Nobel Prize winner, but capable of carrying out relatively high-level tasks.” It will interact with colleagues and remain operational for hours at a time.

Brian González, executive-in-residence at the JFF Centre for Artificial Intelligence and the Future of Work, is well acquainted with Swarstel’s research. In fact, he was one of the people instrumental in Intel hiring her while he was director of training. His reflections are more measured than the bold declarations of Silicon Valley leaders. He argues that the main challenge for AI is to augment human capability.

“When it does that, it lifts us up, giving us time to focus on things more innate to our abilities – what we call enduring skills such as critical thinking and collaboration,” says González. “The human element is the key to successfully extending AI. Someone in Silicon Valley is not going to solve it; we have to think about what we want to achieve, what outcomes we need, and how the suite of technologies can help us deliver in a way that produces a higher-quality result.”

“Domain knowledge and experience are vital for assessing what AI tells us and knowing whether it is correct or not,” insists Swarstel. “The new business imperative is to identify what we need to train for. Leaders must understand the dynamics of their business, what will make them competitive in the long term, and how. It will take human creativity combined with the generativity of tools like AI to drive that kind of growth.” For González, “the new digital divide is not whether we own a device or not, but what we do with it.”

As technology is rolled out, a new sense of trust and transparency must also be fostered. Agent-based AI is a game changer: “If I send it off to do certain things and it does not return with what I need, it may have taken some actions on my behalf. This will open up a new frontier of transparency and trust, because AI will operate independently of us.”

Between the launch of ChatGPT in December 2022 and December 2024, the number of job postings requiring AI skills has doubled. “It is difficult to predict with certainty which roles will be automated,” says González. At JFF, attention is focused on two types of work: jobs that involve creating new things, because the ability to understand a human need and match it with a business proposition is extraordinarily powerful; and jobs that depend on human-to-human connection in some form. “AI will only ever be as good as the human intelligence that drives it.”

Sources: 

• Ces.tech
• Weforum.org