Which Healthcare Tasks Should Stay Human?

 Elon Musk has presented his view of what the future holds for us — or at least what he and his company, Tesla, envision. Within 10 years, there will be a fleet of more than 1 billion humanoid robots, whose productivity will be five times that of a human. As Musk explained, the onset of labor automation will be slow, but once robots are capable of manufacturing other robots, growth will be explosive.

This prediction has sparked widespread public reaction. In fact, shortly before this announcement, Bill Gates published a comprehensive reflection on his websites about the future of AI under the title, “The Turbulent Era of AI Is Already Here. The decisions we make now are crucial.” In this extensive piece, he offered numerous reflections on the well-known benefits AI can bring to society, but also on the many risks if it is not implemented safely — particularly regarding cyberattacks — and even on how it might affect human cognition itself.

Reserved for Humans

For this reason, he made various proposals, among which the idea of creating a “Reserved for Humans” initiative stood out — to safeguard certain jobs or tasks that must continue to be performed by people. In fact, he introduced this concept with a reflection on the healthcare system and his own experience with his father’s treatment because his father suffered from Alzheimer’s disease.

“My family and I will always be grateful to that incredible group of professionals. The care they provided to my father was irreplaceably human. No robot could or should have done that. I think of that team whenever the question arises about which jobs will disappear and which will remain. I believe that, as AI and robots improve, we will reserve certain tasks exclusively for people. I’ve started calling this area ‘Reserved for Humans,’ and it’s an example of the kind of ideas we’ll have to consider.”

A ‘Reserved for Humans’ Category

Reflecting on how this idea might apply to the healthcare sector, El Médico Interactivo, part of the Medscape’s Professional Network, spoke with Julio Mayol, scientific director of Health Research Institute of Hospital Clínico San Carlos in Madrid, Spain, and co-founder of EpistemAI. From his perspective, what needs to be protected — more than professions — are certain functions that should enhance the value of human capital.

“For me, without a doubt, end-of-life care, delivering bad news, genuine informed consent — not just filling out a form — mental health, palliative care, and the ongoing relationship in primary care. All those situations where someone must make a decision amid uncertainty and take responsibility for it. A machine can make suggestions. It cannot be held accountable.”

As the expert reflected, “the human factor is concentrated where uncertainty, conflicting values, and vulnerability converge. It depends not so much on the position as on the decision made within it.”

Conversely, we must also identify which tasks are truly worth handing over to AI. The ones we should hand over with relief are the administrative burden, documentation, coding, preliminary test review, and routine monitoring. Delegating them is not a threat — it’s a relief.

As outlined in a recent analysis published in June 2026 in npj Digital Medicine regarding collaboration between humans and AI in healthcare, the available evidence currently focuses on diagnostic interpretation, with fewer studies on screening and triage, therapeutic decision-making, and administrative workflows. However, although issues such as accountability and patient safety frequently arise in the debate on this collaboration, they remain aspects that have been poorly evaluated empirically.

Embracing Change

Nevertheless, the reality is that as both AI and humanoid robots continue to advance, their potential and capacity to take on tasks we cannot yet even imagine will continue to grow.

On this issue, El Médico Interactivo also spoke with Ismael Said Criado, an internist and chair of the Telemedicine Working Group, innovative technologies, and digital health at the European Federation of Internal Medicine. “Our profession is based on decision-making grounded in clinical information, and that is precisely what AI does much better. We have spent more than a century perfecting that decision-making process, developing medical science to provide solutions to diseases. And now a technology is emerging that harnesses all that knowledge to do so much more efficiently. So I see it as inevitable that this technology will take on cognitive tasks that we humans still perform today.”

Even so, the expert clarified that professions more focused on care will be less likely to disappear with the development of AI, such as nursing or even family medicine, although automation will also have an impact in this area.

Considering this reflection, Criado noted that we must accept that the change in the workforce brought about by AI is unstoppable. “We’ll likely change the way we work significantly and will only provide support to patients to build trust and reassurance.” However, he added, “I’d like to think that legislation will be enacted in the near future to ensure that humans remain the guarantors of the clinical process, but we’ll probably only be the ones who initiate a process and validate it upon completion.”

In any case, Criado’s conclusion is that “we must continue to evolve to improve healthcare so that patients derive the maximum benefit from this entire transformation. We should not limit ourselves to merely defending our professions, thereby missing the opportunity for significant improvement for the patient.”

https://www.medscape.com/viewarticle/which-healthcare-tasks-should-stay-human-2026a1000ydg