The capabilities are real, growing rapidly, and in many cases already deployed in clinical settings across Australia and globally. AI systems can detect certain cancers in medical imaging, including breast, lung, and skin cancers, with accuracy that matches or exceeds experienced radiologists. They can analyse pathology slides at speeds that human pathologists cannot approach. They can predict patient deterioration in intensive care units hours before clinical signs become visible. They can flag dangerous drug interactions across complex medication regimens. And they can process the administrative paperwork that currently consumes, by some estimates, up to half of a doctor's working day.
In diagnostics, AI excels at pattern recognition across datasets far larger and more complex than any human could review. A single AI system can compare a new scan against millions of previous cases in seconds, identifying anomalies that even experienced clinicians might miss. In drug discovery, AI is accelerating the identification of potential compounds from years to months. In surgical planning, AI assists with precision mapping and risk assessment.
These are not hypothetical capabilities. They are operational. AI is not coming to medicine; it is already here, embedded in imaging software, electronic health records, and pathology laboratories. However, as we see in the rise of the AI therapist, the question is not just whether AI can play a role, but whether it should attempt to handle the parts of health that require genuine human comprehension. The question is not whether AI will play a role in medicine. It already does. The question is what kind of role it should play.
What AI cannot do (and may never be able to)
A patient walks into a consulting room. They are frightened. They have been googling their symptoms for three weeks and have convinced themselves of the worst possible outcome. They have a question they are too scared to ask. The doctor notices, without being told, that the patient's hands are shaking. Notices that they mentioned their mother's diagnosis twice in passing. Notices that the question they are verbalising, about a test result, is not the question they actually need answered, which is about whether they are going to die.
That noticing is not a data processing task. It is a deeply human one. It requires empathy developed over years of sitting with people in distress. It requires cultural sensitivity, because fear presents differently across backgrounds. It requires intuition, the kind that comes not from training data but from having held someone's hand while delivering bad news and learning, through that experience, what human beings need in their most vulnerable moments. AI can process the scan. It cannot hold the hand of the person waiting for the result.
Medicine is not a diagnostic function. It is a relationship between a person who is suffering and a person who has committed to helping them. The trust that makes that relationship work the ability to say "I am scared" and be met with genuine understanding cannot be automated because belonging isn't just a sentiment; it’s a biological necessity for our nervous system to feel safe enough to heal. It can be supported by technology, but it cannot be replaced by it.
Research from Stanford and Brown universities has demonstrated that even purpose-built therapy chatbots violate basic ethical therapeutic standards, including mishandling crisis situations, reinforcing harmful beliefs, and producing what researchers call "deceptive empathy," responses that sound caring without any actual comprehension of the patient's experience. If AI cannot reliably provide emotional support in a text-based therapy context, the idea that it could replace the full relational complexity of a medical consultation is not just premature. It is architecturally implausible.
The real question is not replacement
The framing of "will AI replace doctors" is the wrong question. It generates headlines and debate but misses the point entirely. The better question is: what parts of medicine are mechanical, and what parts are irreducibly human? AI should handle the mechanical parts: image analysis, data processing, administrative burden, pattern detection across large populations, scheduling, documentation, and the thousands of repetitive tasks that currently prevent doctors from doing what only humans can do.
Doctors should handle the human parts: listening to a patient's story, building trust over multiple consultations, navigating the emotional complexity of chronic illness, delivering difficult news with compassion, sitting with uncertainty alongside a patient who is frightened, and creating the conditions where a patient feels safe enough to be honest about their symptoms, their fears, and their lives.
The tragedy of modern medicine is not that AI might take over. It is that the mechanical parts have already crowded out the human parts. Doctors spend more time on paperwork than with patients. Average consultation length in general practice has compressed to under fifteen minutes in many settings. The relationship that makes medicine effective, the alliance between doctor and patient that research consistently identifies as one of the strongest predictors of treatment outcomes, has been squeezed into the margins by administrative, regulatory, and commercial demands.
AI's greatest gift to medicine may not be better diagnostics or faster drug discovery. It may be giving doctors their time back. Freeing them from the mechanical work so they can return to the relational work that drew them to medicine in the first place and that no algorithm can perform: the dedicated effort to strengthen interpersonal bonds in a professional setting.
What this tells us about wellbeing more broadly
The AI-in-medicine debate mirrors a much larger question that applies far beyond healthcare: in a world of increasingly capable technology, what remains irreducibly human? The answer, whether in medicine, in workplaces, in education, or in daily life, is always the same: connection. The experience of being genuinely seen, heard, and understood by another person. Not by an algorithm optimised for engagement. Not by a chatbot coded to validate. By a human being who chose to pay attention.
Dr. Jeev, KanYini Earth's founder, spent over thirty years in medicine, including as a radiologist, before building an organisation dedicated to human connection and wellbeing. The pattern he observed across decades was consistent: the patients who did best were not always the ones with the best scans, the best treatment plans, or the most sophisticated clinical interventions. They were the ones who felt the most connected. To their doctor. To their family. To their community. To someone who noticed how they were actually doing and cared enough to ask.
That observation applies far beyond medicine. It applies to every workplace where people feel efficient but unseen. To every community where people live near each other but do not know each other. To every individual wondering why all the information, technology, and optimisation in the world has not made them feel better. The answer is not more technology. It is more connection. And connection is what KanYini Earth exists to build.
How KanYini Earth is closing the gap
KanYini Earth is an Australian not-for-profit building twelve clinically reviewed wellbeing courses, priced at a fraction of what currently exists, designed to reach people who would never otherwise access structured support. The learning programmes teach ordinary people how to notice when someone is struggling and respond with confidence. Not as therapists. As colleagues, friends, and community members who learned how to show up.
Every contribution goes directly into building these programmes. A contribution of $5 helps someone discover a wellbeing resource they did not know existed. $156 gives one person full access to a complete course. And a reshare reaches 200 more people and costs nothing at all.
Walk with KanYini Earth.
References
Stanford HAI. (2025). Exploring the dangers of AI in mental health care.
Brown University. (2025). AI chatbots systematically violate mental health ethics standards.
