Artificial intelligence (AI) is disrupting healthcare from the ground up. Patients now arrive at (virtual) appointments equipped with AI-generated diagnostics, laboratory result interpretations, and curated treatment recommendations. They often cite journal articles that their doctors haven’t even read. Large language models (LLMs) digest more medical knowledge than any human could hope to retain, and are now widely used for symptom triage, treatment planning, and even procedural implementation.
Patients (and their relatives), especially the more engaged or discerning, do their own research. AI analyses are often highly sophisticated, and may include:
- referenced recent clinical studies,
- detailed diagnostic probabilities,
- prognostic estimates,
- varying specialist opinion and
- various potential management strategies.
The waiting time to see a specialist might be 20 days. An AI interpretation of laboratory tests? Less than three minutes.
This all heavily impacts on the role of doctors, with an increasing emphasis for them to now serve as 'guides' and providers of medically meaningful advice. Helping patients to:
- put their data into context,
- interpret uncertainty, and
- make deeply personal choices that are rooted in their individual situations.
AI is handling low-risk decisions, automating pharmacy logistics, and even leading surgeries under human oversight. Meanwhile, patients are driving change by demanding faster, more personalised, and proactive care. The economic pressure for medical providers to move to full medical automation is also mounting, but the human need for empathy, trust, and interpretation remains irreplaceable.
Compassionate guides who bring meaning to the metrics
As part of building the doctor-patient relationship and building deeper patient partnerships, doctors will increasingly need to stay abreast of AI capabilities (and pitfalls), and be actively involved in helping people navigate divergent expert opinions and the often complex care pathways. It's about making medicine and the treatment / no-treatment options 'meaningful'.
This shift also reveals that clinical consensus is not a constant. When a patient presents their full electronic record (EPR) to multiple specialists, they may receive radically different treatment recommendations. Intra- and inter-specialist variation, driven by training, geography, culture, experience, and even philosophical leanings, is a long-standing reality of medicine. AI may now expose these differences more starkly than ever before. This will inevitably also heavily impact on the role of the MDT (Multi-Disciplinary Team).
Making medicine meaningful
The doctor’s role must therefore evolve, not as an arbiter of fixed truths, but as a translator and guide through complexity. Patients will increasingly need help comparing divergent clinical opinions, understanding the implications of each option (including the do-nothing option), and making informed decisions that reflect their own values and goals. This calls for a deeper, more independent form of engagement, where the clinician is not just a 'proceduralist', but rather a partner in meaning-making.
There are always treatment options
AI cannot understand what health means in the context of someone’s life. It cannot comfort, motivate, or weigh deeply personal trade-offs. Is the age of the discerning, relational, and deeply human doctor is just beginning? In this new era, the doctor’s greatest power lies not in the answers they provide, but in the questions they help patients ask. Why? Because there are always options.
OK I have the report, so what now!?
Access to information doesn’t guarantee understanding, and that’s where the new challenge lies. A patient may now face a complex AI-generated report, multiple expert opinions that diverge sharply, and no clear guidance on how to synthesise these perspectives. For instance, one spinal surgery patient might receive three different management plans from three equally highly credentialed surgeons, each shaped by their training, region, and preferred techniques. Who is right? What are the risks of doing nothing?
This is where doctors as guides and meaning-makers become indispensable. Their role is not simply to add another opinion to the pile, but to help patients evaluate the range of perspectives in front of them. This includes explaining the reasons for clinical variation, exploring the values and preferences that matter most to the patient, and considering the full spectrum of options, including the often-overlooked choice to defer or decline intervention.
In a world where data is abundant, but discernment is rare, the doctor’s true value is emerging in a new way, not as gatekeeper of information, but as a companion in complexity. The doctor becomes an interpreter of multiple signals including the clinical data, AI recommendations, personal history, and patient goals, and a co-navigator of decisions that are rarely black and white.
Medical decisions are rarely black and white
The change means prioritising conversations that matter, exploring the nuances of “what’s right for you” rather than merely “what’s standard practice for the population,” and having the time to help patients understand their choices in the context of their lives.
Can't get to see your GP?
Today, we are witnessing the first credible steps toward fully autonomous care. Systems are emerging that don’t just support physicians, they can operate without them. These tools can triage symptoms, interpret imaging, propose treatment plans, and monitor chronic conditions with minimal human oversight. Digital-first companies such as Total Doctor already integrate with AI agents to manage prescription refills, follow-up reminders, and even initial consultations. In surgical settings, robotic systems are moving from assistive to independent procedures, with FDA-cleared platforms performing increasingly complex interventions under remote supervision.
This transition follows a predictable path seen in other industries: automation starts with bounded, low-risk tasks and gradually scales up. In healthcare, this path might look like:
Phase 0: Digital triage and remote prescribing
Phase 1: Automated management and logistics (e.g., booking, billing, pharmacy operations, sample collection)
Phase 2: Semi-autonomous procedures with human oversight
Phase 3: Closed-loop autonomous care, where AI not only plans and delivers treatment but adjusts it in real time based on patient response
Interpreting the grey zones of medicine
The broad suggestion is that clinicians can refocus on what truly matters i.e. developing relationships, guiding complex decisions, and interpreting the grey zones of medicine.
What remains uniquely human in medicine?
The answer lies in interpretation. In a world saturated with information, sometimes conflicting, often overwhelming, patients need someone to help them make sense of it all. Not someone who adds more noise, but someone who can clarify signal.
Assisting patient choice
Consider the patient who receives three different management plans from three specialists, or who has just run their lab results through an AI tool and received a series of probability-weighted prognoses. These are no longer rare instances, they are fast becoming the norm. What that patient often lacks isn’t more data, but the information underpinning ‘informed choice’ and the confidence to choose.
Doctors must now step into the role of clinical translator, which means helping patients to:
- weigh the evidence,
- understand trade-offs, and
- align decisions with personal values and life goals.
That might mean explaining why one expert prefers conservative treatment while another recommends surgery. It may involve helping a patient understand the potential implications of doing nothing, or how comorbidities, lifestyle, or psychological readiness affect outcomes. In short, it requires presence, nuance, and time.
Doctors are cultivating new skills: not just clinical knowledge, but ethical discernment, narrative listening, and psychological insight. The role will evolve from problem-solver to pattern-seeker; from information provider to a partner in the provision of medical ‘meaningful’ advise.
