How many times has a nurse or doctor delivered the hardest line of their day without ever having rehearsed it? Words are the most used—and least trained—clinical tool in healthcare. We learn to insert an IV into a mannequin; we learn to deliver bad news, almost always, in real time.
What if the question isn’t how much we know about communication, but how much we’ve actually practised it?
Is the current system outdated?
It isn’t outdated—it’s incomplete. Lectures and clinical placements remain essential, but they don’t teach the craft of conversation. Instead, they teach something no one intended: the "hidden curriculum" (Hafferty and Franks, 1994). Its effects are measurable: empathy declines during medical school and residency (Neumann et al., 2011).
The obsolescence isn’t in the content but in the method: we teach conversation as we teach history—by explaining it—and we learn at the expense of real people.
A generation learning differently
No educational method is neutral in its era. According to sociologist Hartmut Rosa (2013), we live in an age of acceleration that shrinks the present, and 46 % of US teenagers report being online "almost constantly" (Pew Research Center, 2024). The label "digital native" has little empirical basis (Kirschner and van Merriënboer, 2013), but new habits carry weight: Sherry Turkle (2015) argues that the ever-present phone impoverishes conversation and empathy.
Hence the paradox: those who have lived most through screens need more practice in human conversation, and technology may be the very door it recognises. Adapting isn’t surrendering—it’s teaching in the learner’s language so they can finally learn the language of the other.
What practitioners gain from practice
Simulation with deliberate practice yields better results than traditional clinical training (McGaghie et al., 2011), and motivational interviewing skills fade after a workshop unless reinforced with feedback (Schwalbe, Oh and Zweben, 2014). An AI-simulated patient provides what’s missing: error without victims, on-demand repetition, and varied patients—including the difficult ones.
To be clear: a scoping review of 28 studies (Zeng et al., 2025) describes a promising but fledgling field, and only five measured learning objectively. AI doesn’t replace the tutor.
What kind of professionals do we want to train?
Aristotle put it plainly: what we must do after learning it, we learn by doing (Nicomachean Ethics, II, 1). Empathy isn’t just a trait you’re born with; it’s a habit you cultivate or lose. The deeper question isn’t whether a machine can teach humanity, but what kind of professionals we want to train—and at whose expense.
How many more difficult conversations will we let debut live?
Entrena-T is an AI patient simulator for practising motivational interviewing in a safe environment. Rehearse today the conversation you’ll face tomorrow.
Request your Entrena-T trial
