Nobody Told Us How to Fly It: The Career Inside Healthcare’s AI Training Gap
Written by Rod on August 10, 2026
Series: From Bedside to Build · Episode 6 · Pillar 3: Expertise and Opportunity
By Rod Gamble | Week 33, 2026 | Pillar 3: Expertise and Opportunity
The new AI tools appeared at work before anyone was taught to use them — and everyone is quietly hoping a colleague figures it out first.
That sentence could have been written in almost any staff room, in almost any health system, this year. I want to talk about the gap it describes — the space between the technology arriving and anyone being shown what to do with it — because most clinicians see that gap as a problem. I see it as the clearest doorway into digital health that this decade has offered.
The Tools Landed Before the Training
Consider the scale of what is happening right now. In June, NHS England announced that 505,000 clinicians and staff will have Microsoft 365 Copilot on their desktops by October, following a pilot of 30,000 staff which NHS England reports saved an average of 43 minutes per person per day. A month later came the follow-up: an accelerated national rollout including ambient voice technology, with an emergency department pilot at St George’s reporting an average of 47 minutes saved per clinician per shift. Those are pilot claims from press releases, so hold the exact figures lightly — but the direction of travel is unmistakable. This is the largest AI deployment most clinicians will ever personally feel, and similar waves are moving through US and Irish systems.
But listen to the front line. On the Patient Safety Learning hub, a patient safety manager described what the rollout feels like from the inside: “Copilot has arrived in the NHS — but no one told us how to fly it.” No training. Real anxiety about where patient data goes. Wi-Fi that can’t carry the load. Digital literacy that varies wildly from one desk to the next. The tools landed before the training did — and the question forming in every break room is the same: who exactly is supposed to close that gap?
The Night the Macros Did the Work
I have been on both sides of that question, and the first time changed my life. In the mid-1990s I was in nursing school, working nights at a local cell-phone company because if I didn’t work, I didn’t eat. The job on the slow shift was medieval: retype thousands of customer records into a new system by hand, one field at a time, off fan-fold dot-matrix printouts. People had been grinding through it for months.
I wanted to study. So I asked the day team for the raw data as a file, taught the macro recorder in Windows 3.11 to repeat the keystrokes, split the work across the dozen customer-service computers, and let them run overnight. By morning, the machines had finished about three-quarters of a months-long job in a single evening. It earned me a raise — enough for food, gas, and the rent on my trailer lot, with a little left over for the first time.
It was plain automation — not AI, not machine learning. But it taught me the lesson I have leaned on ever since: the win never comes from the technology. It comes from understanding the work well enough to rebuild it around the tool. And here is the detail that matters for you — nobody trained me to do that either. There was no course. The gap was the training.
The Gap Isn’t the Obstacle. It’s the Opening.
Here is what most clinicians expect to hear next: wait for the training team, escalate to IT, the organisation will sort it out eventually. But the honest answer is different. The gap between tool and training is not something to route around — for the clinician willing to step into it, it is the opening itself.
Look at what happened to the last profession AI was confidently supposed to erase. A decade ago, radiologists were told their field was finished. Ten years on, demand has grown, the number of active radiologists in the US is up around 10%, and salaries have reached record levels. Clinical judgement did not get replaced when the machines arrived — it got repriced, upward. A radiologist writing on KevinMD put the frame precisely: FDA-cleared AI in medicine is assistive, not autonomous. The physician stays in the loop. Therefore someone has to be the loop.
Your colleagues already sense this. A recent analysis of clinician discussions of ambient AI scribes on Reddit found sentiment running roughly 61% neutral, 22% positive, and 17% negative — not hype, not panic, but practical problem-solving talk: the editing burden, the accents the tools mishear, the clumsy copy-paste into the EHR. The profession has quietly moved past “will this replace me?” and into “how do I make this work?” Read that second question again. It is a job description.
What the Market Is Saying
The demand side is measurable. The US Bureau of Labor Statistics projects health information technologist roles to grow around 15% between 2024 and 2034, and industry surveys suggest roughly 70% of healthcare employers now treat digital competency as a critical hiring factor even for clinical roles. Most of these paths — workflow design, clinical informatics, training, AI governance — require no coding at all. Meanwhile, physician burnout has fallen for a fourth straight year to 41.9%, yet one in four physicians is still considering leaving clinical medicine. Put those together and the picture is striking: the systems rolling out AI and digital transformation at national scale need exactly the people who are thinking about walking away — just on different terms. Clinical innovation needs clinicians.
Steps You Can Take Now
Master the one tool in front of you. Not every AI — the one your employer actually deployed. Give it two focused weeks: learn its settings, its limits, and its failure modes. Expertise in healthcare technology starts embarrassingly small, and that is fine.
Map it to the real workflow. Document where it genuinely saves time and where it fails — the editing burden, the integration gaps, the steps it doesn’t understand. Then write the one-page guidance nobody gave your team. You will have produced your first piece of digital transformation work without leaving your job.
Teach three colleagues. Fifteen minutes each. Then raise your hand for the super-user role, the training group, or the AI governance committee the moment it forms. Nearly every digital health career I know — including mine — began with a hand going up.
Keep the evidence. Your notes, your guide, the minutes saved. That file is a portfolio proving clinical innovation experience that most CVs merely claim — and it is what turns a workaround into a career.
The tools will keep arriving before the training does; that is simply how healthcare technology moves now. But somebody has to write the instructions the box never included. The clinician who does isn’t just using the system anymore. They’ve started building it.
When you’re ready to talk, rodgamble.com is where to find me.
References
1. NHS England. “500,000 NHS staff to get new artificial intelligence tools to help free up more time for patients” (June 2026). https://www.england.nhs.uk/2026/06/500000-nhs-staff-to-get-new-artificial-intelligence-tools-to-help-free-up-more-time-for-patients/
2. NHS England. “NHS accelerates artificial intelligence rollout to cut waiting times and improve care for millions” (July 2026). https://www.england.nhs.uk/2026/07/nhs-accelerates-artificial-intelligence-rollout-to-cut-waiting-times-and-improve-care-for-millions/
3. Patient Safety Learning hub. “Copilot has arrived in the NHS — but no one told us how to fly it!” https://www.pslhub.org/learn/digital-health-and-care-service-provision/288_artificial-intelligence/380_large-language-models-llms-and-generative-ai/copilot-has-arrived-in-the-nhs-%E2%80%94-but-no-one-told-us-how-to-fly-it-r13820/
4. medRxiv preprint. Topic modelling and sentiment analysis of clinician Reddit discourse on ambient AI scribes (2026). https://www.medrxiv.org/content/10.64898/2026.04.26.26351798v1
5. Fortune. “A decade after Geoffrey Hinton said radiologists were obsolete” (May 2026). https://fortune.com/2026/05/04/godfather-of-ai-geoffrey-hinton-radiologists-future-of-work-tech-ai-job-anxiety/
6. KevinMD. “Is AI replacing radiologists? A radiologist disagrees” (June 2026). https://kevinmd.com/2026/06/is-ai-replacing-radiologists-a-radiologist-disagrees.html
7. Barton Associates. “Physician burnout remains high in 2026: latest rates, top causes, staffing shortages and schedule control” (May 2026). https://www.bartonassociates.com/blog/physician-burnout-remains-high-in-2026-see-latest-rates-top-causes-and-how-staffing-shortages-and-schedule-control-impact-clinicians/
8. Prosum. “Healthcare IT jobs in 2026: hiring trends, essential skills, and top roles” (Dec 2025); INNOVA People. “How digital health is changing healthcare careers” (May 2026). https://www.prosum.com/2025/12/03/healthcare-it-jobs-in-2026-hiring-trends-essential-skills-and-top-roles/ · http://innovapeople.com/2026/05/25/how-digital-health-is-changing-healthcare-careers/
