Reflections
Lifelong learning is a professional obligation medical professionals sign up for. Reflection has been a powerful personal vehicle in this regard throughout my career.
I have always reflected on my work; but here I do some of it in the open. Each entry follows a simple structure, what happened, so what, and now what, and records the time it took, because these reflections also form part of my appraisal and revalidation with the GMC.
Reflection is about my own practice and learning, so nothing on this page identifies a patient or a clinical colleague, and no confidential detail appears. If a reflection cannot be shared safely in public, it stays in my private record instead. Peers who appear do so with their explicit consent; where I remember someone I worked with in public life, I do so as a tribute.
A pattern, not a bet
- Knowledge, skills & development
- Colleagues, culture & safety
On a decade of adopting the next medium early: WatMed Media in 2015, ChatGPT within a month of launch for my MD analysis, and a site built for machine readers in 2026.
On air in ten minutes
- Patients, partnership & communication
- Trust & professionalism
- Knowledge, skills & development
On going live with Dermot Murnaghan as the Prime Minister entered intensive care, with barely ten minutes to prepare, and why doing the science justice is the antithesis of shock-and-awe health influencing.
Undoctored Truths: turning a moment into a movement
- Trust & professionalism
- Patients, partnership & communication
On co-founding the Undoctored Truths Alliance, and why medical misinformation needs people to organise rather than resign.
B2LLM: when the reader is a language model
- Knowledge, skills & development
- Patients, partnership & communication
- Trust & professionalism
On coining B2LLM, and what the shift from search to language models means for anyone responsible for accurate medical information.
The model I recommended was not the model I used
- Knowledge, skills & development
- Trust & professionalism
On benchmarking Claude against GPT for doctoral qualitative analysis, and why being honest about the instrument is part of the finding.
Why I built GME:X with a deliberately early-career team
- Knowledge, skills & development
- Colleagues, culture & safety
- Patients, partnership & communication
On building GME:X and D:CAM at Bayer, and the deliberate choice to do it with an early-career team.
Education fails when it cannot adapt to the learner
- Knowledge, skills & development
- Colleagues, culture & safety
On building an adaptive teaching method years before a dyslexia diagnosis explained, from the inside, why it mattered.
Leading a fireside at short notice
- Knowledge, skills & development
- Trust & professionalism
On being asked to lead an AI and trust fireside at 48 hours notice, and choosing a format that kept the session credible and humans at the core.