Artificial intelligence (AI) has gone from a niche, often misunderstood concept to a highly relevant (albeit still frequently misunderstood) topic that is prevalent in discussions across medicine. While many clinicians accept that AI will eventually reshape healthcare, few can say with confidence how or when. Many of us are therefore left wondering: if I can’t beat it, should I join it?
This article is aimed at UK doctors and medical students who already have an interest in AI but are unsure where to start. It is intended to act as an entry point; outlining the key information, resources, advice and questions that you may want to consider if you aspire to make AI a meaningful aspect of your career in medicine.
If you are completely new to the topic of medical AI, you might prefer to start with these introductions first:
- MindTheBleep: An Introduction to Artificial Intelligence in Healthcare
- MindTheBleep: Introduction to AI in Healthcare – Types of AI and replacing Healthcare Professionals
Contents
Why should you get involved?
The most straightforward motivation is genuine interest. You might enjoy the problem-solving it offers or believe that it will transform healthcare and want to be part of that.
Those who anticipate a growing presence of AI in medicine may also be incentivised by future-proofing their careers through developing skills and experience that will counteract any potential devaluing of clinical skills which the rise of this new technology may cause.
Finally, the rapid growth of interest in AI and its impact on healthcare has outpaced the development of relevant skills within the medical workforce, creating opportunities for clinicians willing to bridge that gap. With the increasingly competitive landscape for speciality training within the UK, this vacuum of skills in such an important area may allow those interested in AI to leverage their interest into opportunities that also help them to progress through training.
Why should you not get involved?
As many working doctors will already know, the job is demanding. The basic workload can be exhausting, and that’s without considering the audits, conferences, publications and other unpaid hours necessary to progress through training. Taking on additional pursuits in AI on top of these responsibilities is something that you must consider carefully. Even those who are energised by the topic may struggle to maintain the necessary devotion and should be honest with themselves about whether such a pursuit is sustainable.
The future of AI, especially in medicine, is also uncertain. You may commit your limited time and effort to an area of research that fails to provide any benefit, or may choose to develop certain skills that become significantly less valuable due to new developments in AI.
Important Considerations
Defining your role
Saying that you want to “work in AI” is similar to saying you want to “work in medicine”; it’s much too broad to be meaningful. The first step is to clarify what interests you the most and what role you want to play. This will guide the skills you need to develop.
You may choose to consider what it is that actually interests you about AI:
- Problem-solving: If your main interest is finding new ways to solve problems, you might be interested in AI development, such as collaborating with data scientists and working in start-ups or NHS innovation hubs.
- Implementation: If implementation interests you, you may prefer roles within departments, hospitals, trusts and other organisations that provide you with opportunities to make decisions and lobby for the use of certain AI systems, such as a trust Chief Clinical Information Officer.
- Research: You may also find yourself drawn to AI research, which can range from considering novel ways that AI may be developed, to evaluating clinical outcomes from deployed systems, to using AI processes to provide insights into physiology and pathology.
You should also consider what role you’d actually play in your chosen domain. Your clinical background will be useful, but is this where your contribution would end? If you are drawn to the technical side, developing this further may set you apart from both the non-clinicians and the non-technical clinicians in the field.
A key point in the spectrum of technical involvement is deciding whether to learn to code. Mastering a programming language, especially Python, can separate those involved in development from those focused on implementation or solely clinical expertise. However, with the rise of large language models and their ability to translate ideas into code, traditional coding skills appear to be less essential. For this reason, it would not be advisable to commit the limited time you have to develop your career on learning a skill like programming as it may provide very little in return.
Still, if you’re generally interested, Python remains the most relevant and accessible major language in AI development. I personally found the University of Helsinki’s free Massive Open Online Course (MOOC) to be an excellent starting point.
Alternatively, you may want to act as a link between the medical and the technical teams. However, you should bear in mind that even if you don’t intend to become an expert in AI, lacking a conceptual grasp of how these systems work will limit your ability to translate your medical insight into any useful expertise, and potentially weaken the communication within your teams.
Education
Degrees
Pursuing a relevant degree is a major commitment but can be transformative. It has the potential to expand your knowledge and experience, introduce you to a new and relevant network, and generate new opportunities. However, a degree demands significant time, money and energy. You should already have a clear idea of what you want from a career in medical AI and be confident in how the specific degree would help move you towards it.
Examples of degrees focused on AI and digital health include:
- Imperial College London: MRes in Clinical Robotics and AI
- University of Birmingham: MSc in Artificial Intelligence Implementation
- University of Oxford: MSc in Applied Digital Health
Courses
If a full degree isn’t feasible, online courses provide flexible and affordable alternatives. While less formally recognised, they’re still valuable for building foundational knowledge.
Coursera is a popular choice, offering an introduction to key foundational concepts in AI. In particular, courses by Andrew Ng, while not specifically medical, provide an accessible entry point that will help you to better understand more applied topics.
Opportunities to develop your portfolio
Networking
If you’re starting from zero, collaboration is vital. It opens doors to projects and develops both your career and your contacts, which can snowball into greater opportunities.
Everyone will have different paths to collaboration, whether that is through existing contacts, institutional links or entirely new circles. Conferences and workshops that focus on AI in medicine are excellent starting points, as they often focus on networking. Presenting relevant posters or projects can amplify your visibility and invite meaningful engagement from other attendees.
LinkedIn is an alternative path which lets you directly identify and approach people of interest, although it may lack the personal touch which conferences can offer.
You should also look closer to home: trusts, hospitals and medical schools may have individuals or small teams exploring AI and digital health. Searching intranets, websites or simply asking around can uncover surprisingly accessible opportunities.
Mentorship
A mentor is one of the most valuable relationships you can cultivate. The right mentor offers tailored guidance, shares experience and connects you with opportunities otherwise out of reach. They may be a formal supervisor or someone you collaborate with who naturally becomes a mentor over time.
For informal mentors, consider people you’ve previously interacted with in a professional capacity with whom you had a good working relationship, whether that was during university, your career or other professional environments. While many mentors will view their role of supporting new generations altruistically, some will have their own projects to which you may be expected to contribute.
If you lack suitable contacts, proactive networking becomes key. Reach out to individuals in local trusts, universities or professional groups to offer help with ongoing projects. Conferences can also connect you with potential mentors and collaborators. Persistence matters here; many approaches will likely lead nowhere, but one meaningful connection can open major doors going forward.
Programmes
Structured fellowships and innovation programmes are an alternative way to meet the right people and get involved in the projects on which you want to work. Many of these have been set up in recent years to promote digital health and combine teaching, mentorship and project-based work to help clinicians apply technology to healthcare challenges.
For example:
- NHS Fellowship in Clinical AI note that this is limited to doctors on a training programme that leads directly to CCT e.g. registrars and run-through trainees (Eligibility criteria)
- Topol Digital Fellowship
- NHS Clinical Entrepreneur programme
- NHS Innovation Accelerator
- The Turing Institute’s Clinical AI group
Self-driven projects
Finally, if formal routes aren’t feasible, you may choose to create your own opportunities. A well-designed independent project can:
- Build relevant skills and experience
- Strengthen your portfolio for future roles
- Generate publishable work
- Provide material to present at conferences and attract collaborators
However, establishing an independent project demands time, discipline and resilience. Success is less certain without established support, so commit only if the project is manageable in scale, relevant to your clinical interests and within your skill set.
A well-chosen self-directed project can be transformative for your career and demonstrate commitment and perseverance that set you apart. However, a clear focus and realistic expectations are essential for success.
