What AI teams can learn from Medics

March 16, 2024

A friend, an engineer, once told me "Doctors are like mechanics and carpenters - they are not scientists". A close friend, a doctor, said "I agree!".

Doctors apply cutting edge research and science to achieve an outcome. They see themselves as experts making use of the best of what human science and technology can offer to treat patients. Doctors are not preoccupied with building things from scratch or working from the ground up - if not required. The stakes are too high.

AI teams should learn from this. Unless they are in research functions they should make maximum use of tools and services that others build and operate. Outside of research, AI teams should focus, like medics, on delivering a valuable and pressing outcome and using all the tools required to achieve that goal.

Think of a surgery - there's another lesson: a core deliverable of the medical profession can only happen in a cross functional team of different professions working hands on together. In a surgery there will be a surgeon, an anaesthetist, nurses and assistants. They all have different training, skills, and communities, and every day they deliver together. This is beyond "collaboration": they can only do their jobs when they are all operating together.

AI and Machine Learning experts too should organise their core deliverables by working in cross functional teams. More of the work of these professions should be done at the table with software engineers, developers, and others. The professional has a common training and a community, but should be organised to work embeded in teams of other related specialties. Their success is the extent to which the core job (and their role in it) is a success.

If you work in AI, be like medics: build less, buy/use more advanced tools, and operate together with the other specialists.