Implementing AI in Healthcare

Society of Automotive Engineers years back came up with a five-level framework for vehicle automation. It’s a clear roadmap that starts with zero automation and moves toward full autonomy, where cars operate without any human involvement. There are overlapping similarities between the automotive and healthcare industry. Both are highly regulated fields with life-or-death stakes and entrenched legacy systems. Yet, the auto industry has managed to establish a structured and safe path to integrating AI that can rival - and may eventually surpass - human drivers.

Modern cars already reach Level 2, with features like automated braking and steering, and high-end vehicles like Teslas can navigate highways on their own under human supervision. In select cities like San Francisco, Waymo has achieved Level 4 with driverless taxis operating in limited conditions. Level 5 - full autonomy everywhere - is within sight.

A similar five-level process could apply to AI in healthcare. At Level 1, AI handles administrative tasks, staying entirely out of patient care. Level 2 involves reducing the busywork that bogs down clinical professionals. At Level 3, AI aids doctors and nurses in decision-making, improving accuracy and efficiency with human oversight. Level 4 delves into the “black box” territory, where AI makes recommendations that may be trusted even without full transparency - high regulation, akin to medical devices, will be essential here. Level 5 imagines AI operating as independent medical practitioners, a future closer than we might think.

The risks of misdiagnosis by AI could be weighed against the enormous potential to make healthcare more accessible and affordable. The major problem isn’t that doctors are making bad decisions; it’s that people are locked out of care because they can’t afford it, or the system doesn’t have the capacity.

Such analogy might help reframing how we think about progress, trust, and the trade-offs required in the journey toward integrating AI into sensitive, impactful industries.